RFC 9055




Internet Engineering Task Force (IETF)                  E. Grossman, Ed.
Request for Comments: 9055                                         DOLBY
Category: Informational                                       T. Mizrahi
ISSN: 2070-1721                                                   HUAWEI
                                                               A. Hacker
                                                                 THOUGHT
                                                               June 2021


       Deterministic Networking (DetNet) Security Considerations

Abstract



   A DetNet (deterministic network) provides specific performance
   guarantees to its data flows, such as extremely low data loss rates
   and bounded latency (including bounded latency variation, i.e.,
   "jitter").  As a result, securing a DetNet requires that in addition
   to the best practice security measures taken for any mission-critical
   network, additional security measures may be needed to secure the
   intended operation of these novel service properties.

   This document addresses DetNet-specific security considerations from
   the perspectives of both the DetNet system-level designer and
   component designer.  System considerations include a taxonomy of
   relevant threats and attacks, and associations of threats versus use
   cases and service properties.  Component-level considerations include
   ingress filtering and packet arrival-time violation detection.

   This document also addresses security considerations specific to the
   IP and MPLS data plane technologies, thereby complementing the
   Security Considerations sections of those documents.

Status of This Memo



   This document is not an Internet Standards Track specification; it is
   published for informational purposes.

   This document is a product of the Internet Engineering Task Force
   (IETF).  It represents the consensus of the IETF community.  It has
   received public review and has been approved for publication by the
   Internet Engineering Steering Group (IESG).  Not all documents
   approved by the IESG are candidates for any level of Internet
   Standard; see Section 2 of RFC 7841.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   https://www.rfc-editor.org/info/rfc9055.

Copyright Notice



   Copyright (c) 2021 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (https://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.  Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents



   1.  Introduction
   2.  Abbreviations and Terminology
   3.  Security Considerations for DetNet Component Design
     3.1.  Resource Allocation
       3.1.1.  Inviolable Flows
       3.1.2.  Design Trade-Off Considerations in the Use Cases
               Continuum
       3.1.3.  Documenting the Security Properties of a Component
       3.1.4.  Fail-Safe Component Behavior
       3.1.5.  Flow Aggregation Example
     3.2.  Explicit Routes
     3.3.  Redundant Path Support
     3.4.  Timing (or Other) Violation Reporting
   4.  DetNet Security Considerations Compared with Diffserv Security
           Considerations
   5.  Security Threats
     5.1.  Threat Taxonomy
     5.2.  Threat Analysis
       5.2.1.  Delay
       5.2.2.  DetNet Flow Modification or Spoofing
       5.2.3.  Resource Segmentation (Inter-segment Attack)
               Vulnerability
       5.2.4.  Packet Replication and Elimination
         5.2.4.1.  Replication: Increased Attack Surface
         5.2.4.2.  Replication-Related Header Manipulation
       5.2.5.  Controller Plane
         5.2.5.1.  Path Choice Manipulation
         5.2.5.2.  Compromised Controller
       5.2.6.  Reconnaissance
       5.2.7.  Time-Synchronization Mechanisms
     5.3.  Threat Summary
   6.  Security Threat Impacts
     6.1.  Delay Attacks
       6.1.1.  Data Plane Delay Attacks
       6.1.2.  Controller Plane Delay Attacks
     6.2.  Flow Modification and Spoofing
       6.2.1.  Flow Modification
       6.2.2.  Spoofing
         6.2.2.1.  Data Plane Spoofing
         6.2.2.2.  Controller Plane Spoofing
     6.3.  Segmentation Attacks (Injection)
       6.3.1.  Data Plane Segmentation
       6.3.2.  Controller Plane Segmentation
     6.4.  Replication and Elimination
       6.4.1.  Increased Attack Surface
       6.4.2.  Header Manipulation at Elimination Routers
     6.5.  Control or Signaling Packet Modification
     6.6.  Control or Signaling Packet Injection
     6.7.  Reconnaissance
     6.8.  Attacks on Time-Synchronization Mechanisms
     6.9.  Attacks on Path Choice
   7.  Security Threat Mitigation
     7.1.  Path Redundancy
     7.2.  Integrity Protection
     7.3.  DetNet Node Authentication
     7.4.  Synthetic Traffic Insertion
     7.5.  Encryption
       7.5.1.  Encryption Considerations for DetNet
     7.6.  Control and Signaling Message Protection
     7.7.  Dynamic Performance Analytics
     7.8.  Mitigation Summary
   8.  Association of Attacks to Use Cases
     8.1.  Association of Attacks to Use Case Common Themes
       8.1.1.  Sub-network Layer
       8.1.2.  Central Administration
       8.1.3.  Hot Swap
       8.1.4.  Data Flow Information Models
       8.1.5.  L2 and L3 Integration
       8.1.6.  End-to-End Delivery
       8.1.7.  Replacement for Proprietary Fieldbuses and
               Ethernet-Based Networks
       8.1.8.  Deterministic vs. Best-Effort Traffic
       8.1.9.  Deterministic Flows
       8.1.10. Unused Reserved Bandwidth
       8.1.11. Interoperability
       8.1.12. Cost Reductions
       8.1.13. Insufficiently Secure Components
       8.1.14. DetNet Network Size
       8.1.15. Multiple Hops
       8.1.16. Level of Service
       8.1.17. Bounded Latency
       8.1.18. Low Latency
       8.1.19. Bounded Jitter (Latency Variation)
       8.1.20. Symmetrical Path Delays
       8.1.21. Reliability and Availability
       8.1.22. Redundant Paths
       8.1.23. Security Measures
     8.2.  Summary of Attack Types per Use Case Common Theme
   9.  Security Considerations for OAM Traffic
   10. DetNet Technology-Specific Threats
     10.1.  IP
     10.2.  MPLS
   11. IANA Considerations
   12. Security Considerations
   13. Privacy Considerations
   14. References
     14.1.  Normative References
     14.2.  Informative References
   Contributors

   Authors' Addresses



1.  Introduction



   A deterministic IP network ("Deterministic Networking Architecture"
   [RFC8655]) can carry data flows for real-time applications with
   extremely low data loss rates and bounded latency.  The bounds on
   latency defined by DetNet (as described in [RFC9016]) include both
   worst-case latency (Maximum Latency, Section 5.9.2 of [RFC9016]) and
   worst-case jitter (Maximum Latency Variation, Section 5.9.3 of
   [RFC9016]).  Data flows with deterministic properties are well
   established for Ethernet networks (see Time-Sensitive Networking
   (TSN), [IEEE802.1BA]); DetNet brings these capabilities to the IP
   network.

   Deterministic IP networks have been successfully deployed in real-
   time Operational Technology (OT) applications for some years;
   however, such networks are typically isolated from external access,
   and thus the security threat from external attackers is low.  An
   example of such an isolated network is a network deployed within an
   aircraft, which is "air gapped" from the outside world.  DetNet
   specifies a set of technologies that enable creation of deterministic
   flows on IP-based networks of a potentially wide area (on the scale
   of a corporate network), potentially merging OT traffic with best-
   effort Information Technology (IT) traffic, and placing OT network
   components into contact with IT network components, thereby exposing
   the OT traffic and components to security threats that were not
   present in an isolated OT network.

   These DetNet (OT-type) technologies may not have previously been
   deployed on a wide area IP-based network that also carries IT
   traffic, and thus they can present security considerations that may
   be new to IP-based wide area network designers; this document
   provides insight into such system-level security considerations.  In
   addition, designers of DetNet components (such as routers) face new
   security-related challenges in providing DetNet services, for
   example, maintaining reliable isolation between traffic flows in an
   environment where IT traffic co-mingles with critical reserved-
   bandwidth OT traffic; this document also examines security
   implications internal to DetNet components.

   Security is of particularly high importance in DetNet because many of
   the use cases that are enabled by DetNet [RFC8578] include control of
   physical devices (power grid devices, industrial controls, building
   controls, etc.) that can have high operational costs for failure and
   present potentially attractive targets for cyber attackers.

   This situation is even more acute given that one of the goals of
   DetNet is to provide a "converged network", i.e., one that includes
   both IT traffic and OT traffic, thus exposing potentially sensitive
   OT devices to attack in ways that were not previously common (usually
   because they were under a separate control system or otherwise
   isolated from the IT network, for example [ARINC664P7]).  Security
   considerations for OT networks are not a new area, and there are many
   OT networks today that are connected to wide area networks or the
   Internet; this document focuses on the issues that are specific to
   the DetNet technologies and use cases.

   Given the above considerations, securing a DetNet starts with a
   scrupulously well-designed and well-managed engineered network
   following industry best practices for security at both the data plane
   and controller plane, as well as for any Operations, Administration,
   and Maintenance (OAM) implementation; this is the assumed starting
   point for the considerations discussed herein.  Such assumptions also
   depend on the network components themselves upholding the security-
   related properties that are to be assumed by DetNet system-level
   designers; for example, the assumption that network traffic
   associated with a given flow can never affect traffic associated with
   a different flow is only true if the underlying components make it
   so.  Such properties, which may represent new challenges to component
   designers, are also considered herein.

   Starting with a "well-managed network", as noted above, enables us to
   exclude some of the more powerful adversary capabilities from the
   Internet Threat Model of [BCP72], such as the ability to arbitrarily
   drop or delay any or all traffic.  Given this reduced attacker
   capability, we can present security considerations based on attacker
   capabilities that are more directly relevant to a DetNet.

   In this context, we view the "conventional" (i.e., non-time-
   sensitive) network design and management aspects of network security
   as being primarily concerned with preventing denial of service, i.e.,
   they must ensure that DetNet traffic goes where it's supposed to and
   that an external attacker can't inject traffic that disrupts the
   delivery timing assurance of the DetNet.  The time-specific aspects
   of DetNet security presented here take up where those "conventional"
   design and management aspects leave off.

   However, note that "conventional" methods for mitigating (among all
   the others) denial-of-service attacks (such as throttling) can only
   be effectively used in a DetNet when their use does not compromise
   the required time-sensitive or behavioral properties required for the
   OT flows on the network.  For example, a "retry" protocol is
   typically not going to be compatible with a low-latency (worst-case
   maximum latency) requirement; however, if in a specific use case and
   implementation such a retry protocol is able to meet the timing
   constraints, then it may well be used in that context.  Similarly, if
   common security protocols such as TLS/DTLS or IPsec are to be used,
   it must be verified that their implementations are able to meet the
   timing and behavioral requirements of the time-sensitive network as
   implemented for the given use case.  An example of "behavioral
   properties" might be that dropping of more than a specific number of
   packets in a row is not acceptable according to the service level
   agreement.

   The exact security requirements for any given DetNet are necessarily
   specific to the use cases handled by that network.  Thus, the reader
   is assumed to be familiar with the specific security requirements of
   their use cases, for example, those outlined in the DetNet Use Cases
   [RFC8578] and the Security Considerations sections of the DetNet
   documents applicable to the network technologies in use, for example,
   [RFC8939] for an IP data plane and [RFC8964] for an MPLS data plane.
   Readers can find a general introduction to the DetNet Architecture in
   [RFC8655], the DetNet Data Plane in [RFC8938], and the Flow
   Information Model in [RFC9016].

   The DetNet technologies include ways to:

   *  Assign data plane resources for DetNet flows in some or all of the
      intermediate nodes (routers) along the path of the flow

   *  Provide explicit routes for DetNet flows that do not dynamically
      change with the network topology in ways that affect the quality
      of service received by the affected flow(s)

   *  Distribute data from DetNet flow packets over time and/or space to
      ensure delivery of the data in each packet in spite of the loss of
      a path

   This document includes sections considering DetNet component design
   as well as system design.  The latter includes a taxonomy and
   analysis of threats, threat impacts and mitigations, and an
   association of attacks with use cases (based on Section 11 of
   [RFC8578]).

   This document is based on the premise that there will be a very broad
   range of DetNet applications and use cases, ranging in size and scope
   from individual industrial machines to networks that span an entire
   country [RFC8578].  Thus, no single set of prescriptions (such as
   exactly which mitigation should be applied to which segment of a
   DetNet) can be applicable to all of them, and indeed any single one
   that we might prescribe would inevitably prove impractical for some
   use case, perhaps one that does not even exist at the time of this
   writing.  Thus, we are not prescriptive here; we are stating the
   desired end result, with the understanding that most DetNet use cases
   will necessarily differ from each other, and there is no "one size
   fits all".

2.  Abbreviations and Terminology



   Information Technology (IT):  The application of computers to store,
      study, retrieve, transmit, and manipulate data or information,
      often in the context of a business or other enterprise [IT-DEF].

   Operational Technology (OT):  The hardware and software dedicated to
      detecting or causing changes in physical processes through direct
      monitoring and/or control of physical devices such as valves,
      pumps, etc.  [OT-DEF].

   Component:  A component of a DetNet system -- used here to refer to
      any hardware or software element of a DetNet that implements
      DetNet-specific functionality, for example, all or part of a
      router, switch, or end system.

   Device:  Used here to refer to a physical entity controlled by the
      DetNet, for example, a motor.

   Resource Segmentation:  Used as a more general form for Network
      Segmentation (the act or practice of splitting a computer network
      into sub-networks, each being a network segment [NS-DEF]).

   Controller Plane:  In DetNet, the Controller Plane corresponds to the
      aggregation of the Control and Management Planes (see [RFC8655],
      Section 4.4.2).

3.  Security Considerations for DetNet Component Design



   This section provides guidance for implementers of components to be
   used in a DetNet.

   As noted above, DetNet provides resource allocation, explicit routes,
   and redundant path support.  Each of these has associated security
   implications, which are discussed in this section, in the context of
   component design.  Detection, reporting and appropriate action in the
   case of packet arrival-time violations are also discussed.

3.1.  Resource Allocation



3.1.1.  Inviolable Flows



   A DetNet system security designer relies on the premise that any
   resources allocated to a resource-reserved (OT-type) flow are
   inviolable; in other words, there is no physical possibility within a
   DetNet component that resources allocated to a given DetNet flow can
   be compromised by any type of traffic in the network.  This includes
   malicious traffic as well as inadvertent traffic such as might be
   produced by a malfunctioning component, or due to interactions
   between components that were not sufficiently tested for
   interoperability.  From a security standpoint, this is a critical
   assumption, for example, when designing against DoS attacks.  In
   other words, with correctly designed components and security
   mechanisms, one can prevent malicious activities from impacting other
   resources.

