Boeing Research & Technology

P.O. Box 3707 Seattle WA 98124 USA fltemplin@acm.org

I-D Internet-Draft Mobile Ad-hoc NETworks (MANETs) present an interesting challenge for IPv6 addressing due to the indeterminant neighborhood properties of MANET interfaces. MANET routers must assign an IPv6 address to each MANET interface that is both unique and routable within the MANET but must not be forwarded to other networks. MANET routers must be able to assign self-generated addresses to their MANET interfaces when there is no infrastructure present that can coordinate topology-relative IPv6 addresses or prefixes. This document therefore specifies a means for MANET routers to generate and assign address types useful for MANET local communications.

When two or more IPv6 nodes come together within a common local operating region (e.g., during the formation of a Mobile Ad-hoc Network (MANET)), they must be able to assign unique addresses, discover multihop routes and exchange IPv6 packets with local network peers over their MANET interfaces even if there is no operator infrastructure present. MANETs include routers that configure interfaces to links with undetermined connectivity, in particular where the transitive property of connectivity for traditional shared links is not assured. MANET routers must nonetheless assign and use IPv6 addresses that are unique within the MANET. This is true even for nodes that configure multiple interface connections to the same MANET as a multilink routing domain. Section 6 of the "IP Addressing Model in Ad Hoc Networks" states that: "an IP address configured on this (MANET) interface should be unique, at least within the routing domain" and: "no on-link subnet prefix is configured on this (MANET) interface". The section then continues to explain why IPv6 Link-Local Addresses (LLAs) are of limited utility on links with undetermined connectivity, to the point that they cannot be used exclusively within multilink routing domains. suggests that global (aka "GUA") and unique-local (aka "ULA") addresses are MANET addressing candidates. However, provisioning of unique GUAs and ULAs must be coordinated either through administrative actions or through an automated address delegation service that all MANET routers can access. This document asserts that new forms of self-generated and unique MANET local IPv6 addresses are needed. The key feature of these MANET local IPv6 addresses is that they must be assured unique so that there is no chance of conflicting with an address assigned by another node. There is no requirement that the addresses have topologically-oriented prefixes, since the (newly-formed) local network may not (yet) connect to any other Internetworking topologies. The MANET local IPv6 addresses could then be used for continuous local communications and/or to coordinate topologically-oriented addresses for assignment on other interfaces. This would also manifest a new "MANET local" scope for the IPv6 scoped addressing architecture with scope greater than link-local but lesser than global/unique-local unicast. This document proposes a new unique local unicast address space known as "MANET-Local Addresses (MLAs)". MLAs use the formerly-deprecated IPv6 site-local prefix fec0::10 according to the address generation procedures specified in this document. The document further discusses the utility of the Hierarchical Host Identity Tag (HHIT) specified in for MANET local addressing purposes.

The IPv6 addressing architecture specified in , and defines the supported IPv6 unicast/multicast/anycast address forms with various scopes including link-local, site-local and others. Unique-local and global unicast addresses are typically obtained through Stateless Address AutoConfiguration (SLAAC) and/or the Dynamic Host Configuration Protocol for IPv6 (DHCPv6) , but these services require the presence of IPv6 network infrastructure which may not be immediately available in spontaneously-formed MANETs or other isolated local networks. A new IPv6 address type known as the Hierarchical Host Identity Tag (HHIT) (aka DRIP Entity Tag (DET)) provides a well-structured address format with exceptional uniqueness properties. A portion of the address includes the node's self-generated Overlay Routable Cryptographic Hash IDentifier (ORCHID) while the remainder of the address includes a well-formed IPv6 prefix plus bits corresponding to an attestation service that supports address proof-of-ownership. Verification of the attestation aspect of the address requires access to network infrastructure, but this may not always be available. Hence, a fully self-generated MLA may be necessary in environments where an HHIT cannot be used. MANET interfaces have the interesting property that a MANET router R will often need to forward packets between MANET nodes A and B even though R uses the same interface in the inbound and outbound directions. Since nodes A and B may not be able to communicate directly even though can both communicate directly with R, the link connectivity property is intransitive and the IPv6 Neighbor Discovery (ND) Redirect service cannot be used. Conversely, R may need to forward packets between nodes A and B via different MANET interfaces within a single MANET that includes multiple distinct links/regions. Due to these indeterminant (multi-)link properties, exclusive use of IPv6 Link Local Addresses (LLAs) is also out of scope. This document therefore introduces the MLA as a new fully self-generated IPv6 unicast address type that can be used either instead of or in addition to other IPv6 unicast address types. MLAs use the formerly-deprecated Site-Local IPv6 Address prefix fec0::10 according to the modified format shown in :

