Pekka P„„kk÷nen Document: draft-paakkonen-addressing-htr-manet-00.txt Mika Rantonen Expires: June 2004 Juhani Latvakoski VTT Electronics December 2003 IPv6 addressing in a heterogeneous MANET-network draft-paakkonen-addressing-htr-manet-00.txt Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC 2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet- Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet- Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http://www.ietf.org/1id-abstracts.html The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html Copyright Notice Copyright (C) The Internet Society (2003). All Rights Reserved. Abstract This document presents IPv6 addressing related to end-to-end connectivity in a Mobile Ad hoc Network with heterogeneous nodes. Internet connectivity for the mobile ad hoc network has to be provided by an Access Router. The nodes in such a network could only have IPv6 functionality, or contain support for Mobile IPv6 (MIPv6) or facilitate ad hoc routing by using a MANET-based routing protocol. End-to-end connectivity means that the nodes must be able to communicate with each other locally and over the Internet. The draft focuses on the Ad hoc On-demand Distance Vector (AODV) as a MANET-based routing protocol. Table of contents Abstract i 1. Introduction i 2. Terminology 2 3. Terms 2 P„„kk÷nen et al. Expires July 2004 [Page i] INTERNET-DRAFT End-to-end connectivity December 2003 4. IPv6/MIPv6 addressing 2 4.1. Functionality it the AR.......................................4 5. MANET-addressing 4 5.1. Duplicate Address Detection (DAD).............................4 5.2. Global connectivity...........................................5 5.3. Global vs. site-local addresses...............................5 5.3.1. AODV-MIPv6 co-operation.................................5 6. IPv6-MANET addressing 6 6.1. AODV communication algorithm..................................6 6.2. Address autoconfiguration and communication issues............6 7. NEMO addressing 8 8. Security issues 8 9. Open issues 8 References 9 Author's addresses 9 APPENDIX A IPv6/MIPv6 addressing example 10 A.1. Proxy Neighbor Discovery (PND) in the AR.....................10 A.2. Redirect ICMP message........................................11 APPENDIX B Global connectivity via next hop routing 11 Full Copyright Statement 13 1. Introduction This document describes IPv6 addressing, which is related to the end-to-end connectivity in a heterogeneous Mobile Ad hoc NETwork (MANET). The heterogeneity means that the nodes in the MANET-network may contain different capabilities, which has been illustrated in figure 1. First of all nodes with only IPv6 functionality might be present. Also Mobile IPv6 (MIPv6) [1] enabled nodes might be present, either MIPv6 Corresponding Nodes (MIPv6_CN) or MIPv6 Mobile Nodes (MIPv6_MN). MIPv6_CNs have MIPv6 Route Optimization (RO) capability, but don't support mobility [1]. MIPv6_MNs have MIPv6 mobility extensions [1]. These nodes are referred in this document to as IPv6/MIPv6-nodes. MANET routing protocol supported nodes might also be present, which in addition have similarly different capabilities as the IPv6/MIPv6-nodes (IPv6/ MIPv6_CN/ MIPv6_MN). In this document Ad hoc On-demand Distance Vector (AODV) routing protocol functionality has been focused on. The nodes of the MANET-network have to be able to communicate with each other, and also with a Corresponding Node (CN) of the Internet (end-to-end connectivity). P„„kk÷nen et al. Expires July 2004 [Page 1] INTERNET-DRAFT End-to-end connectivity December 2003 CN HA_MN | | | | _________|_________|________ \ / \ Internet /---------HA_MR \_____________________/ ______________________|__________________________ ( | ) ( AR/MR --------MANETnode ) ( / \ | | ) ( / | \ | | ) ( / | \ | | ) ( / | \ | MANET ) ( / | \ | node ) ( / | \ | / ) ( IPv6------MIPv6_CN---MIPv6_MN / ) ( node ) ( ) ( Heterogeneous MANET- ) ( network ) (_______________________________________________) Figure 1. Heterogeneity in a hybrid MANET-network. If the heterogeneous MANET-network needs global Internet connectivity, it has to have one or more Access Routers (AR), which is/are connected to the Internet. This AR is static if the MANET-network doesn't move in relation to the Internet topology. Network Mobility might also be supported as defined in the NEMO working group [2]. In this case the ARs are considered as Mobile Routers (MR), which maintain Internet connectivity by using bi-directional tunneling with their corresponding Home Agent (HA), when the mobile network is away from home. IPv6 addressing related to IPv6/MIPv6-nodes can be referred to as IPv6/MIPv6-addressing, and IPv6 addressing