   However, achieving the goal of absolutely inviolable flows may not be
   technically or economically feasible for any given use case, given
   the broad range of possible use cases (e.g., [RFC8578]) and their
   associated security considerations as outlined in this document.  It
   can be viewed as a continuum of security requirements, from isolated
   ultra-low latency systems that may have little security vulnerability
   (such as an industrial machine) to broadly distributed systems with
   many possible attack vectors and OT security concerns (such as a
   utility network).  Given this continuum, the design principle
   employed in this document is to specify the desired end results,
   without being overly prescriptive in how the results are achieved,
   reflecting the understanding that no individual implementation is
   likely to be appropriate for every DetNet use case.

3.1.2.  Design Trade-Off Considerations in the Use Cases Continuum



   For any given DetNet use case and its associated security
   requirements, it is important for the DetNet system designer to
   understand the interaction and design trade-offs that inevitably need
   to be reconciled between the desired end results and the DetNet
   protocols, as well as the DetNet system and component design.

   For any given component, as designed for any given use case (or scope
   of use cases), it is the responsibility of the component designer to
   ensure that the premise of inviolable flows is supported to the
   extent that they deem necessary to support their target use cases.

   For example, the component may include traffic shaping and policing
   at the ingress to prevent corrupted, malicious, or excessive packets
   from entering the network, thereby decreasing the likelihood that any
   traffic will interfere with any DetNet OT flow.  The component may
   include integrity protection for some or all of the header fields
   such as those used for flow ID, thereby decreasing the likelihood
   that a packet whose flow ID has been compromised might be directed
   into a different flow path.  The component may verify every single
   packet header at every forwarding location, or only at certain
   points.  In any of these cases, the component may use dynamic
   performance analytics (Section 7.7) to cause action to be initiated
   to address the situation in an appropriate and timely manner, either
   at the data plane or controller plane, or both in concert.  The
   component's software and hardware may include measures to ensure the
   integrity of the resource allocation/deallocation process.  Other
   design aspects of the component may help ensure that the adverse
   effects of malicious traffic are more limited, for example, by
   protecting network control interfaces or minimizing cascade failures.
   The component may include features specific to a given use case, such
   as configuration of the response to a given sequential packet loss
   count.

   Ultimately, due to cost and complexity factors, the security
   properties of a component designed for low-cost systems may be (by
   design) far inferior to a component with similar intended
   functionality, but designed for highly secure or otherwise critical
   applications, perhaps at substantially higher cost.  Any given
   component is designed for some set of use cases and accordingly will
   have certain limitations on its security properties and
   vulnerabilities.  It is thus the responsibility of the system
   designer to assure themselves that the components they use in their
   design are capable of satisfying their overall system security
   requirements.

3.1.3.  Documenting the Security Properties of a Component



   In order for the system designer to adequately understand the
   security-related behavior of a given component, the designer of any
   component intended for use with DetNet needs to clearly document the
   security properties of that component.  For example, to address the
   case where a corrupted packet in which the flow identification
   information is compromised and thus may incidentally match the flow
   ID of another ("victim") DetNet flow, resulting in additional
   unauthorized traffic on the victim, the documentation might state
   that the component employs integrity protection on the flow
   identification fields.

3.1.4.  Fail-Safe Component Behavior



   Even when the security properties of a component are understood and
   well specified, if the component malfunctions, for example, due to
   physical circumstances unpredicted by the component designer, it may
   be difficult or impossible to fully prevent malfunction of the
   network.  The degree to which a component is hardened against various
   types of failures is a distinguishing feature of the component and
   its design, and the overall system design can only be as strong as
   its weakest link.

   However, all networks are subject to this level of uncertainty; it is
   not unique to DetNet.  Having said that, DetNet raises the bar by
   changing many added latency scenarios from tolerable annoyances to
   unacceptable service violations.  That in turn underscores the
   importance of system integrity, as well as correct and stable
   configuration of the network and its nodes, as discussed in
   Section 1.

3.1.5.  Flow Aggregation Example



   As another example regarding resource allocation implementation,
   consider the implementation of Flow Aggregation for DetNet flows (as
   discussed in [RFC8938]).  In this example, say there are N flows that
   are to be aggregated; thus, the bandwidth resources of the aggregate
   flow must be sufficient to contain the sum of the bandwidth
   reservation for the N flows.  However, if one of those flows were to
   consume more than its individually allocated bandwidth, this could
   cause starvation of the other flows.  Thus, simply providing and
   enforcing the calculated aggregate bandwidth may not be a complete
   solution; the bandwidth for each individual flow must still be
   guaranteed, for example, via ingress policing of each flow (i.e.,
   before it is aggregated).  Alternatively, if by some other means each
   flow to be aggregated can be trusted not to exceed its allocated
   bandwidth, the same goal can be achieved.

3.2.  Explicit Routes



   The DetNet-specific purpose for constraining the ability of the
   DetNet to reroute OT traffic is to maintain the specified service
   parameters (such as upper and lower latency boundaries) for a given
   flow.  For example, if the network were to reroute a flow (or some
   part of a flow) based exclusively on statistical path usage metrics,
   or due to malicious activity, it is possible that the new path would
   have a latency that is outside the required latency bounds that were
   designed into the original TE-designed path, thereby violating the
   quality of service for the affected flow (or part of that flow).

   However, it is acceptable for the network to reroute OT traffic in
   such a way as to maintain the specified latency bounds (and any other
   specified service properties) for any reason, for example, in
   response to a runtime component or path failure.

   So from a DetNet security standpoint, the DetNet system designer can
   expect that any component designed for use in a DetNet will deliver
   the packets within the agreed-upon service parameters.  For the
   component designer, this means that in order for a component to
   achieve that expectation, any component that is involved in
   controlling or implementing any change of the initially TE-configured
   flow routes must prevent rerouting of OT flows (whether malicious or
   accidental) that might adversely affect delivering the traffic within
   the specified service parameters.

3.3.  Redundant Path Support



   The DetNet provision for redundant paths (i.e., PREOF, or "Packet
   Replication, Elimination, and Ordering Functions"), as defined in the
   DetNet Architecture [RFC8655], provides the foundation for high
   reliability of a DetNet by virtually eliminating packet loss (i.e.,
   to a degree that is implementation dependent) through hitless
   redundant packet delivery.

      |  Note: At the time of this writing, PREOF is not defined for the
      |  IP data plane.

   It is the responsibility of the system designer to determine the
   level of reliability required by their use case and to specify
   redundant paths sufficient to provide the desired level of
   reliability (in as much as that reliability can be provided through
   the use of redundant paths).  It is the responsibility of the
   component designer to ensure that the relevant PREOF operations are
   executed reliably and securely to avoid potentially catastrophic
   situations for the operational technology relying on them.

   However, note that not all PREOF operations are necessarily
   implemented in every network; for example, a packet reordering
   function may not be necessary if the packets are either not required
   to be in order or if the ordering is performed in some other part of
   the network.

   Ideally, a redundant path for a flow could be specified from end to
   end; however, given that this is not always possible (as described in
   [RFC8655]), the system designer will need to consider the resulting
   end-to-end reliability and security resulting from any given
   arrangement of network segments along the path, each of which
   provides its individual PREOF implementation and thus its individual
   level of reliability and security.

   At the data plane, the implementation of PREOF depends on the correct
   assignment and interpretation of packet sequence numbers, as well as
   the actions taken based on them, such as elimination (including
   elimination of packets with spurious sequence numbers).  Thus, the
   integrity of these values must be maintained by the component as they
   are assigned by the DetNet Data Plane Service sub-layer and
   transported by the Forwarding sub-layer.  This is no different than
   the integrity of the values in any header used by the DetNet (or any
   other) data plane and is not unique to redundant paths.  The
   integrity protection of header values is technology dependent; for
   example, in Layer 2 networks, the integrity of the header fields can
   be protected by using MACsec [IEEE802.1AE-2018].  Similarly, from the
   sequence number injection perspective, it is no different from any
   other protocols that use sequence numbers; for particulars of
   integrity protection via IPsec Authentication Headers, useful
   insights are provided by Section 3 of [RFC4302].

3.4.  Timing (or Other) Violation Reporting



   A task of the DetNet system designer is to create a network such that
   for any incoming packet that arrives with any timing or bandwidth
   violation, an appropriate action can be taken in order to prevent
   damage to the system.  The reporting step may be accomplished through
   dynamic performance analysis (see Section 7.7) or by any other means
   as implemented in one or more components.  The action to be taken for
   any given circumstance within any given application will depend on
   the use case.  The action may involve intervention from the
   controller plane, or it may be taken "immediately" by an individual
   component, for example, if a very fast response is required.

   The definitions and selections of the actions that can be taken are
   properties of the components.  The component designer implements
   these options according to their expected use cases, which may vary
   widely from component to component.  Clearly, selecting an
   inappropriate response to a given condition may cause more problems
   than it is intending to mitigate; for example, a naive approach might
   be to have the component shut down the link if a packet arrives
   outside of its prescribed time window.  However, such a simplistic
   action may serve the attacker better than it serves the network.
   Similarly, simple logging of such issues may not be adequate since a
   delay in response could result in material damage, for example, to
   mechanical devices controlled by the network.  Thus, a breadth of
   possible and effective security-related actions and their
   configuration is a positive attribute for a DetNet component.

   Some possible violations that warrant detection include cases where a
   packet arrives:

   *  Outside of its prescribed time window

   *  Within its time window but with a compromised timestamp that makes
      it appear that it is not within its window

   *  Exceeding the reserved flow bandwidth

   Some possible direct actions that may be taken at the data plane
   include traffic policing and shaping functions (e.g., those described
   in [RFC2475]), separating flows into per-flow rate-limited queues,
   and potentially applying active queue management [RFC7567].  However,
   if those (or any other) actions are to be taken, the system designer
   must ensure that the results of such actions do not compromise the
   continued safe operation of the system.  For example, the network
   (i.e., the controller plane and data plane working together) must
   mitigate in a timely fashion any potential adverse effect on
   mechanical devices controlled by the network.

4.  DetNet Security Considerations Compared with Diffserv Security
    Considerations



   DetNet is designed to be compatible with Diffserv [RFC2474] as
   applied to IT traffic in the DetNet.  DetNet also incorporates the
   use of the 6-bit value of the Differentiated Services Code Point
   (DSCP) field of the Type of Service (IPv4) and Traffic Class (IPv6)
   bytes for flow identification.  However, the DetNet interpretation of
   the DSCP value for OT traffic is not equivalent to the per-hop
   behavior (PHB) selection behavior as defined by Diffserv.

   Thus, security considerations for DetNet have some aspects in common
   with Diffserv, in fact overlapping 100% with respect to IP IT
   traffic.  Security considerations for these aspects are part of the
   existing literature on IP network security, specifically the Security
   Considerations sections of [RFC2474] and [RFC2475].  However, DetNet
   also introduces timing and other considerations that are not present
   in Diffserv, so the Diffserv security considerations are a subset of
   the DetNet security considerations.

   In the case of DetNet OT traffic, the DSCP value is interpreted
   differently than in Diffserv and contributes to determination of the
   service provided to the packet.  In DetNet, there are similar
   consequences to Diffserv for lack of detection of, or incorrect
   handling of, packets with mismarked DSCP values, and many of the
   points made in the Diffserv Security discussions (Section 6.1 of
   [RFC2475], Section 7 of [RFC2474], and Section 3.3.2.1 of [RFC6274])
   are also relevant to DetNet OT traffic though perhaps in modified
   form.  For example, in DetNet, the effect of an undetected or
   incorrectly handled maliciously mismarked DSCP field in an OT packet
   is not identical to affecting the PHB of that packet, since DetNet
   does not use the PHB concept for OT traffic.  Nonetheless, the
   service provided to the packet could be affected, so mitigation
   measures analogous to those prescribed by Diffserv would be
   appropriate for DetNet.  For example, mismarked DSCP values should
   not cause failure of network nodes.  The remarks in [RFC2474]
   regarding IPsec and Tunneling Interactions are also relevant (though
   this is not to say that other sections are less relevant).

   In this discussion, interpretation (and any possible intentional re-
   marking) of the DSCP values of packets destined for DetNet OT flows
   is expected to occur at the ingress to the DetNet domain; once inside
   the domain, maintaining the integrity of the DSCP values is subject
   to the same handling considerations as any other field in the packet.

5.  Security Threats



   This section presents a taxonomy of threats and analyzes the possible
   threats in a DetNet-enabled network.  The threats considered in this
   section are independent of any specific technologies used to
   implement the DetNet; Section 10 considers attacks that are
   associated with the DetNet technologies encompassed by [RFC8938].

   We distinguish controller plane threats from data plane threats.  The
   attack surface may be the same, but the types of attacks, as well as
   the motivation behind them, are different.  For example, a Delay
   attack is more relevant to the data plane than to the controller
   plane.  There is also a difference in terms of security solutions;
   the way you secure the data plane is often different than the way you
   secure the controller plane.

5.1.  Threat Taxonomy



   This document employs organizational elements of the threat models of
   [RFC7384] and [RFC7835].  This model classifies attackers based on
   two criteria:

   Internal vs. external:
      Internal attackers either have access to a trusted segment of the
      network or possess the encryption or authentication keys.
      External attackers, on the other hand, do not have the keys and
      have access only to the encrypted or authenticated traffic.

   On-path vs. off-path:
      On-path attackers are located in a position that allows
      interception, modification, or dropping of in-flight protocol
      packets, whereas off-path attackers can only attack by generating
      protocol packets.

   Regarding the boundary between internal vs. external attackers as
   defined above, note that in this document we do not make concrete
   recommendations regarding which specific segments of the network are
   to be protected in any specific way, for example, via encryption or
   authentication.  As a result, the boundary as defined above is not
   unequivocally specified here.  Given that constraint, the reader can
   view an internal attacker as one who can operate within the perimeter
   defined by the DetNet Edge Nodes (as defined in the DetNet
   Architecture [RFC8655]), allowing that the specifics of what is
   encrypted or authenticated within this perimeter will vary depending
   on the implementation.

   Care has also been taken to adhere to Section 5 of [RFC3552], both
   with respect to which attacks are considered out of scope for this
   document, and also which are considered to be the most common threats
   (explored further in Section 5.2).  Most of the direct threats to
   DetNet are active attacks (i.e., attacks that modify DetNet traffic),
   but it is highly suggested that DetNet application developers take
   appropriate measures to protect the content of the DetNet flows from
   passive attacks (i.e., attacks that observe but do not modify DetNet
   traffic), for example, through the use of TLS or DTLS.

   DetNet-Service, one of the service scenarios described in
   [DETNET-SERVICE-MODEL], is the case where a service connects DetNet
   islands, i.e., two or more otherwise independent DetNets are
   connected via a link that is not intrinsically part of either
   network.  This implies that there could be DetNet traffic flowing
   over a non-DetNet link, which may provide an attacker with an
   advantageous opportunity to tamper with DetNet traffic.  The security
   properties of non-DetNet links are outside of the scope of DetNet
   Security, but it should be noted that use of non-DetNet services to
   interconnect DetNets merits security analysis to ensure the integrity
   of the networks involved.