The node sets the first 10 bits of the MLA to the constant string '1111111011' then sets the 11th bit (i.e., the "(L)ocal" bit) to 1. The node next sets subnet ID to a 53 bit random value calculated the same as specified in Section 3.2.1 of for the Unique Local Address Global ID. The node finally generates and assigns a semantically opaque interface ID based on this self-generated prefix as specified in ; the resulting 128-bit MLA then has the proper format of an IPv6 address with a 64-bit "prefix" followed by a 64-bit interface identifier as required by the IPv6 addressing architecture. For example: fee7:6c29:de12:4b74:884e:9d2a:73fc:2d94 After a node creates an MLA, it can use the address within the context of spontaneously-organized local networks in which two or more nodes come together in the absence of supporting infrastructure and can still exchange IPv6 packets with little or no chance of address collisions. The use could be limited to bootstrapping the assignment of topologically correct IPv6 addresses through other means mentioned earlier, or it could extend to longer term usage patterns such as sustained communications with single-hop neighbors on a local link or even between multi-hop peers within a MANET. Note: the above MLA generation procedures apply when the L bit is set to 1; MLA generation procedures for L=0 may be specified by future documents.

IPv6 MLAs and HHITS have no topological orientation and can therefore be assigned to any of a node's MANET interfaces with a /128 prefix length (i.e., as a singleton address). The node can then begin to use MLAs or HHITs as the source/destination addresses of IPv6 packets that are forwarded over the interface within a local routing region. The node can assign the same MLA or HHIT to multiple interfaces all members of the same MANET, but must assign a different MLA or HHIT to the interfaces of each interface set connected to different MANETs. MLAs and HHITs may then serve as a basis for multihop forwarding over a MANET interface and/or for local neighborhood discovery over other IPv6 interface types. Due to their uniqueness properties, the node can assign an IPv6 MLA or HHIT to an interface without invoking (pre-service) Duplicate Address Detection (DAD), however it should deprecate an address if it detects a duplicate through (in-service) DAD.

Returning to a debate from more than 20 years ago, this document now proposes to reclaim the deprecated prefix "fec0::/10" for use as the MLA top-level prefix. documents the reasons for deprecation including the assertion that "Site is an Ill-Defined Concept". However, the concept of a MANET is a logical one based on (multihop) connectivity and not necessarily one constrained by physical boundaries. For example, a MANET router may connect to multiple distinct MANETs with a first set of interfaces connected to MANET "A", a second set of interfaces connected to MANET "B", etc. According to the scoped IPv6 addressing architecture, the router would assign a separate MLA for each interface set A, B, etc. and maintain separate MANET routing protocol instances for each set. MLAs A, B, etc. then become the router IDs for the separate routing protocol instances, but the MANET router may elect to redistribute discovered MLA routes between the instances. The uniqueness properties of MLAs and HHITs therefore transcends logical MANET boundaries but without "leaking" into external networks. The MLA prefix (formerly known as "Site-Local") has the distinct advantage that it is reserved and available for reclamation by a future standards track publication, for which this document qualifies. Upon publication as a standards track RFC, the RFC Editor is instructed to update , , and to reflect this new use for "fec0::/10".