related to MANET-nodes MANET-addressing. IPv6-MANET addressing comprises communication between IPv6/MIPv6-nodes and MANET-nodes. The heterogeneous nodes of the mobile ad hoc network should also be able to communicate with a Corresponding Node (CN) of the Internet outside of the MANET-network, which can be referred to as global communication. The CN might also be an IPv6-node, a MIPv6_CN or a MIPv6_MN. End-to-end connectivity in a heterogeneous MANET-network consists of the different local and global communication use cases between the heterogeneous nodes. This document describes the IPv6 addressing related to end-to-end connectivity in such a heterogeneous multi-hop MANET-network. The structure of the document is as follows: Chapter 3 describes terms used in the document. The next three chapters deal with the IPv6 addressing related to the heterogeneous MANET-network. NEMO addressing is focused on in chapter 7. Chapter 8 discusses security issues, and chapter 9 presents issues for further P„„kk÷nen et al. Expires July 2004 [Page 2] INTERNET-DRAFT End-to-end connectivity December 2003 development. 2. Terminology The keywords "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [3]. 3. Terms This document uses mobility related terms as defined in [15], and network mobility related terms as defined in [16]. In addition the following terms are used: Access Router (AR) AR provides IP connectivity for the nodes by default routing, and acts as an Internet gateway. The AR could also be considered as a Mobile Router (MR) if network mobility is supported by it. Flat Routing Flat routing considers the ad hoc network without subnet partitioning Hierarchical Routing The ad hoc network is considered as logically separated subnets. IPv6-node Node with plain IPv6 functionality [1]. Could also be considered as a Fixed Node (FN), because an IPv6-node doesn't support mobility. Mobile IPv6 Corresponding Node (MIPv6_CN) Mobile IPv6 node with Route Optimization capabilities as defined in [1]. Mobile IPv6 Mobile Node (MIPv6_MN) Mobile IPv6 node with mobility support as defined in [1]. 4. IPv6/MIPv6 addressing The IPv6/MIPv6 nodes configure a link-local address which may be used only if the destination is on a local link. Global addresses are needed if the nodes need to communicate over the local link or over the Internet. Either stateless [4] or stateful [5] address autoconfiguration could be used for global address generation. In this document stateless address autoconfiguration has been focused on. In case of stateless address autoconfiguration, the AR sends Router Advertisements (RA), and includes an IPv6 prefix to the Prefix P„„kk÷nen et al. Expires July 2004 [Page 3] INTERNET-DRAFT End-to-end connectivity December 2003 Information -option [6], which the IPv6/MIPv6-nodes attach to the interface-ID to create a global IPv6 address [7]. To enable default routing for the nodes, the Router Lifetime field of the Router Advertisement (RA) MUST be non-zero [6]. When the nodes receive a RA with a non-zero Router Lifetime field, an entry is added to the default router list [6]. (Implementation note: Some platforms create a Routing Table (RT) entry for the AR's link-local address (::/0 -> AR's link-local address) ). If the Router Lifetime would be zero, default routing would be disabled, and outside communication with a CN of the Internet would not be possible. The IPv6 prefix is advertised as ON-link by setting the L-flag in the Prefix Information -option of the RA [6]. When a IPv6/MIPv6-node considers the IPv6 prefix to be ON-link, it sends the packets for the destination to the interface. (Implementation note: On some platforms an RT entry is created for the IPv6 prefix towards the interface of the AR (IPv6 prefix -> ethx) ). The ON-link prefix causes the IPv6/MIPv6- nodes to send packets to the interface for destinations, which have configured an IPv6 address from the IPv6 prefix (local destinations). If the L-flag is not set, it "conveys no information concerning on-link determination and MUST NOT be interpreted to mean that addresses covered by the prefix are off-link" [6]. This causes the IPv6/MIPv6-nodes to use the default router for all destinations. For clarity in this document the ON-link model is used, when the L-flag is set, and the OFF-link model is used, when the L-flag is not set. Figure 2 describes an example of a situation in which the AR advertises an IPv6 Prefix, with a Router Lifetime of 5000 in the ON-link addressing model. It causes the IPv6/MIPv6-node to add entries for the default router and IPv6 prefix to the Routing Table (RT). In the example the default router list has been implemented by using the RT. AR | | | Router