5.2.  Threat Analysis



5.2.1.  Delay



   An attacker can maliciously delay DetNet data flow traffic.  By
   delaying the traffic, the attacker can compromise the service of
   applications that are sensitive to high delays or to high delay
   variation.  The delay may be constant or modulated.

5.2.2.  DetNet Flow Modification or Spoofing



   An attacker can modify some header fields of en route packets in a
   way that causes the DetNet flow identification mechanisms to
   misclassify the flow.  Alternatively, the attacker can inject traffic
   that is tailored to appear as if it belongs to a legitimate DetNet
   flow.  The potential consequence is that the DetNet flow resource
   allocation cannot guarantee the performance that is expected when the
   flow identification works correctly.

5.2.3.  Resource Segmentation (Inter-segment Attack) Vulnerability



   DetNet components are expected to split their resources between
   DetNet flows in a way that prevents traffic from one DetNet flow from
   affecting the performance of other DetNet flows and also prevents
   non-DetNet traffic from affecting DetNet flows.  However, perhaps due
   to implementation constraints, some resources may be partially
   shared, and an attacker may try to exploit this property.  For
   example, an attacker can inject traffic in order to exhaust network
   resources such that DetNet packets that share resources with the
   injected traffic may be dropped or delayed.  Such injected traffic
   may be part of DetNet flows or non-DetNet traffic.

   Another example of a Resource Segmentation attack is the case in
   which an attacker is able to overload the exception path queue on the
   router, i.e., a "slow path" typically taken by control or OAM packets
   that are diverted from the data plane because they require processing
   by a CPU.  DetNet OT flows are typically configured to take the "fast
   path" through the data plane to minimize latency.  However, if there
   is only one queue from the forwarding Application-Specific Integrated
   Circuit (ASIC) to the exception path, and for some reason the system
   is configured such that any DetNet packets must be handled on this
   exception path, then saturating the exception path could result in
   the delaying or dropping of DetNet packets.

5.2.4.  Packet Replication and Elimination



5.2.4.1.  Replication: Increased Attack Surface



   Redundancy is intended to increase the robustness and survivability
   of DetNet flows, and replication over multiple paths can potentially
   mitigate an attack that is limited to a single path.  However, the
   fact that packets are replicated over multiple paths increases the
   attack surface of the network, i.e., there are more points in the
   network that may be subject to attacks.

5.2.4.2.  Replication-Related Header Manipulation



   An attacker can manipulate the replication-related header fields.
   This capability opens the door for various types of attacks.  For
   example:

   Forward both replicas:
      Malicious change of a packet SN (Sequence Number) can cause both
      replicas of the packet to be forwarded.  Note that this attack has
      a similar outcome to a replay attack.

   Eliminate both replicas:
      SN manipulation can be used to cause both replicas to be
      eliminated.  In this case, an attacker that has access to a single
      path can cause packets from other paths to be dropped, thus
      compromising some of the advantage of path redundancy.

   Flow hijacking:
      An attacker can hijack a DetNet flow with access to a single path
      by systematically replacing the SNs on the given path with higher
      SN values.  For example, an attacker can replace every SN value S
      with a higher value S+C, where C is a constant integer.  Thus, the
      attacker creates a false illusion that the attacked path has the
      lowest delay, causing all packets from other paths to be
      eliminated in favor of the attacked path.  Once the flow from the
      compromised path is favored by the eliminating bridge, the flow
      has effectively been hijacked by the attacker.  It is now possible
      for the attacker to either replace en route packets with malicious
      packets, or to simply inject errors into the packets, causing the
      packets to be dropped at their destination.

   Amplification:
      An attacker who injects packets into a flow that is to be
      replicated will have their attack amplified through the
      replication process.  This is no different than any attacker who
      injects packets that are delivered through multicast, broadcast,
      or other point-to-multi-point mechanisms.

5.2.5.  Controller Plane



5.2.5.1.  Path Choice Manipulation



5.2.5.1.1.  Control or Signaling Packet Modification


   An attacker can maliciously modify en route control packets in order
   to disrupt or manipulate the DetNet path/resource allocation.

5.2.5.1.2.  Control or Signaling Packet Injection


   An attacker can maliciously inject control packets in order to
   disrupt or manipulate the DetNet path/resource allocation.

5.2.5.1.3.  Increased Attack Surface


   One of the possible consequences of a Path Manipulation attack is an
   increased attack surface.  Thus, when the attack described in the
   previous subsection is implemented, it may increase the potential of
   other attacks to be performed.

5.2.5.2.  Compromised Controller



   An attacker can subvert a legitimate controller (or subvert another
   component such that it represents itself as a legitimate controller)
   with the result that the network nodes incorrectly believe it is
   authorized to instruct them.

   The presence of a compromised node or controller in a DetNet is not a
   threat that arises as a result of determinism or time sensitivity;
   the same techniques used to prevent or mitigate against compromised
   nodes in any network are equally applicable in the DetNet case.  The
   act of compromising a controller may not even be within the
   capabilities of our defined attacker types -- in other words, it may
   not be achievable via packet traffic at all, whether internal or
   external, on path or off path.  It might be accomplished, for
   example, by a human with physical access to the component, who could
   upload bogus firmware to it via a USB stick.  All of this underscores
   the requirement for careful overall system security design in a
   DetNet, given that the effects of even one bad actor on the network
   can be potentially catastrophic.

   Security concerns specific to any given controller plane technology
   used in DetNet will be addressed by the DetNet documents associated
   with that technology.

5.2.6.  Reconnaissance



   A passive eavesdropper can identify DetNet flows and then gather
   information about en route DetNet flows, e.g., the number of DetNet
   flows, their bandwidths, their schedules, or other temporal or
   statistical properties.  The gathered information can later be used
   to invoke other attacks on some or all of the flows.

   DetNet flows are typically uniquely identified by their 6-tuple,
   i.e., fields within the L3 or L4 header.  However, in some
   implementations, the flow ID may also be augmented by additional per-
   flow attributes known to the system, e.g., above L4.  For the purpose
   of this document, we assume any such additional fields used for flow
   ID are encrypted and/or integrity protected from external attackers.
   Note however that existing OT protocols designed for use on dedicated
   secure networks may not intrinsically provide such protection, in
   which case IPsec or transport-layer security mechanisms may be
   needed.

5.2.7.  Time-Synchronization Mechanisms



   An attacker can use any of the attacks described in [RFC7384] to
   attack the synchronization protocol, thus affecting the DetNet
   service.

5.3.  Threat Summary



   A summary of the attacks that were discussed in this section is
   presented in Table 1.  For each attack, the table specifies the type
   of attackers that may invoke the attack.  In the context of this
   summary, the distinction between internal and external attacks is
   under the assumption that a corresponding security mechanism is being
   used, and that the corresponding network equipment takes part in this
   mechanism.

    +======================+=========================================+
    |        Attack        |              Attacker Type              |
    |                      +====================+====================+
    |                      |      Internal      |      External      |
    |                      +=========+==========+=========+==========+
    |                      | On-Path | Off-Path | On-Path | Off-Path |
    +======================+=========+==========+=========+==========+
    | Delay Attack         |    +    |          |    +    |          |
    +----------------------+---------+----------+---------+----------+
    | DetNet Flow          |    +    |    +     |         |          |
    | Modification or      |         |          |         |          |
    | Spoofing             |         |          |         |          |
    +----------------------+---------+----------+---------+----------+
    | Inter-segment Attack |    +    |    +     |    +    |    +     |
    +----------------------+---------+----------+---------+----------+
    | Replication:         |    +    |    +     |    +    |    +     |
    | Increased Attack     |         |          |         |          |
    | Surface              |         |          |         |          |
    +----------------------+---------+----------+---------+----------+
    | Replication-Related  |    +    |          |         |          |
    | Header Manipulation  |         |          |         |          |
    +----------------------+---------+----------+---------+----------+
    | Path Manipulation    |    +    |    +     |         |          |
    +----------------------+---------+----------+---------+----------+
    | Path Choice:         |    +    |    +     |    +    |    +     |
    | Increased Attack     |         |          |         |          |
    | Surface              |         |          |         |          |
    +----------------------+---------+----------+---------+----------+
    | Control or Signaling |    +    |          |         |          |
    | Packet Modification  |         |          |         |          |
    +----------------------+---------+----------+---------+----------+
    | Control or Signaling |    +    |    +     |         |          |
    | Packet Injection     |         |          |         |          |
    +----------------------+---------+----------+---------+----------+
    | Reconnaissance       |    +    |          |    +    |          |
    +----------------------+---------+----------+---------+----------+
    | Attacks on Time-     |    +    |    +     |    +    |    +     |
    | Synchronization      |         |          |         |          |
    | Mechanisms           |         |          |         |          |
    +----------------------+---------+----------+---------+----------+

                     Table 1: Threat Analysis Summary

6.  Security Threat Impacts



   When designing security for a DetNet, as with any network, it may be
   prohibitively expensive or technically infeasible to thoroughly
   protect against every possible threat.  Thus, the security designer
   must be informed (for example, by an application domain expert such
   as a product manager) regarding the relative significance of the
   various threats and their impact if a successful attack is carried
   out.  In this section, we present an example of a possible template
   for such a communication, culminating in a table (Table 2) that lists
   a set of threats under consideration, and some values characterizing
   their relative impact in the context of a given industry.  The
   specific threats, industries, and impact values in the table are
   provided only as an example of this kind of assessment and its
   communication; they are not intended to be taken literally.

   This section considers assessment of the relative impacts of the
   attacks described in Section 5.  In this section, the impacts as
   described assume that the associated mitigation is not present or has
   failed.  Mitigations are discussed in Section 7.

   In computer security, the impact (or consequence) of an incident can
   be measured in loss of confidentiality, integrity, or availability of
   information.  In the case of OT or time sensitive networks (though
   not to the exclusion of IT or non-time-sensitive networks), the
   impact of an exploit can also include failure or malfunction of
   mechanical and/or other physical systems.

   DetNet raises these stakes significantly for OT applications,
   particularly those that may have been designed to run in an OT-only
   environment and thus may not have been designed for security in an IT
   environment with its associated components, services, and protocols.

   The extent of impact of a successful vulnerability exploit varies
   considerably by use case and by industry; additional insight
   regarding the individual use cases is available from "Deterministic
   Networking Use Cases" [RFC8578].  Each of those use cases is
   represented in Table 2, including Pro Audio, Electrical Utilities,
   Industrial M2M (split into two areas: M2M Data Gathering and M2M
   Control Loop), and others.

   Aspects of Impact (left column) include Criticality of Failure,
   Effects of Failure, Recovery, and DetNet Functional Dependence.
   Criticality of failure summarizes the seriousness of the impact.  The
   impact of a resulting failure can affect many different metrics that
   vary greatly in scope and severity.  In order to reduce the number of
   variables, only the following were included: Financial, Health and
   Safety, Effect on a Single Organization, and Effect on Multiple
   Organizations.  Recovery outlines how long it would take for an
   affected use case to get back to its pre-failure state (Recovery Time
   Objective, RTO) and how much of the original service would be lost in
   between the time of service failure and recovery to original state
   (Recovery Point Objective, RPO).  DetNet dependence maps how much the
   following DetNet service objectives contribute to impact of failure:
   time dependency, data integrity, source node integrity, availability,
   and latency/jitter.

   The scale of the Impact mappings is low, medium, and high.  In some
   use cases, there may be a multitude of specific applications in which
   DetNet is used.  For simplicity, this section attempts to average the
   varied impacts of different applications.  This section does not
   address the overall risk of a certain impact that would require the
   likelihood of a failure happening.

   In practice, any such ratings will vary from case to case; the
   ratings shown here are given as examples.

   +==============+=====+======+======+==========+======+======+======+
   |              | PRO | Util | Bldg | Wireless | Cell | M2M  | M2M  |
   |              | A   |      |      |          |      | Data | Ctrl |
   +==============+=====+======+======+==========+======+======+======+
   | Criticality  | Med | Hi   | Low  | Med      | Med  | Med  | Med  |
   +==============+=====+======+======+==========+======+======+======+
   | Effects                                                          |
   +==============+=====+======+======+==========+======+======+======+
   | Financial    | Med | Hi   | Med  | Med      | Low  | Med  | Med  |
   +--------------+-----+------+------+----------+------+------+------+
   | Health/      | Med | Hi   | Hi   | Med      | Med  | Med  | Med  |
   | Safety       |     |      |      |          |      |      |      |
   +--------------+-----+------+------+----------+------+------+------+
   | Affects 1    | Hi  | Hi   | Med  | Hi       | Med  | Med  | Med  |
   | org          |     |      |      |          |      |      |      |
   +--------------+-----+------+------+----------+------+------+------+
   | Affects >1   | Med | Hi   | Low  | Med      | Med  | Med  | Med  |
   | org          |     |      |      |          |      |      |      |
   +==============+=====+======+======+==========+======+======+======+
   | Recovery                                                         |
   +==============+=====+======+======+==========+======+======+======+
   | Recov Time   | Med | Hi   | Med  | Hi       | Hi   | Hi   | Hi   |
   | Obj          |     |      |      |          |      |      |      |
   +--------------+-----+------+------+----------+------+------+------+
   | Recov Point  | Med | Hi   | Low  | Med      | Low  | Hi   | Hi   |
   | Obj          |     |      |      |          |      |      |      |
   +==============+=====+======+======+==========+======+======+======+
   | DetNet Dependence                                                |
   +==============+=====+======+======+==========+======+======+======+
   | Time         | Hi  | Hi   | Low  | Hi       | Med  | Low  | Hi   |
   | Dependence   |     |      |      |          |      |      |      |
   +--------------+-----+------+------+----------+------+------+------+
   | Latency/     | Hi  | Hi   | Med  | Med      | Low  | Low  | Hi   |
   | Jitter       |     |      |      |          |      |      |      |
   +--------------+-----+------+------+----------+------+------+------+
   | Data         | Hi  | Hi   | Med  | Hi       | Low  | Hi   | Hi   |
   | Integrity    |     |      |      |          |      |      |      |
   +--------------+-----+------+------+----------+------+------+------+
   | Src Node     | Hi  | Hi   | Med  | Hi       | Med  | Hi   | Hi   |
   | Integ        |     |      |      |          |      |      |      |
   +--------------+-----+------+------+----------+------+------+------+
   | Availability | Hi  | Hi   | Med  | Hi       | Low  | Hi   | Hi   |
   +--------------+-----+------+------+----------+------+------+------+

             Table 2: Impact of Attacks by Use Case Industry

   The rest of this section will cover impact of the different groups in
   more detail.