MANET routers assign MLAs and/or HHITs to their MANET interfaces for use only within the scope of their locally connected MANETs. The MLAs and HHITs are carried in MANET routing protocol control messages and can also appear as the source and destination addresses for IPv6 packets forwarded within the locally connected MANETs. MLAs and HHITs cannot appear in the source or destination addresses for IPv6 packets forwarded beyond the locally connected MANETs, however, where an IPv6 Globally-Unique (GUA) and possibly also a companion Unique-Local (ULA) address is necessary. In order to support global-scope communications, each MANET router is required to obtain an IPv6 GUA through a border router/proxy that connects the MANET to the global IPv6 Internet. Since the proxy may be multiple MANET hops away, however, the MANET router configures and engages an Overlay Multilink Network (OMNI) Interface as specified in . The MANET router assigns the GUA to the OMNI interface which forwards original packets by inserting an IPv6 encapsulation header that uses MLAs or HHITs as the source/destination addresses while the original packet uses GUAs. The proxy may be configured as an IPv6-to-IPv6 Network Prefix Translation (NPTv6) gateway that connects the rest of the MANET to the outside Internet. In that case, the proxy supplies each MANET router with a ULA internally and maintains a 1:1 relationship between the ULA on the "inside" and a GUA on the "outside" as discussed in . Each MANET router will then assign the ULA to its OMNI interface which encapsulates each original packet in an IPv6 header that uses MLAs or HHITs as the encapsulation source/destination addresses while the original packet uses ULAs/GUAs. The NPTv6 gateway will then statelessly translate each ULA into its corresponding GUA (and vice versa) for packets that transit the proxy. The proxy delegates a ULA and its companion Provider-Aggregated (PA) GUA for each MANET router which assigns them according to either the "ULA+PA" or "ULA-Only" connected network models . If the MANET router employs the "ULA+PA" model, it can assign both the ULA and GUA to the OMNI interface. The MANET router can then use the ULA for local-scoped communications with peers inside the MANET and the GUA for global-scoped communications with external peers via the proxy as a "NPTv6 pass-through". MANET routers can then select address pair combinations according to IPv6 default address selection rules . After receiving a ULA plus PA GUA delegation, MANET routers that require Provider-Independent (PI) GUAs can use the OMNI interface in conjunction with the Automatic Extended Route Optimization (AERO) global distributed mobility management service to request and maintain IPv6 and/or IPv4 PI prefixes from the mobility service. The MANET router can then sub-delegate GUAs from the PI prefixes to its attached downstream local networks which may in turn engage an arbitrarily large IPv6 and/or IPv4 "Internet of Things".

"Default Address Selection for Internet Protocol Version 6 (IPv6)" provides a policy table that specifies precedence values and preferred source prefixes for destination prefixes. "Preference for IPv6 ULAs over IPv4 addresses in RFC6724" updates the policy table entries for ULAs, IPv4 addresses and the 6to4 prefix (2002::/16). This document proposes a further update to the policy table for IPv6 MLAs (prefix fec0::/10) and HHITs (prefix 2001:30::/28). The proposed updates appear in the table below:

With the proposed updates, IPv6 HHITs now appear as a lesser precedence than IPv6 GUAs, IPv6 ULAs and IPv4 addresses but as a greater precedence than IPv6 MLAs. IPv6 MLAs now appear as a greater precedence than deprecated IPv6 prefixes but a lesser precedence than all other address types.

IPv6 nodes MAY assign self-generated IPv6 MLAs and/or HHITs to their interface connections to local networks (or MANETs). If the node becomes aware that the address is already in use by another node, it instead generates and assigns a new MLA/HHIT. IPv6 routers MAY forward IPv6 packets with MLA/HHIT source or destination addresses over multiple hops within the same local network (or MANET). IPv6 routers MUST NOT forward packets with MLA/HHIT source or destination addresses to a link outside the packet's local network (or MANET) of origin. IPv6 routers MUST NOT advertise the prefix fec0::/10 (or any IPv6 prefixes reserved for HHITs) in routing protocol exchanges with correspondents outside the local network (or MANET). The default behavior of exterior routing protocol sessions between administrative routing regions must be to ignore receipt of and not advertise prefixes in the fee0::/11 block. At the present time, AAAA and PTR records for MLAs in the fee0::/11 block are not recommended to be installed in the global DNS.

In progress.

instructed IANA to mark the fec0::/10 prefix as "deprecated", and as such it does not appear in the IANA IPv6 Special-Purpose Address Registry. Upon publication, IANA is instructed to add the prefix fec0::/10 to the 'iana-ipv6-special-registry' registry with the name "MANET-Local Unicast" and with RFC set to "[RFCXXXX]" (i.e., this document).

IPv6 MLAs include very large uniquely-assigned bit strings in both the prefix and interface identifier components. With the random prefix generation procedures specified in and the semantically opaque interface identifier generation procedures specified in the only apparent opportunity for address duplication would be through either intentional or unintentional misconfiguration. A node that generates an MLA and assigns it to an interface should therefore be prepared to deprecate the MLA and generate/assign a new one if it detects a legitimate duplicate.

This work was inspired by continued investigations into 5G MANET operations in cooperation with the Virginia Tech National Security Institute (VTNSI). Emerging discussions on the IPv6 maintenance (6man) mailing list continue to shape updated versions of this document. The author acknowledges all those whose useful comments have helped further the understanding of this proposal. Honoring life, liberty and the pursuit of happiness.

<< RFC Editor - remove prior to publication >> Differences from earlier versions: First draft publication.