Advertisement | || Router Lifetime = 5000 | || Prefix Information Option | \ / L-flag = 1 | \/ A-flag = 1 | Prefix = IPv6 Prefix | Source Link-layer Address Option | Link-layer address of AR | IPv6/MIPv6-node RT: ::/0 -> AR's link-local address IPv6 Prefix -> ethx Figure 2. Stateless address autoconfiguration. When the AR advertises the IPv6 prefix with the ON-link model, it causes the communicating nodes to send packets directly to each other (to the P„„kk÷nen et al. Expires July 2004 [Page 4] INTERNET-DRAFT End-to-end connectivity December 2003 interface). The ON-link addressing model might be a problem with certain access technologies. For example consider a case when IEEE 802.11b Wireless LAN (WLAN) is used in the ad-hoc mode, and the IPv6/MIPv6 nodes communicate locally on the same link. If the global addresses are used, communication isn't possible if the nodes are positioned in such a way in which the nodes are not on each other's WLAN communication range. This is caused by the distance between the nodes (direct radio connectivity isn't possible), when the nodes send packets directly to each other (to the interface). The communication could however be enabled by advertising the IPv6 prefix as OFF-link or by executing Proxy Neighbor Discovery in the AR, if the AR is spatially positioned in such a way, that it is able to provide routing for the nodes (i.e. AR is in the radio communication range of both nodes). This would route the packets via the AR. In this particular case, routing via the AR is also the downside of the OFF-link model, when the peers are in the radio communication range of each other (unoptimal routing). Specific details on how the ON-link addressing model, Proxy Neighbor Discovery (PND) and Redirect ICMP messages relate to the ad hoc mode of WLAN are described in Appendix A. 4.1. Functionality in the AR The access technology used and the network topology of the Access Network SHOULD be taken into account when making the choice of the ON-link addressing model of the IPv6 prefix in the AR. As described in appendix A the OFF-link model could be used for WLAN environments which use the ad-hoc mode (if the PND function is not used in the AR in the ON-link model). But if the WLAN infrastructure mode with access points would be used the choice of the addressing model could be different. The PND function MAY be executed with the ON-link model with some access technologies as described in appendix A. The access technology SHOULD also be taken into account, when making the choice of enabling/disabling Redirect ICMP message sending in the AR. As described in appendix A the sending of Redirect-ICMP messages MAY be disabled in the AR, when WLAN is used in the ad-hoc mode. 5. MANET-addressing The MANET-nodes configure a MANET-address, which is used for communication. Either a site-local or global address could be configured (although the use of the site-local address has been deprecated by the IETF [8]). To create a unique MANET-address, Duplicate Address Detection (DAD) must be used. 5.1. Duplicate Address Detection (DAD) The main purpose of the DAD procedure is to guarantee the uniqueness of the IPv6 address to be used in the MANET. The uniqueness test messages (Address Request (AREQ), Address Reply (AREP)) have to disseminate over the MANET. The DAD described in [9] could be used, and the DAD could also be improved as described in [10]. The latter solution performs the DAD only on the interface-ID part of the MANET-address to be created, P„„kk÷nen et al. Expires July 2004 [Page 5] INTERNET-DRAFT End-to-end connectivity December 2003 and also reduces the unnecessary forwarding of the AREQ-message. 5.2. Global connectivity To enable global connectivity for a MANET-node, the MANET-node must have a globally unique address, the packets from the Internet must be routable to the MANET-node, and the MANET-node must be able to send packets to the Internet. The MANET-nodes have to configure a global MANET-address by acquiring an IPv6 prefix from the AR, which is attached to the interface-ID of the node. This functionality could be achieved as described in [11]. In this case the nodes configure three routing table entries to enable global routing (default-route/0 -> AR; AR -> next-hop towards AR; next-hop -> ethx). Appendix B describes how global connectivity could be achieved via next-hop routing by using a reactive MANET-protocol (AODV). In this case only two routing table entries are needed (default-route/0 -> next-hop; next-hop -> ethx). 