6.1.  Delay Attacks



6.1.1.  Data Plane Delay Attacks



   Note that "Delay attack" also includes the possibility of a "negative
   delay" or early arrival of a packet, or possibly adversely changing
   the timestamp value.

   Delayed messages in a DetNet link can result in the same behavior as
   dropped messages in ordinary networks, since the services attached to
   the DetNet flow are likely to have strict delivery time requirements.

   For a single-path scenario, disruption within the single flow is a
   real possibility.  In a multipath scenario, large delays or
   instabilities in one DetNet flow can also lead to increased buffer
   and processor resource consumption at the eliminating router.

   A data plane Delay attack on a system controlling substantial moving
   devices, for example, in industrial automation, can cause physical
   damage.  For example, if the network promises a bounded latency of 2
   ms for a flow, yet the machine receives it with 5 ms latency, the
   control loop of the machine may become unstable.

6.1.2.  Controller Plane Delay Attacks



   In and of itself, this is not directly a threat to the DetNet
   service, but the effects of delaying control messages can have quite
   adverse effects later.

   *  Delayed teardown can lead to resource leakage, which in turn can
      result in failure to allocate new DetNet flows, finally giving
      rise to a denial-of-service attack.

   *  Failure to deliver, or severely delaying, controller plane
      messages adding an endpoint to a multicast group will prevent the
      new endpoint from receiving expected frames thus disrupting
      expected behavior.

   *  Delaying messages that remove an endpoint from a group can lead to
      loss of privacy, as the endpoint will continue to receive messages
      even after it is supposedly removed.

6.2.  Flow Modification and Spoofing



6.2.1.  Flow Modification



   If the contents of a packet header or body can be modified by the
   attacker, this can cause the packet to be routed incorrectly or
   dropped, or the payload to be corrupted or subtly modified.  Thus,
   the potential impact of a Modification attack includes disrupting the
   application as well as the network equipment.

6.2.2.  Spoofing



6.2.2.1.  Data Plane Spoofing



   Spoofing data plane messages can result in increased resource
   consumption on the routers throughout the network as it will increase
   buffer usage and processor utilization.  This can lead to resource
   exhaustion and/or increased delay.

   If the attacker manages to create valid headers, the false messages
   can be forwarded through the network, using part of the allocated
   bandwidth.  This in turn can cause legitimate messages to be dropped
   when the resource budget has been exhausted.

   Finally, the endpoint will have to deal with invalid messages being
   delivered to the endpoint instead of (or in addition to) a valid
   message.

6.2.2.2.  Controller Plane Spoofing



   A successful Controller Plane Spoofing attack will potentially have
   adverse effects.  It can do virtually anything from:

   *  modifying existing DetNet flows by changing the available
      bandwidth

   *  adding or removing endpoints from a DetNet flow

   *  dropping DetNet flows completely

   *  falsely creating new DetNet flows (exhausting the systems
      resources or enabling DetNet flows that are outside the control of
      the network engineer)

6.3.  Segmentation Attacks (Injection)



6.3.1.  Data Plane Segmentation



   Injection of false messages in a DetNet flow could lead to exhaustion
   of the available bandwidth for that flow if the routers attribute
   these false messages to the resource budget of that flow.

   In a multipath scenario, injected messages will cause increased
   processor utilization in elimination routers.  If enough paths are
   subject to malicious injection, the legitimate messages can be
   dropped.  Likewise, it can cause an increase in buffer usage.  In
   total, it will consume more resources in the routers than normal,
   giving rise to a resource-exhaustion attack on the routers.

   If a DetNet flow is interrupted, the end application will be affected
   by what is now a non-deterministic flow.  Note that there are many
   possible sources of flow interruptions, for example, but not limited
   to, such physical-layer conditions as a broken wire or a radio link
   that is compromised by interference.

6.3.2.  Controller Plane Segmentation



   In a successful Controller Plane Segmentation attack, control
   messages are acted on by nodes in the network, unbeknownst to the
   central controller or the network engineer.  This has the potential
   to:

   *  create new DetNet flows (exhausting resources)

   *  drop existing DetNet flows (denial of service)

   *  add end stations to a multicast group (loss of privacy)

   *  remove end stations from a multicast group (reduction of service)

   *  modify the DetNet flow attributes (affecting available bandwidth)

   If an attacker can inject control messages without the central
   controller knowing, then one or more components in the network may
   get into a state that is not expected by the controller.  At that
   point, if the controller initiates a command, the effect of that
   command may not be as expected, since the target of the command may
   have started from a different initial state.

6.4.  Replication and Elimination



   The Replication and Elimination functions are relevant only to data
   plane messages as controller plane messages are not subject to
   multipath routing.

6.4.1.  Increased Attack Surface



   The impact of an increased attack surface is that it increases the
   probability that the network can be exposed to an attacker.  This can
   facilitate a wide range of specific attacks, and their respective
   impacts are discussed in other subsections of this section.

6.4.2.  Header Manipulation at Elimination Routers



   This attack can potentially cause DoS to the application that uses
   the attacked DetNet flows or to the network equipment that forwards
   them.  Furthermore, it can allow an attacker to manipulate the
   network paths and the behavior of the network layer.

6.5.  Control or Signaling Packet Modification



   If control packets are subject to manipulation undetected, the
   network can be severely compromised.

6.6.  Control or Signaling Packet Injection



   If an attacker can inject control packets undetected, the network can
   be severely compromised.

6.7.  Reconnaissance



   Of all the attacks, this is one of the most difficult to detect and
   counter.

   An attacker can, at their leisure, observe over time various aspects
   of the messaging and signaling, learning the intent and purpose of
   the traffic flows.  Then at some later date, possibly at an important
   time in the operational context, they might launch an attack based on
   that knowledge.

   The flow ID in the header of the data plane messages gives an
   attacker a very reliable identifier for DetNet traffic, and this
   traffic has a high probability of going to lucrative targets.

   Applications that are ported from a private OT network to the higher
   visibility DetNet environment may need to be adapted to limit
   distinctive flow properties that could make them susceptible to
   reconnaissance.

6.8.  Attacks on Time-Synchronization Mechanisms



   DetNet relies on an underlying time-synchronization mechanism;
   therefore, a compromised synchronization mechanism may cause DetNet
   nodes to malfunction.  Specifically, DetNet flows may fail to meet
   their latency requirements and deterministic behavior, thus causing
   DoS to DetNet applications.

6.9.  Attacks on Path Choice



   This is covered in part in Section 6.3 (Segmentation Attacks
   (Injection)) and, as with Replication and Elimination (see
   Section 6.4), this is relevant for data plane messages.

7.  Security Threat Mitigation



   This section describes a set of measures that can be taken to
   mitigate the attacks described in Section 5.  These mitigations
   should be viewed as a set of tools, any of which can be used
   individually or in concert.  The DetNet component and/or system and/
   or application designer can apply these tools as necessary based on a
   system-specific threat analysis.

   Some of the technology-specific security considerations and
   mitigation approaches are further discussed in DetNet data plane
   solution documents, such as [RFC8938], [RFC8939], [RFC8964],
   [RFC9025], and [RFC9056].

7.1.  Path Redundancy



   Description:  Path redundancy is a DetNet flow that can be forwarded
      simultaneously over multiple paths.  Packet Replication and
      Elimination [RFC8655] provide resiliency to dropped or delayed
      packets.  This redundancy improves the robustness to failures and
      to on-path attacks.

         |  Note: At the time of this writing, PREOF is not defined for
         |  the IP data plane.

   Related attacks:  Path redundancy can be used to mitigate various on-
      path attacks, including attacks described in Sections 5.2.1,
      5.2.2, 5.2.3, and 5.2.7.  However, it is also possible that
      multiple paths may make it more difficult to locate the source of
      an on-path attacker.

      A Delay Modulation attack could result in extensively exercising
      otherwise unused code paths to expose hidden flaws.  Subtle race
      conditions and memory allocation bugs in error-handling paths are
      classic examples of this.

7.2.  Integrity Protection



   Description:  Integrity protection in the scope of DetNet is the
      ability to detect if a packet header has been modified
      (maliciously or otherwise) and if so, take some appropriate action
      (as discussed in Section 7.7).  The decision on where in the
      network to apply integrity protection is part of the DetNet system
      design, and the implementation of the protection method itself is
      a part of a DetNet component design.

      The most common technique for detecting header modification is the
      use of a Message Authentication Code (MAC) (see Section 10 for
      examples).  The MAC can be distributed either in line (included in
      the same packet) or via a side channel.  Of these, the in-line
      method is generally preferred due to the low latency that may be
      required on DetNet flows and the relative complexity and
      computational overhead of a sideband approach.

      There are different levels of security available for integrity
      protection, ranging from the basic ability to detect if a header
      has been corrupted in transit (no malicious attack) to stopping a
      skilled and determined attacker capable of both subtly modifying
      fields in the headers as well as updating an unkeyed checksum.
      Common for all are the 2 steps that need to be performed in both
      ends.  The first is computing the checksum or MAC.  The
      corresponding verification step must perform the same steps before
      comparing the provided with the computed value.  Only then can the
      receiver be reasonably sure that the header is authentic.

      The most basic protection mechanism consists of computing a simple
      checksum of the header fields and providing it to the next entity
      in the packets path for verification.  Using a MAC combined with a
      secret key provides the best protection against Modification and
      Replication attacks (see Sections 5.2.2 and 5.2.4).  This MAC
      usage needs to be part of a security association that is
      established and managed by a security association protocol (such
      as IKEv2 for IPsec security associations).  Integrity protection
      in the controller plane is discussed in Section 7.6.  The secret
      key, regardless of the MAC used, must be protected from falling
      into the hands of unauthorized users.  Once key management becomes
      a topic, it is important to understand that this is a delicate
      process and should not be undertaken lightly.  BCP 107 [BCP107]
      provides best practices in this regard.

      DetNet system and/or component designers need to be aware of these
      distinctions and enforce appropriate integrity-protection
      mechanisms as needed based on a threat analysis.  Note that adding
      integrity-protection mechanisms may introduce latency; thus, many
      of the same considerations in Section 7.5.1 also apply here.

   Packet Sequence Number Integrity Considerations:  The use of PREOF in
      a DetNet implementation implies the use of a sequence number for
      each packet.  There is a trust relationship between the component
      that adds the sequence number and the component that removes the
      sequence number.  The sequence number may be end-to-end source to
      destination, or it may be added/deleted by network edge
      components.  The adder and remover(s) have the trust relationship
      because they are the ones that ensure that the sequence numbers
      are not modifiable.  Thus, sequence numbers can be protected by
      using authenticated encryption or by a MAC without using
      encryption.  Between the adder and remover there may or may not be
      replication and elimination functions.  The elimination functions
      must be able to see the sequence numbers.  Therefore, if
      encryption is done between adders and removers, it must not
      obscure the sequence number.  If the sequence removers and the
      eliminators are in the same physical component, it may be possible
      to obscure the sequence number; however, that is a layer violation
      and is not recommended practice.

         |  Note: At the time of this writing, PREOF is not defined for
         |  the IP data plane.

   Related attacks:  Integrity protection mitigates attacks related to
      modification and tampering, including the attacks described in
      Sections 5.2.2 and 5.2.4.

7.3.  DetNet Node Authentication



   Description:  Authentication verifies the identity of DetNet nodes
      (including DetNet Controller Plane nodes), and this enables
      mitigation of Spoofing attacks.  While integrity protection
      (Section 7.2) prevents intermediate nodes from modifying
      information, authentication can provide traffic origin
      verification, i.e., to verify that each packet in a DetNet flow is
      from a known source.  Although node authentication and integrity
      protection are two different goals of a security protocol, in most
      cases, a common protocol (such as IPsec [RFC4301] or MACsec
      [IEEE802.1AE-2018]) is used for achieving both purposes.

   Related attacks:  DetNet node authentication is used to mitigate
      attacks related to spoofing, including the attacks of Sections
      5.2.2 and 5.2.4.

7.4.  Synthetic Traffic Insertion



   Description:  With some queuing methods such as [IEEE802.1Qch-2017],
      it is possible to introduce synthetic traffic in order to
      regularize the timing of packet transmission.  (Synthetic traffic
      typically consists of randomly generated packets injected in the
      network to mask observable transmission patterns in the flows,
      which may allow an attacker to gain insight into the content of
      the flows).  This can subsequently reduce the value of passive
      monitoring from internal threats (see Section 5) as it will be
      much more difficult to associate discrete events with particular
      network packets.

   Related attacks:  Removing distinctive temporal properties of
      individual packets or flows can be used to mitigate against
      reconnaissance attacks (Section 5.2.6).  For example, synthetic
      traffic can be used to maintain constant traffic rate even when no
      user data is transmitted, thus making it difficult to collect
      information about the times at which users are active and the
      times at which DetNet flows are added or removed.

   Traffic Insertion Challenges:  Once an attacker is able to monitor
      the frames traversing a network to such a degree that they can
      differentiate between best-effort traffic and traffic belonging to
      a specific DetNet flow, it becomes difficult to not reveal to the
      attacker whether a given frame is valid traffic or an inserted
      frame.  Thus, having the DetNet components generate and remove the
      synthetic traffic may or may not be a viable option unless certain
      challenges are solved; for example, but not limited to:

      *  Inserted traffic must be indistinguishable from valid stream
         traffic from the viewpoint of the attacker.

      *  DetNet components must be able to safely identify and remove
         all inserted traffic (and only inserted traffic).

      *  The controller plane must manage where to insert and remove
         synthetic traffic, but this information must not be revealed to
         an attacker.

         An alternative design is to have the insertion and removal of
         synthetic traffic be performed at the application layer rather
         than by the DetNet itself.  For example, the use of RTP padding
         to reduce information leakage from variable-bit-rate audio
         transmission via the Secure Real-time Transport Protocol (SRTP)
         is discussed in [RFC6562].

7.5.  Encryption



   Description:  Reconnaissance attacks (Section 5.2.6) can be mitigated
      to some extent through the use of encryption, thereby preventing
      the attacker from accessing the packet header or contents.
      Specific encryption protocols will depend on the lower layers that
      DetNet is forwarded over.  For example, IP flows may be forwarded
      over IPsec [RFC4301], and Ethernet flows may be secured using
      MACsec [IEEE802.1AE-2018].

      However, despite the use of encryption, a reconnaissance attack
      can provide the attacker with insight into the network, even
      without visibility into the packet.  For example, an attacker can
      observe which nodes are communicating with which other nodes,
      including when, how often, and with how much data.  In addition,
      the timing of packets may be correlated in time with external
      events such as action of an external device.  Such information may
      be used by the attacker, for example, in mapping out specific
      targets for a different type of attack at a different time.

      DetNet nodes do not have any need to inspect the payload of any
      DetNet packets, making them data agnostic.  This means that end-
      to-end encryption at the application layer is an acceptable way to
      protect user data.

      Note that reconnaissance is a threat that is not specific to
      DetNet flows; therefore, reconnaissance mitigation will typically
      be analyzed and provided by a network operator regardless of
      whether DetNet flows are deployed.  Thus, encryption requirements
      will typically not be defined in DetNet technology-specific
      specifications, but considerations of using DetNet in encrypted
      environments will be discussed in these specifications.  For
      example, Section 5.1.2.3 of [RFC8939] discusses flow
      identification of DetNet flows running over IPsec.

   Related attacks:  As noted above, encryption can be used to mitigate
      reconnaissance attacks (Section 5.2.6).  However, for a DetNet to
      provide differentiated quality of service on a flow-by-flow basis,
      the network must be able to identify the flows individually.  This
      implies that in a reconnaissance attack, the attacker may also be
      able to track individual flows to learn more about the system.

7.5.1.  Encryption Considerations for DetNet



   Any compute time that is required for encryption and decryption
   processing ("crypto") must be included in the flow latency
   calculations.  Thus, cryptographic algorithms used in a DetNet must
   have bounded worst-case execution times, and these values must be
   used in the latency calculations.  Fortunately, encryption and
   decryption operations typically are designed to have constant
   execution times in order to avoid side channel leakage.

   Some cryptographic algorithms are symmetric in encode/decode time
   (such as AES), and others are asymmetric (such as public key
   algorithms).  There are advantages and disadvantages to the use of
   either type in a given DetNet context.  The discussion in this
   document relates to the timing implications of crypto for DetNet; it
   is assumed that integrity considerations are covered elsewhere in the
   literature.

   Asymmetrical crypto is typically not used in networks on a packet-by-
   packet basis due to its computational cost.  For example, if only
   endpoint checks or checks at a small number of intermediate points
   are required, asymmetric crypto can be used to authenticate
   distribution or exchange of a secret symmetric crypto key; a
   successful check based on that key will provide traffic origin
   verification as long as the key is kept secret by the participants.
   TLS (v1.3 [RFC8446], in particular, Section 4.1 ("Key Exchange
   Messages")) and IKEv2 [RFC6071] are examples of this for endpoint
   checks.

   However, if secret symmetric keys are used for this purpose, the key
   must be given to all relays, which increases the probability of a
   secret key being leaked.  Also, if any relay is compromised or
   faulty, then it may inject traffic into the flow.  Group key
   management protocols can be used to automate management of such
   symmetric keys; for an example in the context of IPsec, see
   [IPSECME-G-IKEV2].

   Alternatively, asymmetric crypto can provide traffic origin
   verification at every intermediate node.  For example, a DetNet flow
   can be associated with an (asymmetric) keypair, such that the private
   key is available to the source of the flow and the public key is
   distributed with the flow information, allowing verification at every
   node for every packet.  However, this is more computationally
   expensive.

   In either case, origin verification also requires replay detection as
   part of the security protocol to prevent an attacker from recording
   and resending traffic, e.g., as a denial-of-service attack on flow
   forwarding resources.

   In the general case, cryptographic hygiene requires the generation of
   new keys during the lifetime of an encrypted flow (e.g., see
   Section 9 of [RFC4253]), and any such key generation (or key
   exchange) requires additional computing time, which must be accounted
   for in the latency calculations for that flow.  For modern ECDH
   (Elliptical Curve Diffie-Hellman) key-exchange operations (such as
   x25519 [RFC7748]), these operations can be performed in constant
   (predictable) time; however, this is not universally true (for
   example, for legacy RSA key exchange [RFC4432]).  Thus, implementers
   should be aware of the time properties of these algorithms and avoid
   algorithms that make constant-time implementation difficult or
   impossible.

7.6.  Control and Signaling Message Protection



   Description:  Control and signaling messages can be protected through
      the use of any or all of encryption, authentication, and
      integrity-protection mechanisms.  Compared with data flows, the
      timing constraints for controller and signaling messages may be
      less strict, and the number of such packets may be fewer.  If that
      is the case in a given application, then it may enable the use of
      asymmetric cryptography for the signing of both payload and
      headers for such messages, as well as encrypting the payload.
      Given that a DetNet is managed by a central controller, the use of
      a shared public key approach for these processes is well proven.
      This is further discussed in Section 7.5.1.

   Related attacks:  These mechanisms can be used to mitigate various
      attacks on the controller plane, as described in Sections 5.2.5,
      5.2.7, and 5.2.5.1.

7.7.  Dynamic Performance Analytics



   Description:  Incorporating Dynamic Performance Analytics (DPA)
      implies that the DetNet design includes a performance monitoring
      system to validate that timing guarantees are being met and to
      detect timing violations or other anomalies that may be the
      symptom of a security attack or system malfunction.  If this
      monitoring system detects unexpected behavior, it must then cause
      action to be initiated to address the situation in an appropriate
      and timely manner, either at the data plane or controller plane or
      both in concert.

      The overall DPA system can thus be decomposed into the "detection"
      and "notification" functions.  Although the time-specific DPA
      performance indicators and their implementation will likely be
      specific to a given DetNet, and as such are nascent technology at
      the time of this writing, DPA is commonly used in existing
      networks so we can make some observations on how such a system
      might be implemented for a DetNet given that it would need to be
      adapted to address the time-specific performance indicators.

   Detection Mechanisms:  Measurement of timing performance can be done
      via "passive" or "active" monitoring, as discussed below.

      Examples of passive monitoring strategies include:

      *  Monitoring of queue and buffer levels, e.g., via active queue
         management (e.g., [RFC7567]).

      *  Monitoring of per-flow counters.

      *  Measurement of link statistics such as traffic volume,
         bandwidth, and QoS.

      *  Detection of dropped packets.

      *  Use of commercially available Network Monitoring tools.

      Examples of active monitoring include:

      *  In-band timing measurements (such as packet arrival times),
         e.g., by timestamping and packet inspection.

      *  Use of OAM.  For DetNet-specific OAM considerations, see
         [DETNET-IP-OAM] and [DETNET-MPLS-OAM].  Note: At the time of
         this writing, specifics of DPA have not been developed for the
         DetNet OAM but could be a subject for future investigation.

         -  For OAM for Ethernet specifically, see also Connectivity
            Fault Management (CFM [IEEE802.1Q]), which defines protocols
            and practices for OAM for paths through 802.1 bridges and
            LANs.

      *  Out-of-band detection.  Following the data path or parts of a
         data path, for example, Bidirectional Forwarding Detection
         (BFD, e.g., [RFC5880]).

      Note that for some measurements (e.g., packet delay), it may be
      necessary to make and reconcile measurements from more than one
      physical location (e.g., a source and destination), possibly in
      both directions, in order to arrive at a given performance
      indicator value.

   Notification Mechanisms:  Making DPA measurement results available at
      the right place(s) and time(s) to effect timely response can be
      challenging.  Two notification mechanisms that are in general use
      are NETCONF/YANG Notifications and the proprietary local telemetry
      interfaces provided with components from some vendors.  The
      Constrained Application Protocol (CoAP) Observe Option [RFC7641]
      could also be relevant to such scenarios.

      At the time of this writing, YANG Notifications are not addressed
      by the DetNet YANG documents; however, this may be a topic for
      future work.  It is possible that some of the passive mechanisms
      could be covered by notifications from non-DetNet-specific YANG
      modules; for example, if there is OAM or other performance
      monitoring that can monitor delay bounds, then that could have its
      own associated YANG data model, which could be relevant to DetNet,
      for example, some "threshold" values for timing measurement
      notifications.

      At the time of this writing, there is an IETF Working Group for
      network/performance monitoring (IP Performance Metrics (IPPM)).
      See also previous work by the completed Remote Network Monitoring
      Working Group (RMONMIB).  See also "An Overview of the IETF
      Network Management Standards", [RFC6632].

      Vendor-specific local telemetry may be available on some
      commercially available systems, whereby the system can be
      programmed (via a proprietary dedicated port and API) to monitor
      and report on specific conditions, based on both passive and
      active measurements.

   Related attacks:  Performance analytics can be used to detect various
      attacks, including the ones described in Section 5.2.1 (Delay
      attack), Section 5.2.3 (Resource Segmentation attack), and
      Section 5.2.7 (Time-Synchronization attack).  Once detection and
      notification have occurred, the appropriate action can be taken to
      mitigate the threat.

      For example, in the case of data plane Delay attacks, one possible
      mitigation is to timestamp the data at the source and timestamp it
      again at the destination, and if the resulting latency does not
      meet the service agreement, take appropriate action.  Note that
      DetNet specifies packet sequence numbering; however, it does not
      specify use of packet timestamps, although they may be used by the
      underlying transport (for example, TSN [IEEE802.1BA]) to provide
      the service.

7.8.  Mitigation Summary



   The following table maps the attacks of Section 5 (Security Threats)
   to the impacts of Section 6 (Security Threat Impacts) and to the
   mitigations of the current section.  Each row specifies an attack,
   the impact of this attack if it is successfully implemented, and
   possible mitigation methods.

   +======================+======================+=====================+
   | Attack               | Impact               | Mitigations         |
   +======================+======================+=====================+
   | Delay Attack         | *  Non-deterministic | *  Path redundancy  |
   |                      |    delay             |                     |
   |                      |                      | *  Performance      |
   |                      | *  Data disruption   |    analytics        |
   |                      |                      |                     |
   |                      | *  Increased         |                     |
   |                      |    resource          |                     |
   |                      |    consumption       |                     |
   +----------------------+----------------------+---------------------+
   | Reconnaissance       | *  Enabler for other | *  Encryption       |
   |                      |    attacks           |                     |
   |                      |                      | *  Synthetic        |
   |                      |                      |    traffic          |
   |                      |                      |    insertion        |
   +----------------------+----------------------+---------------------+
   | DetNet Flow          | *  Increased         | *  Path redundancy  |
   | Modification or      |    resource          |                     |
   | Spoofing             |    consumption       | *  Integrity        |
   |                      |                      |    protection       |
   |                      | *  Data disruption   |                     |
   |                      |                      | *  DetNet Node      |
   |                      |                      |    authentication   |
   +----------------------+----------------------+---------------------+
   | Inter-segment Attack | *  Increased         | *  Path redundancy  |
   |                      |    resource          |                     |
   |                      |    consumption       | *  Performance      |
   |                      |                      |    analytics        |
   |                      | *  Data disruption   |                     |
   +----------------------+----------------------+---------------------+
   | Replication:         | *  All impacts of    | *  Integrity        |
   | Increased Attack     |    other attacks     |    protection       |
   | Resource             |                      |                     |
   |                      |                      | *  DetNet Node      |
   |                      |                      |    authentication   |
   |                      |                      |                     |
   |                      |                      | *  Encryption       |
   +----------------------+----------------------+---------------------+
   | Replication-Related  | *  Non-deterministic | *  Integrity        |
   | Header Manipulation  |    delay             |    protection       |
   |                      |                      |                     |
   |                      | *  Data disruption   | *  DetNet Node      |
   |                      |                      |    authentication   |
   +----------------------+----------------------+---------------------+
   | Path Manipulation    | *  Enabler for other | *  Control and      |
   |                      |    attacks           |    signaling        |
   |                      |                      |    message          |
   |                      |                      |    protection       |
   +----------------------+----------------------+---------------------+
   | Path Choice:         | *  All impacts of    | *  Control and      |
   | Increased Attack     |    other attacks     |    signaling        |
   | Surface              |                      |    message          |
   |                      |                      |    protection       |
   +----------------------+----------------------+---------------------+
   | Control or Signaling | *  Increased         | *  Control and      |
   | Packet Modification  |    resource          |    signaling        |
   |                      |    consumption       |    message          |
   |                      |                      |    protection       |
   |                      | *  Non-deterministic |                     |
   |                      |    delay             |                     |
   |                      |                      |                     |
   |                      | *  Data disruption   |                     |
   +----------------------+----------------------+---------------------+
   | Control or Signaling | *  Increased         | *  Control and      |
   | Packet Injection     |    resource          |    signaling        |
   |                      |    consumption       |    message          |
   |                      |                      |    protection       |
   |                      | *  Non-deterministic |                     |
   |                      |    delay             |                     |
   |                      |                      |                     |
   |                      | *  Data disruption   |                     |
   +----------------------+----------------------+---------------------+
   | Attacks on Time-     | *  Non-deterministic | *  Path redundancy  |
   | Synchronization      |    delay             |                     |
   | Mechanisms           |                      | *  Control and      |
   |                      | *  Increased         |    signaling        |
   |                      |    resource          |    message          |
   |                      |    consumption       |    protection       |
   |                      |                      |                     |
   |                      | *  Data disruption   | *  Performance      |
   |                      |                      |    analytics        |
   +----------------------+----------------------+---------------------+

             Table 3: Mapping Attacks to Impact and Mitigations

8.  Association of Attacks to Use Cases



   Different attacks can have different impact and/or mitigation
   depending on the use case, so we would like to make this association
   in our analysis.  However, since there is a potentially unbounded
   list of use cases, we categorize the attacks with respect to the
   common themes of the use cases as identified in Section 11 of
   [RFC8578].

   See also Table 2 for a mapping of the impact of attacks per use case
   by industry.

8.1.  Association of Attacks to Use Case Common Themes



   In this section, we review each theme and discuss the attacks that
   are applicable to that theme, as well as anything specific about the
   impact and mitigations for that attack with respect to that theme.
   Table 5, Mapping between Themes and Attacks, then provides a summary
   of the attacks that are applicable to each theme.

8.1.1.  Sub-network Layer



   DetNet is expected to run over various transmission mediums, with
   Ethernet being the first identified.  Attacks such as Delay or
   Reconnaissance might be implemented differently on a different
   transmission medium; however, the impact on the DetNet as a whole
   would be essentially the same.  We thus conclude that all attacks and
   impacts that would be applicable to DetNet over Ethernet (i.e., all
   those named in this document) would also be applicable to DetNet over
   other transmission mediums.

   With respect to mitigations, some methods are specific to the
   Ethernet medium, for example, time-aware scheduling using 802.1Qbv
   [IEEE802.1Qbv-2015] can protect against excessive use of bandwidth at
   the ingress -- for other mediums, other mitigations would have to be
   implemented to provide analogous protection.

8.1.2.  Central Administration



   A DetNet network can be controlled by a centralized network
   configuration and control system.  Such a system may be in a single
   central location, or it may be distributed across multiple control
   entities that function together as a unified control system for the
   network.