5.3. Global vs. site-local addresses When the AODV-nodes communicate with each other, route discovery for on-demand routing protocols are used as described in [12]. This means that the route is discovered on-demand based on the current ad hoc network topology. The MANET-node may use either a site-local address or a global address configured via IPv6 prefix discovery as a MANET-address, which means that both addresses cannot be used at the same time. If the site-local address is used as a MANET-address, Internet communication is not possible, because packets with a site-local source address cannot be routed to the Internet. This means also that it is not possible to communicate with local MANET-nodes, which are away from home and use the home address as the source address for local communication. The site-local address can be used in multiple sites and the address itself doesn't contain an indication about a particular site. This kind of ambiguousness presents problems to the application developers and multi-sited routers. This has led the IETF to deprecating the use of site-local address [8], and search for alternative solutions [13]. This discourages further the use of site-local addresses instead of global addresses. It remains to be seen how the alternative addressing solution can be used with AODV [13]. 5.3.1. AODV-MIPv6 co-operation A typical way for a MIPv6_MN to choose a default router and configure a COA is to receive RAs sent by the AR. If AODV is used as a MANET-routing protocol, the default router and COA are configured via a reactive IPv6 Prefix discovery sequence. It is desirable for an AODV supported MIPv6_MN to use its home address when at home, and its COA when away from home to avoid unnecessary tunneling. To enable this, the choice of being in the home or foreign network should be done after the IPv6 Prefix discovery sequence by comparing the received IPv6 Prefix to the home network prefix similarly as the IPv6 prefixes advertised with P„„kk÷nen et al. Expires July 2004 [Page 6] INTERNET-DRAFT End-to-end connectivity December 2003 RAs are used for movement detection in standard MIPv6 networks. If the home network prefix is equal to the IPv6 Prefix, the MIPv6_MN uses the configured address as its home address and refreshes the route for it. Otherwise the address is used as a COA, which is registered after the route refresh sequence to the HA of the MIPv6_MN. If the MIPv6_MN moves away from the home MANET-network, and continues to use the configured MANET-address as its home address, its HA has to defend the home address by answering to AODV's DAD messages, so that no other node can configure the same address in the MANET-network. This functionality should be similar as the standard MIPv6 home address defending at the HA [1]. 6. IPv6-MANET addressing The MANET-nodes should be able to communicate with local IPv6/MIPv6- nodes and CNs in the Internet. This chapter describes IPv6 addressing issues related to such use cases. 6.1. AODV communication algorithm The following communication algorithm supports hierarchical routing for COAs and flat routing for site-local addresses: 1. If the destination is a site-local MANET-address or the destination's IPv6-prefix is equal with the IPv6 Prefix received via IPv6 Prefix Discovery => execute route discovery 2. Otherwise send packets via the default router The flat routing approach has to be used for the site-local addresses, because site-local addresses don't contain any information about a specific subnet. The need for route discovery to COAs prior to communication is based on the IPv6 Prefix. In this case the packets between multiple MANETs using COAs would flow via the ARs of the different MANETs. In [11] it has been defined that the MANET-node MAY use route discovery always when sending packets whether or not the destination is in the Internet or not. If in this case the CN is in the Internet and no answer to the RREQ is received, the CN is deduced to be outside the MANET, and the default router is used. The algorithm presented in this document uses the default router immediately, when packets are sent to local IPv6/MIPv6-nodes or CNs in the Internet. This means that the initial route discovery phase is not needed, which results in faster communication initiation with the peer compared to the solution presented in [11]. There is also no need to add destination entries to the routing table via the default router, (destination -> default-router) which are needed, if route discovery is used always when initiating communication with a random node at the first time. 6.2. Address autoconfiguration and communication issues IPv6/MIPv6-node related DAD guarantees that the interface-ID of the P„„kk÷nen et al. Expires July 2004 [Page 7] INTERNET-DRAFT End-to-end connectivity December 2003 IPv6-address is unique on a link [4]. The IPv6/MIPv6-nodes discover