   All attacks named in this document that are relevant to controller
   plane packets (and the controller itself) are relevant to this theme,
   including Path Manipulation, Path Choice, Control Packet Modification
   or Injection, Reconnaissance, and Attacks on Time-Synchronization
   Mechanisms.

8.1.3.  Hot Swap



   A DetNet network is not expected to be "plug and play"; it is
   expected that there is some centralized network configuration and
   control system.  However, the ability to "hot swap" components (e.g.,
   due to malfunction) is similar enough to "plug and play" that this
   kind of behavior may be expected in DetNet networks, depending on the
   implementation.

   An attack surface related to hot swap is that the DetNet network must
   at least consider input at runtime from components that were not part
   of the initial configuration of the network.  Even a "perfect" (or
   "hitless") replacement of a component at runtime would not
   necessarily be ideal, since presumably one would want to distinguish
   it from the original for OAM purposes (e.g., to report hot swap of a
   failed component).

   This implies that an attack such as Flow Modification, Spoofing, or
   Inter-segment (which could introduce packets from a "new" component,
   i.e., one heretofore unknown on the network) could be used to exploit
   the need to consider such packets (as opposed to rejecting them out
   of hand as one would do if one did not have to consider introduction
   of a new component).

   To mitigate this situation, deployments should provide a method for
   dynamic and secure registration of new components, and (possibly
   manual) deregistration and re-keying of retired components.  This
   would avoid the situation in which the network must accommodate
   potentially insecure packet flows from unknown components.

   Similarly, if the network was designed to support runtime replacement
   of a clock component, then presence (or apparent presence) and thus
   consideration of packets from a new such component could affect the
   network, or the time synchronization of the network, for example, by
   initiating a new Best Master Clock selection process.  These types of
   attacks should therefore be considered when designing hot-swap-type
   functionality (see [RFC7384]).

8.1.4.  Data Flow Information Models



   DetNet specifies new YANG data models [DETNET-YANG] that may present
   new attack surfaces.  Per IETF guidelines, security considerations
   for any YANG data model are expected to be part of the YANG data
   model specification, as described in [IETF-YANG-SEC].

8.1.5.  L2 and L3 Integration



   A DetNet network integrates Layer 2 (bridged) networks (e.g., AVB/TSN
   LAN) and Layer 3 (routed) networks (e.g., IP) via the use of well-
   known protocols such as IP, MPLS Pseudowire, and Ethernet.  Various
   DetNet documents address many specific aspects of Layer 2 and Layer 3
   integration within a DetNet, and these are not individually
   referenced here; security considerations for those aspects are
   covered within those documents or within the related subsections of
   the present document.

   Please note that although there are no entries in the L2 and L3
   Integration line of the Mapping between Themes and Attacks table
   (Table 5), this does not imply that there could be no relevant
   attacks related to L2-L3 integration.

8.1.6.  End-to-End Delivery



   Packets that are part of a resource-reserved DetNet flow are not to
   be dropped by the DetNet due to congestion.  Packets may however be
   dropped for intended reasons, for example, security measures.  For
   example, consider the case in which a packet becomes corrupted
   (whether incidentally or maliciously) such that the resulting flow ID
   incidentally matches the flow ID of another DetNet flow, potentially
   resulting in additional unauthorized traffic on the latter.  In such
   a case, it may be a security requirement that the system report and/
   or take some defined action, perhaps when a packet drop count
   threshold has been reached (see also Section 7.7).

   A data plane attack may force packets to be dropped, for example, as
   a result of a Delay attack, Replication/Elimination attack, or Flow
   Modification attack.

   The same result might be obtained by a Controller plane attack, e.g.,
   Path Manipulation or Signaling Packet Modification.

   An attack may also cause packets that should not be delivered to be
   delivered, such as by forcing packets from one (e.g., replicated)
   path to be preferred over another path when they should not be
   (Replication attack), or by Flow Modification, or Path Choice or
   Packet Injection.  A Time-Synchronization attack could cause a system
   that was expecting certain packets at certain times to accept
   unintended packets based on compromised system time or time windowing
   in the scheduler.

8.1.7.  Replacement for Proprietary Fieldbuses and Ethernet-Based
        Networks



   There are many proprietary "fieldbuses" used in Industrial and other
   industries, as well as proprietary non-interoperable deterministic
   Ethernet-based networks.  DetNet is intended to provide an open-
   standards-based alternative to such buses/networks.  In cases where a
   DetNet intersects with such fieldbuses/networks or their protocols,
   such as by protocol emulation or access via a gateway, new attack
   surfaces can be opened.

   For example, an Inter-segment or Controller plane attack such as Path
   Manipulation, Path Choice, or Control Packet Modification/Injection
   could be used to exploit commands specific to such a protocol or that
   are interpreted differently by the different protocols or gateway.

8.1.8.  Deterministic vs. Best-Effort Traffic



   Most of the themes described in this document address OT (reserved)
   DetNet flows -- this item is intended to address issues related to IT
   traffic on a DetNet.

   DetNet is intended to support coexistence of time-sensitive
   operational (OT, deterministic) traffic and informational (IT, "best
   effort") traffic on the same ("unified") network.

   With DetNet, this coexistence will become more common, and
   mitigations will need to be established.  The fact that the IT
   traffic on a DetNet is limited to a corporate-controlled network
   makes this a less difficult problem compared to being exposed to the
   open Internet; however, this aspect of DetNet security should not be
   underestimated.

   An Inter-segment attack can flood the network with IT-type traffic
   with the intent of disrupting the handling of IT traffic and/or the
   goal of interfering with OT traffic.  Presumably, if the DetNet flow
   reservation and isolation of the DetNet is well designed (better-
   designed than the attack), then interference with OT traffic should
   not result from an attack that floods the network with IT traffic.

   The handling of IT traffic (i.e., traffic that by definition is not
   guaranteed any given deterministic service properties) by the DetNet
   will by definition not be given the DetNet-specific protections
   provided to DetNet (resource-reserved) flows.  The implication is
   that the IT traffic on the DetNet network will necessarily have its
   own specific set of product (component or system) requirements for
   protection against attacks such as DoS; presumably they will be less
   stringent than those for OT flows, but nonetheless, component and
   system designers must employ whatever mitigations will meet the
   specified security requirements for IT traffic for the given
   component or DetNet.

   The network design as a whole also needs to consider possible
   application-level dependencies of OT-type applications on services
   provided by the IT part of the network; for example, does the OT
   application depend on IT network services such as DNS or OAM?  If
   such dependencies exist, how are malicious packet flows handled?
   Such considerations are typically outside the scope of DetNet proper,
   but nonetheless need to be addressed in the overall DetNet network
   design for a given use case.

8.1.9.  Deterministic Flows



   Reserved bandwidth data flows (deterministic flows) must provide the
   allocated bandwidth and must be isolated from each other.

   A Spoofing or Inter-segment attack that adds packet traffic to a
   bandwidth-reserved DetNet flow could cause that flow to occupy more
   bandwidth than it was allocated, resulting in interference with other
   DetNet flows.

   A Flow Modification, Spoofing, Header Manipulation, or Control Packet
   Modification attack could cause packets from one flow to be directed
   to another flow, thus breaching isolation between the flows.

8.1.10.  Unused Reserved Bandwidth



   If bandwidth reservations are made for a DetNet flow but the
   associated bandwidth is not used at any point in time, that bandwidth
   is made available on the network for best-effort traffic.  However,
   note that security considerations for best-effort traffic on a DetNet
   network is out of scope of the present document, provided that any
   such attacks on best-effort traffic do not affect performance for
   DetNet OT traffic.

8.1.11.  Interoperability



   The DetNet specifications as a whole are intended to enable an
   ecosystem in which multiple vendors can create interoperable
   products, thus promoting component diversity and potentially higher
   numbers of each component manufactured.  Toward that end, the
   security measures and protocols discussed in this document are
   intended to encourage interoperability.

   Given that the DetNet specifications are unambiguously written and
   that the implementations are accurate, the property of
   interoperability should not in and of itself cause security concerns;
   however, flaws in interoperability between components could result in
   security weaknesses.  The network operator, as well as system and
   component designers, can all contribute to reducing such weaknesses
   through interoperability testing.

8.1.12.  Cost Reductions



   The DetNet network specifications are intended to enable an ecosystem
   in which multiple vendors can create interoperable products, thus
   promoting higher numbers of each component manufactured, promoting
   cost reduction and cost competition among vendors.

   This envisioned breadth of DetNet-enabled products is in general a
   positive factor; however, implementation flaws in any individual
   component can present an attack surface.  In addition, implementation
   differences between components from different vendors can result in
   attack surfaces (resulting from their interaction) that may not exist
   in any individual component.

   Network operators can mitigate such concerns through sufficient
   product and interoperability testing.

8.1.13.  Insufficiently Secure Components



   The DetNet network specifications are intended to enable an ecosystem
   in which multiple vendors can create interoperable products, thus
   promoting component diversity and potentially higher numbers of each
   component manufactured.  However, this raises the possibility that a
   vendor might repurpose for DetNet applications a hardware or software
   component that was originally designed for operation in an isolated
   OT network and thus may not have been designed to be sufficiently
   secure, or secure at all, against the sorts of attacks described in
   this document.  Deployment of such a component on a DetNet network
   that is intended to be highly secure may present an attack surface;
   thus, the DetNet network operator may need to take specific actions
   to protect such components, for example, by implementing a secure
   interface (such as a firewall) to isolate the component from the
   threats that may be present in the greater network.

8.1.14.  DetNet Network Size



   DetNet networks range in size from very small, e.g., inside a single
   industrial machine, to very large, e.g., a Utility Grid network
   spanning a whole country.

   The size of the network might be related to how the attack is
   introduced into the network.  For example, if the entire network is
   local, there is a threat that power can be cut to the entire network.
   If the network is large, perhaps only a part of the network is
   attacked.

   A Delay attack might be as relevant to a small network as to a large
   network, although the amount of delay might be different.

   Attacks sourced from IT traffic might be more likely in large
   networks since more people might have access to the network,
   presenting a larger attack surface.  Similarly, Path Manipulation,
   Path Choice, and Time-Synchronization attacks seem more likely
   relevant to large networks.

8.1.15.  Multiple Hops



   Large DetNet networks (e.g., a Utility Grid network) may involve many
   "hops" over various kinds of links, for example, radio repeaters,
   microwave links, fiber optic links, etc.

   An attacker who has knowledge of the operation of a component or
   device's internal software (such as "device drivers") may be able to
   take advantage of this knowledge to design an attack that could
   exploit flaws (or even the specifics of normal operation) in the
   communication between the various links.

   It is also possible that a large-scale DetNet topology containing
   various kinds of links may not be in as common use as other more
   homogeneous topologies.  This situation may present more opportunity
   for attackers to exploit software and/or protocol flaws in or between
   these components because these components or configurations may not
   have been sufficiently tested for interoperability (in the way they
   would be as a result of broad usage).  This may be of particular
   concern to early adopters of new DetNet components or technologies.

   Of the attacks we have defined, the ones identified in Section 8.1.14
   as germane to large networks are the most relevant.

8.1.16.  Level of Service



   A DetNet is expected to provide means to configure the network that
   include querying network path latency, requesting bounded latency for
   a given DetNet flow, requesting worst-case maximum and/or minimum
   latency for a given path or DetNet flow, and so on.  It is an
   expected case that the network cannot provide a given requested
   service level.  In such cases, the network control system should
   reply that the requested service level is not available (as opposed
   to accepting the parameter but then not delivering the desired
   behavior).

   Controller plane attacks such as Signaling Packet Modification and
   Injection could be used to modify or create control traffic that
   could interfere with the process of a user requesting a level of
   service and/or the reply from the network.

   Reconnaissance could be used to characterize flows and perhaps target
   specific flows for attack via the controller plane as noted in
   Section 6.7.

8.1.17.  Bounded Latency



   DetNet provides the expectation of guaranteed bounded latency.

   Delay attacks can cause packets to miss their agreed-upon latency
   boundaries.

   Time-Synchronization attacks can corrupt the time reference of the
   system, resulting in missed latency deadlines (with respect to the
   "correct" time reference).

8.1.18.  Low Latency



   Applications may require "extremely low latency"; however, depending
   on the application, these may mean very different latency values.
   For example, "low latency" across a Utility Grid network is on a
   different time scale than "low latency" in a motor control loop in a
   small machine.  The intent is that the mechanisms for specifying
   desired latency include wide ranges, and that architecturally there
   is nothing to prevent arbitrarily low latencies from being
   implemented in a given network.

   Attacks on the controller plane (as described in the Level of Service
   theme; see Section 8.1.16) and Delay and Time attacks (as described
   in the Bounded Latency theme; see Section 8.1.17) both apply here.

8.1.19.  Bounded Jitter (Latency Variation)



   DetNet is expected to provide bounded jitter (packet-to-packet
   latency variation).

   Delay attacks can cause packets to vary in their arrival times,
   resulting in packet-to-packet latency variation, thereby violating
   the jitter specification.

8.1.20.  Symmetrical Path Delays



   Some applications would like to specify that the transit delay time
   values be equal for both the transmit and return paths.

   Delay attacks can cause path delays to materially differ between
   paths.

   Time-Synchronization attacks can corrupt the time reference of the
   system, resulting in path delays that may be perceived to be
   different (with respect to the "correct" time reference) even if they
   are not materially different.

8.1.21.  Reliability and Availability



   DetNet-based systems are expected to be implemented with essentially
   arbitrarily high availability (for example, 99.9999% up time, or even
   12 nines).  The intent is that the DetNet designs should not make any
   assumptions about the level of reliability and availability that may
   be required of a given system and should define parameters for
   communicating these kinds of metrics within the network.

   Any attack on the system, of any type, can affect its overall
   reliability and availability; thus, in the mapping table (Table 5),
   we have marked every attack.  Since every DetNet depends to a greater
   or lesser degree on reliability and availability, this essentially
   means that all networks have to mitigate all attacks, which to a
   greater or lesser degree defeats the purpose of associating attacks
   with use cases.  It also underscores the difficulty of designing
   "extremely high reliability" networks.

   In practice, network designers can adopt a risk-based approach in
   which only those attacks are mitigated whose potential cost is higher
   than the cost of mitigation.

8.1.22.  Redundant Paths



   This document expects that each DetNet system will be implemented to
   some essentially arbitrary level of reliability and/or availability,
   depending on the use case.  A strategy used by DetNet for providing
   extraordinarily high levels of reliability when justified is to
   provide redundant paths between which traffic can be seamlessly
   switched, all the while maintaining the required performance of that
   system.

   Replication-related attacks are by definition applicable here.
   Controller plane attacks can also interfere with the configuration of
   redundant paths.

8.1.23.  Security Measures



   If any of the security mechanisms that protect the DetNet are
   attacked or subverted, this can result in malfunction of the network.
   Thus, the security systems themselves need to be robust against
   attacks.

   The general topic of protection of security mechanisms is not unique
   to DetNet; it is identical to the case of securing any security
   mechanism for any network.  This document addresses these concerns
   only to the extent that they are unique to DetNet.

8.2.  Summary of Attack Types per Use Case Common Theme



   The List of Attacks table (Table 4) lists the attacks described in
   Section 5, Security Threats, assigning a number to each type of
   attack.  That number is then used as a short form identifier for the
   attack in Table 5, Mapping between Themes and Attacks.

            +====+============================================+
            |    | Attack                                     |
            +====+============================================+
            | 1  | Delay Attack                               |
            +----+--------------------------------------------+
            | 2  | DetNet Flow Modification or Spoofing       |
            +----+--------------------------------------------+
            | 3  | Inter-segment Attack                       |
            +----+--------------------------------------------+
            | 4  | Replication: Increased Attack Surface      |
            +----+--------------------------------------------+
            | 5  | Replication-Related Header Manipulation    |
            +----+--------------------------------------------+
            | 6  | Path Manipulation                          |
            +----+--------------------------------------------+
            | 7  | Path Choice: Increased Attack Surface      |
            +----+--------------------------------------------+
            | 8  | Control or Signaling Packet Modification   |
            +----+--------------------------------------------+
            | 9  | Control or Signaling Packet Injection      |
            +----+--------------------------------------------+
            | 10 | Reconnaissance                             |
            +----+--------------------------------------------+
            | 11 | Attacks on Time-Synchronization Mechanisms |
            +----+--------------------------------------------+

                          Table 4: List of Attacks

   The Mapping between Themes and Attacks table (Table 5) maps the use
   case themes of [RFC8578] (as also enumerated in this document) to the
   attacks of Table 4.  Each row specifies a theme, and the attacks
   relevant to this theme are marked with a "+".  The row items that
   have no threats associated with them are included in the table for
   completeness of the list of Use Case Common Themes and do not have
   DetNet-specific threats associated with them.

   +====================+=============================================+
   |       Theme        |                    Attack                   |
   |                    +===+===+===+===+===+===+===+===+===+====+====+
   |                    | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 |
   +====================+===+===+===+===+===+===+===+===+===+====+====+
   | Network Layer -    | + | + | + | + | + | + | + | + | + | +  | +  |
   | AVB/TSN Eth.       |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Central            |   |   |   |   |   | + | + | + | + | +  | +  |
   | Administration     |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Hot Swap           |   | + | + |   |   |   |   |   |   |    | +  |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Data Flow          |   |   |   |   |   |   |   |   |   |    |    |
   | Information Models |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | L2 and L3          |   |   |   |   |   |   |   |   |   |    |    |
   | Integration        |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | End-to-End         | + | + | + | + | + | + | + | + |   | +  |    |
   | Delivery           |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Proprietary        |   |   | + |   |   | + | + | + | + |    |    |
   | Deterministic      |   |   |   |   |   |   |   |   |   |    |    |
   | Ethernet Networks  |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Replacement for    |   |   | + |   |   |   |   |   |   |    |    |
   | Proprietary        |   |   |   |   |   |   |   |   |   |    |    |
   | Fieldbuses         |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Deterministic vs.  | + | + | + |   | + | + |   | + |   |    |    |
   | Best-Effort        |   |   |   |   |   |   |   |   |   |    |    |
   | Traffic            |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Deterministic      | + | + | + |   | + | + |   | + |   |    |    |
   | Flows              |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Unused Reserved    |   | + | + |   |   |   |   | + | + |    |    |
   | Bandwidth          |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Interoperability   |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Cost Reductions    |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Insufficiently     |   |   |   |   |   |   |   |   |   |    |    |
   | Secure Components  |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | DetNet Network     | + |   |   |   |   | + | + |   |   |    | +  |
   | Size               |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Multiple Hops      | + | + |   |   |   | + | + |   |   |    | +  |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Level of Service   |   |   |   |   |   |   |   | + | + | +  |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Bounded Latency    | + |   |   |   |   |   |   |   |   |    | +  |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Low Latency        | + |   |   |   |   |   |   | + | + |    | +  |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Bounded Jitter     | + |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Symmetric Path     | + |   |   |   |   |   |   |   |   |    | +  |
   | Delays             |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Reliability and    | + | + | + | + | + | + | + | + | + | +  | +  |
   | Availability       |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Redundant Paths    |   |   |   | + | + |   |   | + | + |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+
   | Security Measures  |   |   |   |   |   |   |   |   |   |    |    |
   +--------------------+---+---+---+---+---+---+---+---+---+----+----+

               Table 5: Mapping between Themes and Attacks

9.  Security Considerations for OAM Traffic



   This section considers DetNet-specific security considerations for
   packet traffic that is generated and transmitted over a DetNet as
   part of OAM (Operations, Administration, and Maintenance).  For the
   purposes of this discussion, OAM traffic falls into one of two basic
   types:

   *  OAM traffic generated by the network itself.  The additional
      bandwidth required for such packets is added by the network
      administration, presumably transparent to the customer.  Security
      considerations for such traffic are not DetNet specific (apart
      from such traffic being subject to the same DetNet-specific
      security considerations as any other DetNet data flow) and are
      thus not covered in this document.

   *  OAM traffic generated by the customer.  From a DetNet security
      point of view, DetNet security considerations for such traffic are
      exactly the same as for any other customer data flows.

   From the perspective of an attack, OAM traffic is indistinguishable
   from DetNet traffic, and the network needs to be secure against
   injection, removal, or modification of traffic of any kind, including
   OAM traffic.  A DetNet is sensitive to any form of packet injection,
   removal, or manipulation, and in this respect DetNet OAM traffic is
   no different.  Techniques for securing a DetNet against these threats
   have been discussed elsewhere in this document.

10.  DetNet Technology-Specific Threats



   Section 5, Security Threats, describes threats that are independent
   of a DetNet implementation.  This section considers threats
   specifically related to the IP- and MPLS-specific aspects of DetNet
   implementations.

   The primary security considerations for the data plane specifically
   are to maintain the integrity of the data and the delivery of the
   associated DetNet service traversing the DetNet network.

   The primary relevant differences between IP and MPLS implementations
   are in flow identification and OAM methodologies.

   As noted in [RFC8655], DetNet operates at the IP layer [RFC8939] and
   delivers service over sub-layer technologies such as MPLS [RFC8964]
   and IEEE 802.1 Time-Sensitive Networking (TSN) [RFC9023].
   Application flows can be protected through whatever means are
   provided by the layer and sub-layer technologies.  For example,
   technology-specific encryption may be used for IP flows (IPsec
   [RFC4301]).  For IP-over-Ethernet (Layer 2) flows using an underlying
   sub-net, MACsec [IEEE802.1AE-2018] may be appropriate.  For some use
   cases, packet integrity protection without encryption may be
   sufficient.

   However, if the DetNet nodes cannot decrypt IPsec traffic, then
   DetNet flow identification for encrypted IP traffic flows must be
   performed in a different way than it would be for unencrypted IP
   DetNet flows.  The DetNet IP data plane identifies unencrypted flows
   via a 6-tuple that consists of two IP addresses, the transport
   protocol ID, two transport protocol port numbers, and the DSCP in the
   IP header.  When IPsec is used, the transport header is encrypted and
   the next protocol ID is an IPsec protocol, usually Encapsulating
   Security Payload (ESP), and not a transport protocol, leaving only
   three components of the 6-tuple, which are the two IP addresses and
   the DSCP.  If the IPsec sessions are established by a controller,
   then this controller could also transmit (in the clear) the Security
   Parameter Index (SPI) and thus the SPI could be used (in addition to
   the pair of IP addresses) for flow identification.  Identification of
   DetNet flows over IPsec is further discussed in Section 5.1.2.3 of
   [RFC8939].

   Sections below discuss threats specific to IP and MPLS in more
   detail.

10.1.  IP



   IP has a long history of security considerations and architectural
   protection mechanisms.  From a data plane perspective, DetNet does
   not add or modify any IP header information, so the carriage of
   DetNet traffic over an IP data plane does not introduce any new
   security issues that were not there before, apart from those already
   described in the data-plane-independent threats section (Section 5).

   Thus, the security considerations for a DetNet based on an IP data
   plane are purely inherited from the rich IP security literature and
   code/application base, and the data-plane-independent section of this
   document.

   Maintaining security for IP segments of a DetNet may be more
   challenging than for the MPLS segments of the network given that the
   IP segments of the network may reach the edges of the network, which
   are more likely to involve interaction with potentially malevolent
   outside actors.  Conversely, MPLS is inherently more secure than IP
   since it is internal to routers and it is well known how to protect
   it from outside influence.

   Another way to look at DetNet IP security is to consider it in the
   light of VPN security.  As an industry, we have a lot of experience
   with VPNs running through networks with other VPNs -- it is well
   known how to secure the network for that.  However, for a DetNet, we
   have the additional subtlety that any possible interaction of one
   packet with another can have a potentially deleterious effect on the
   time properties of the flows.  So the network must provide sufficient
   isolation between flows, for example, by protecting the forwarding
   bandwidth and related resources so that they are available to DetNet
   traffic, by whatever means are appropriate for the data plane of that
   network, for example, through the use of queuing mechanisms.

   In a VPN, bandwidth is generally guaranteed over a period of time
   whereas in DetNet, it is not aggregated over time.  This implies that
   any VPN-type protection mechanism must also maintain the DetNet
   timing constraints.

10.2.  MPLS



   An MPLS network carrying DetNet traffic is expected to be a "well-
   managed" network.  Given that this is the case, it is difficult for
   an attacker to pass a raw MPLS-encoded packet into a network because
   operators have considerable experience at excluding such packets at
   the network boundaries as well as excluding MPLS packets being
   inserted through the use of a tunnel.

   MPLS security is discussed extensively in [RFC5920] ("Security
   Framework for MPLS and GMPLS Networks") to which the reader is
   referred.

   [RFC6941] builds on [RFC5920] by providing additional security
   considerations that are applicable to the MPLS-TP extensions
   appropriate to the MPLS Transport Profile [RFC5921] and thus to the
   operation of DetNet over some types of MPLS network.

   [RFC5921] introduces to MPLS new Operations, Administration, and
   Maintenance (OAM) capabilities; a transport-oriented path protection
   mechanism; and strong emphasis on static provisioning supported by
   network management systems.

   The operation of DetNet over an MPLS network builds on MPLS and
   pseudowire encapsulation.  Thus, for guidance on securing the DetNet
   elements of DetNet over MPLS, the reader is also referred to the
   security considerations of [RFC4385], [RFC5586], [RFC3985],
   [RFC6073], and [RFC6478].

   Having attended to the conventional aspects of network security, it
   is necessary to attend to the dynamic aspects.  The closest
   experience that the IETF has with securing protocols that are
   sensitive to manipulation of delay are the two-way time transfer
   (TWTT) protocols, which are NTP [RFC5905] and the Precision Time
   Protocol [IEEE1588].  The security requirements for these are
   described in [RFC7384].

   One particular problem that has been observed in operational tests of
   TWTT protocols is the ability for two closely but not completely
   synchronized flows to beat and cause a sudden phase hit to one of the
   flows.  This can be mitigated by the careful use of a scheduling
   system in the underlying packet transport.

   Some investigations into protection of MPLS systems against dynamic
   attacks exist, such as [MPLS-OPP-ENCRYPT]; perhaps deployment of
   DetNets will encourage additional such investigations.

11.  IANA Considerations



   This document has no IANA actions.

12.  Security Considerations



   The security considerations of DetNet networks are presented
   throughout this document.

13.  Privacy Considerations



   Privacy in the context of DetNet is maintained by the base
   technologies specific to the DetNet and user traffic.  For example,
   TSN can use MACsec, IP can use IPsec, and applications can use IP
   transport protocol-provided methods, e.g., TLS and DTLS.  MPLS
   typically uses L2/L3 VPNs combined with the previously mentioned
   privacy methods.

   However, note that reconnaissance threats such as traffic analysis
   and monitoring of electrical side channels can still cause there to
   be privacy considerations even when traffic is encrypted.

14.  References



14.1.  Normative References



   [RFC8655]  Finn, N., Thubert, P., Varga, B., and J. Farkas,
              "Deterministic Networking Architecture", RFC 8655,
              DOI 10.17487/RFC8655, October 2019,
              <https://www.rfc-editor.org/info/rfc8655>.

   [RFC8938]  Varga, B., Ed., Farkas, J., Berger, L., Malis, A., and S.
              Bryant, "Deterministic Networking (DetNet) Data Plane
              Framework", RFC 8938, DOI 10.17487/RFC8938, November 2020,
              <https://www.rfc-editor.org/info/rfc8938>.

   [RFC8939]  Varga, B., Ed., Farkas, J., Berger, L., Fedyk, D., and S.
              Bryant, "Deterministic Networking (DetNet) Data Plane:
              IP", RFC 8939, DOI 10.17487/RFC8939, November 2020,
              <https://www.rfc-editor.org/info/rfc8939>.

   [RFC8964]  Varga, B., Ed., Farkas, J., Berger, L., Malis, A., Bryant,
              S., and J. Korhonen, "Deterministic Networking (DetNet)
              Data Plane: MPLS", RFC 8964, DOI 10.17487/RFC8964, January
              2021, <https://www.rfc-editor.org/info/rfc8964>.

14.2.  Informative References



   [ARINC664P7]
              ARINC, "Aircraft Data Network Part 7 Avionics Full-Duplex
              Switched Ethernet Network", ARINC 664 P7, September 2009.

   [BCP107]   Bellovin, S. and R. Housley, "Guidelines for Cryptographic
              Key Management", BCP 107, RFC 4107, June 2005.

              <https://www.rfc-editor.org/info/bcp107>

   [BCP72]    Rescorla, E. and B. Korver, "Guidelines for Writing RFC
              Text on Security Considerations", BCP 72, RFC 3552, July
              2003.

              <https://www.rfc-editor.org/info/bcp72>

   [DETNET-IP-OAM]
              Mirsky, G., Chen, M., and D. Black, "Operations,
              Administration and Maintenance (OAM) for Deterministic
              Networks (DetNet) with IP Data Plane", Work in Progress,
              Internet-Draft, draft-ietf-detnet-ip-oam-02, 30 March
              2021, <https://datatracker.ietf.org/doc/html/draft-ietf-
              detnet-ip-oam-02>.

   [DETNET-MPLS-OAM]
              Mirsky, G. and M. Chen, "Operations, Administration and
              Maintenance (OAM) for Deterministic Networks (DetNet) with
              MPLS Data Plane", Work in Progress, Internet-Draft, draft-
              ietf-detnet-mpls-oam-03, 30 March 2021,
              <https://datatracker.ietf.org/doc/html/draft-ietf-detnet-
              mpls-oam-03>.

   [DETNET-SERVICE-MODEL]
              Varga, B., Ed. and J. Farkas, "DetNet Service Model", Work
              in Progress, Internet-Draft, draft-varga-detnet-service-
              model-02, May 2017,
              <https://datatracker.ietf.org/doc/html/draft-varga-detnet-
              service-model-02>.

   [DETNET-YANG]
              Geng, X., Chen, M., Ryoo, Y., Fedyk, D., Rahman, R., and
              Z. Li, "Deterministic Networking (DetNet) YANG Model",
              Work in Progress, Internet-Draft, draft-ietf-detnet-yang-
              12, 19 May 2021, <https://datatracker.ietf.org/doc/html/
              draft-ietf-detnet-yang-12>.

   [IEEE1588] IEEE, "IEEE 1588 Standard for a Precision Clock
              Synchronization Protocol for Networked Measurement and
              Control Systems", IEEE Std. 1588-2008,
              DOI 10.1109/IEEESTD.2008.4579760, July 2008,
              <https://doi.org/10.1109/IEEESTD.2008.4579760>.

   [IEEE802.1AE-2018]
              IEEE, "IEEE Standard for Local and metropolitan area
              networks-Media Access Control (MAC) Security", IEEE Std.
              802.1AE-2018, DOI 10.1109/IEEESTD.2018.8585421, December
              2018, <https://ieeexplore.ieee.org/document/8585421>.

   [IEEE802.1BA]
              IEEE, "IEEE Standard for Local and metropolitan area
              networks--Audio Video Bridging (AVB) Systems", IEEE Std.
              802.1BA-2011, DOI 10.1109/IEEESTD.2011.6032690, September
              2011, <https://ieeexplore.ieee.org/document/6032690>.

   [IEEE802.1Q]
              IEEE, "IEEE Standard for Local and metropolitan area
              networks--Bridges and Bridged Networks", IEEE Std. 802.1Q-
              2014, DOI 10.1109/IEEESTD.2014.6991462, December 2014,
              <https://ieeexplore.ieee.org/document/6991462>.

   [IEEE802.1Qbv-2015]
              IEEE, "IEEE Standard for Local and metropolitan area
              networks -- Bridges and Bridged Networks - Amendment 25:
              Enhancements for Scheduled Traffic", IEEE Std. 802.1Qbv-
              2015, DOI 10.1109/IEEESTD.2016.8613095, March 2016,
              <https://ieeexplore.ieee.org/document/8613095>.

   [IEEE802.1Qch-2017]
              IEEE, "IEEE Standard for Local and metropolitan area
              networks--Bridges and Bridged Networks--Amendment 29:
              Cyclic Queuing and Forwarding", IEEE Std. 802.1Qch-2017,
              DOI 10.1109/IEEESTD.2017.7961303, June 2017,
              <https://ieeexplore.ieee.org/document/7961303>.

   [IETF-YANG-SEC]
              IETF, "YANG module security considerations", October 2018,
              <https://trac.ietf.org/trac/ops/wiki/yang-security-
              guidelines>.

   [IPSECME-G-IKEV2]
              Smyslov, V. and B. Weis, "Group Key Management using
              IKEv2", Work in Progress, Internet-Draft, draft-ietf-
              ipsecme-g-ikev2-02, 11 January 2021,
              <https://datatracker.ietf.org/doc/html/draft-ietf-ipsecme-
              g-ikev2-02>.

   [IT-DEF]   Wikipedia, "Information technology", March 2020,
              <https://en.wikiquote.org/w/
              index.php?title=Information_technology&oldid=2749907>.

   [MPLS-OPP-ENCRYPT]
              Farrel, A. and S. Farrell, "Opportunistic Security in MPLS
              Networks", Work in Progress, Internet-Draft, draft-ietf-
              mpls-opportunistic-encrypt-03, 28 March 2017,
              <https://datatracker.ietf.org/doc/html/draft-ietf-mpls-
              opportunistic-encrypt-03>.

   [NS-DEF]   Wikipedia, "Network segmentation", December 2020,
              <https://en.wikipedia.org/w/
              index.php?title=Network_segmentation&oldid=993163264>.

   [OT-DEF]   Wikipedia, "Operational technology", March 2021,
              <https://en.wikipedia.org/w/
              index.php?title=Operational_technology&oldid=1011704361>.

   [RFC2474]  Nichols, K., Blake, S., Baker, F., and D. Black,
              "Definition of the Differentiated Services Field (DS
              Field) in the IPv4 and IPv6 Headers", RFC 2474,
              DOI 10.17487/RFC2474, December 1998,
              <https://www.rfc-editor.org/info/rfc2474>.

   [RFC2475]  Blake, S., Black, D., Carlson, M., Davies, E., Wang, Z.,
              and W. Weiss, "An Architecture for Differentiated
              Services", RFC 2475, DOI 10.17487/RFC2475, December 1998,
              <https://www.rfc-editor.org/info/rfc2475>.

   [RFC3985]  Bryant, S., Ed. and P. Pate, Ed., "Pseudo Wire Emulation
              Edge-to-Edge (PWE3) Architecture", RFC 3985,
              DOI 10.17487/RFC3985, March 2005,
              <https://www.rfc-editor.org/info/rfc3985>.

   [RFC4253]  Ylonen, T. and C. Lonvick, Ed., "The Secure Shell (SSH)
              Transport Layer Protocol", RFC 4253, DOI 10.17487/RFC4253,
              January 2006, <https://www.rfc-editor.org/info/rfc4253>.

   [RFC4301]  Kent, S. and K. Seo, "Security Architecture for the
              Internet Protocol", RFC 4301, DOI 10.17487/RFC4301,
              December 2005, <https://www.rfc-editor.org/info/rfc4301>.

   [RFC4302]  Kent, S., "IP Authentication Header", RFC 4302,
              DOI 10.17487/RFC4302, December 2005,
              <https://www.rfc-editor.org/info/rfc4302>.

   [RFC4385]  Bryant, S., Swallow, G., Martini, L., and D. McPherson,
              "Pseudowire Emulation Edge-to-Edge (PWE3) Control Word for
              Use over an MPLS PSN", RFC 4385, DOI 10.17487/RFC4385,
              February 2006, <https://www.rfc-editor.org/info/rfc4385>.

   [RFC4432]  Harris, B., "RSA Key Exchange for the Secure Shell (SSH)
              Transport Layer Protocol", RFC 4432, DOI 10.17487/RFC4432,
              March 2006, <https://www.rfc-editor.org/info/rfc4432>.

   [RFC5586]  Bocci, M., Ed., Vigoureux, M., Ed., and S. Bryant, Ed.,
              "MPLS Generic Associated Channel", RFC 5586,
              DOI 10.17487/RFC5586, June 2009,
              <https://www.rfc-editor.org/info/rfc5586>.

   [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
              (BFD)", RFC 5880, DOI 10.17487/RFC5880, June 2010,
              <https://www.rfc-editor.org/info/rfc5880>.

   [RFC5905]  Mills, D., Martin, J., Ed., Burbank, J., and W. Kasch,
              "Network Time Protocol Version 4: Protocol and Algorithms
              Specification", RFC 5905, DOI 10.17487/RFC5905, June 2010,
              <https://www.rfc-editor.org/info/rfc5905>.

   [RFC5920]  Fang, L., Ed., "Security Framework for MPLS and GMPLS
              Networks", RFC 5920, DOI 10.17487/RFC5920, July 2010,
              <https://www.rfc-editor.org/info/rfc5920>.

   [RFC5921]  Bocci, M., Ed., Bryant, S., Ed., Frost, D., Ed., Levrau,
              L., and L. Berger, "A Framework for MPLS in Transport
              Networks", RFC 5921, DOI 10.17487/RFC5921, July 2010,
              <https://www.rfc-editor.org/info/rfc5921>.

   [RFC6071]  Frankel, S. and S. Krishnan, "IP Security (IPsec) and
              Internet Key Exchange (IKE) Document Roadmap", RFC 6071,
              DOI 10.17487/RFC6071, February 2011,
              <https://www.rfc-editor.org/info/rfc6071>.

   [RFC6073]  Martini, L., Metz, C., Nadeau, T., Bocci, M., and M.
              Aissaoui, "Segmented Pseudowire", RFC 6073,
              DOI 10.17487/RFC6073, January 2011,
              <https://www.rfc-editor.org/info/rfc6073>.

   [RFC6274]  Gont, F., "Security Assessment of the Internet Protocol
              Version 4", RFC 6274, DOI 10.17487/RFC6274, July 2011,
              <https://www.rfc-editor.org/info/rfc6274>.

   [RFC6478]  Martini, L., Swallow, G., Heron, G., and M. Bocci,
              "Pseudowire Status for Static Pseudowires", RFC 6478,
              DOI 10.17487/RFC6478, May 2012,
              <https://www.rfc-editor.org/info/rfc6478>.

   [RFC6562]  Perkins, C. and JM. Valin, "Guidelines for the Use of
              Variable Bit Rate Audio with Secure RTP", RFC 6562,
              DOI 10.17487/RFC6562, March 2012,
              <https://www.rfc-editor.org/info/rfc6562>.

   [RFC6632]  Ersue, M., Ed. and B. Claise, "An Overview of the IETF
              Network Management Standards", RFC 6632,
              DOI 10.17487/RFC6632, June 2012,
              <https://www.rfc-editor.org/info/rfc6632>.

   [RFC6941]  Fang, L., Ed., Niven-Jenkins, B., Ed., Mansfield, S., Ed.,
              and R. Graveman, Ed., "MPLS Transport Profile (MPLS-TP)
              Security Framework", RFC 6941, DOI 10.17487/RFC6941, April
              2013, <https://www.rfc-editor.org/info/rfc6941>.

   [RFC7384]  Mizrahi, T., "Security Requirements of Time Protocols in
              Packet Switched Networks", RFC 7384, DOI 10.17487/RFC7384,
              October 2014, <https://www.rfc-editor.org/info/rfc7384>.

   [RFC7567]  Baker, F., Ed. and G. Fairhurst, Ed., "IETF
              Recommendations Regarding Active Queue Management",
              BCP 197, RFC 7567, DOI 10.17487/RFC7567, July 2015,
              <https://www.rfc-editor.org/info/rfc7567>.

   [RFC7641]  Hartke, K., "Observing Resources in the Constrained
              Application Protocol (CoAP)", RFC 7641,
              DOI 10.17487/RFC7641, September 2015,
              <https://www.rfc-editor.org/info/rfc7641>.

   [RFC7748]  Langley, A., Hamburg, M., and S. Turner, "Elliptic Curves
              for Security", RFC 7748, DOI 10.17487/RFC7748, January
              2016, <https://www.rfc-editor.org/info/rfc7748>.

   [RFC7835]  Saucez, D., Iannone, L., and O. Bonaventure, "Locator/ID
              Separation Protocol (LISP) Threat Analysis", RFC 7835,
              DOI 10.17487/RFC7835, April 2016,
              <https://www.rfc-editor.org/info/rfc7835>.

   [RFC8446]  Rescorla, E., "The Transport Layer Security (TLS) Protocol
              Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018,
              <https://www.rfc-editor.org/info/rfc8446>.

   [RFC8578]  Grossman, E., Ed., "Deterministic Networking Use Cases",
              RFC 8578, DOI 10.17487/RFC8578, May 2019,
              <https://www.rfc-editor.org/info/rfc8578>.

   [RFC9016]  Varga, B., Farkas, J., Cummings, R., Jiang, Y., and D.
              Fedyk, "Flow and Service Information Model for
              Deterministic Networking (DetNet)", RFC 9016,
              DOI 10.17487/RFC9016, March 2021,
              <https://www.rfc-editor.org/info/rfc9016>.

   [RFC9023]  Varga, B., Ed., Farkas, J., Malis, A., and S. Bryant,
              "Deterministic Networking (DetNet) Data Plane: IP over
              IEEE 802.1 Time-Sensitive Networking (TSN)", RFC 9023,
              DOI 10.17487/RFC9023, June 2021,
              <https://www.rfc-editor.org/info/rfc9023>.

   [RFC9025]  Varga, B., Ed., Farkas, J., Berger, L., Malis, A., and S.
              Bryant, "Deterministic Networking (DetNet) Data Plane:
              MPLS over UDP/IP", RFC 9025, DOI 10.17487/RFC9025, April
              2021, <https://www.rfc-editor.org/info/rfc9025>.

   [RFC9056]  Varga, B., Ed., Berger, L., Fedyk, D., Bryant, S., and J.
              Korhonen, "Deterministic Networking (DetNet) Data Plane:
              IP over MPLS", RFC 9056, DOI 10.17487/RFC9056, June 2021,
              <https://www.rfc-editor.org/info/rfc9056>.

Contributors

   The Editor would like to recognize the contributions of the following
   individuals to this document.

   Stewart Bryant
   Futurewei Technologies

   Email: sb@stewartbryant.com


   David Black
   Dell EMC
   176 South Street
   Hopkinton, Massachusetts 01748
   United States of America


   Henrik Austad
   SINTEF Digital
   Klaebuveien 153
   7037 Trondheim
   Norway

   Email: henrik@austad.us


   John Dowdell
   Airbus Defence and Space
   Celtic Springs
   Newport, NP10 8FZ
   United Kingdom

   Email: john.dowdell.ietf@gmail.com


   Norman Finn
   3101 Rio Way
   Spring Valley, California 91977
   United States of America

   Email: nfinn@nfinnconsulting.com


   Subir Das
   Applied Communication Sciences
   150 Mount Airy Road
   Basking Ridge, New Jersey 07920
   United States of America

   Email: sdas@appcomsci.com


   Carsten Bormann
   Universitat Bremen TZI
   Postfach 330440 D-28359 Bremen
   Germany

   Email: cabo@tzi.org


Authors' Addresses



   Ethan Grossman (editor)
   Dolby Laboratories, Inc.
   1275 Market Street
   San Francisco, CA 94103
   United States of America

   Email: ethan@ieee.org
   URI:   https://www.dolby.com


   Tal Mizrahi
   Huawei

   Email: tal.mizrahi.phd@gmail.com


   Andrew J. Hacker
   Thought LLC
   Harrisburg, PA
   United States of America

   Email: andrew@thought.live