the IPv6 prefix via the RAs sent by the AR, which is attached to the interface-ID to configure a unique IPv6 address. Also a link-local address is created by attaching a well-known link-local prefix (fe80::/64) to the interface-ID [4][7]. MANET-address related DAD guarantees that the interface-ID is unique in the MANET [10]. The IPv6 prefix used in the MANET is received via IPv6 Prefix discovery, which is attached to the MANET related interface-ID to create a unique IPv6 address. It is possible that equal interface-IDs are configured in a heterogeneous MANET-network, because the scope of the DAD-procedures related to the interface-IDs overlap. To guarantee unique global IPv6 addresses, the IPv6 prefixes used in the different autoconfiguration procedures have to be different. End-to-end connectivity between heterogeneous nodes is described in figure 3. Because the IPv6 prefixes have to be different, it doesn't matter which addressing model is used with the IPv6-MANET communication. Also PND is not required in the AR, because the IPv6-MIPv6-nodes send always packets for the MANET-node via its configured default router (::/0 -> AR's link-local address). The MANET-node also uses its configured default router for communication with the IPv6/MIPv6-node (::/0 -> next hop towards MR) without using the route discovery procedure, because the IPv6 prefixes are different. Global addresses have to be used always when the default router is used, in case the CN would be in the Internet instead of being in a local IPv6/MIPv6-node. If the IPv6/MIPv6-node and MANET-node are on the same link and in the radio communication range of each other, the only possible way for these nodes to communicate directly with each other is by using link-local addresses , because the default router is used in other situations for routing between the nodes. P„„kk÷nen et al. Expires July 2004 [Page 8] INTERNET-DRAFT End-to-end connectivity December 2003 CN HA_MN | | | | _________|_________|________ \ / \ Internet /---------HA_MR \_____________________/ ___________________|____________________________________ ( | ) ( AR/MR RT: IPv6-prefix -> ethx ) ( /\ /\ MANET1-node -> ethx ) ( / \ MANET2-node -> MANET1-node) ( / MANET1-node ) ( / \ RT: ::/0 -> AR ) ( / \ MANET2 -> ethx ) ( \/ \/ ) ( IPv6/MIPv6<----------->MANET2-node ) ( node link-local ) ( addresses ) ( RT: ::/0 -> AR RT: ::/0 -> next-hop ) ( towards AR ) (___________________________________next-hop -> ethx_____) Figure 3. End-to-end connectivity in a MANET-network with heterogeneous nodes. If the AR sends Redirect ICMP-messages and the AODV-node and NEMO-node are on the same link, the Redirect-messages cause the same problem, which is present in IPv6/MIPv6-communication, but only for communication from the IPv6/MIPv6-node to the AODV-node. 7. NEMO addressing As mentioned before the AR could also be a MR, which would make the heterogeneous MANET-network mobile. In this case the NEMO bi-directional tunneling used on the MR's egress interface is transparent to the end-to-end connectivity of the nodes in the MANET-network, because seamless mobility is supported by the NEMO-approach i.e. the IP addresses configured for the nodes of the mobile network don't change when the network moves. In any case the AR/MR needs two IPv6 prefixes for the autoconfiguration of both MANET and IPv6/MIPv6-nodes. These prefixes have to be either statically configured for the AR/MR, or delegated dynamically from some network entity for example via DHCPv6 [5] [14]. 8. Security issues Security issues have not been considered in this document, but should be taken into account in a future version of this document. 9. Open issues P„„kk÷nen et al. Expires July 2004 [Page 9] INTERNET-DRAFT End-to-end connectivity December 2003 The source address selection of a node in the MANET-network is an issue for further work. For example when a MIPv6_MN communicates with a local IPv6-node, the COAs should be used for communication to avoid tunneling via the HA_MN (instead of using home addresses). Security issues should also be taken into account in the future. References [1] D. B. Johnson, C. E. Perkins, J. Arkko "Mobility Support in IPv6" , Internet Draft, June 2003. [2] NEtwork MObility working group website URL: http://www.ietf.org/html.charters/nemo-charter.html. [3] S. Bradner "Key words for use in RFCs to Indicate Requirement Levels" BCP 14, RFC 2119, March 1997. [4] S. Thomson and T. Narten, "IPv6 Stateless Address Autoconfiguration", RFC 2462, December 1998. [5] R. Droms et al. "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)", RFC 3315, July 2003. [6] T. Narten, E. Nordmark and W. Simpson, "Neighbor Discovery for IP Version 6 (IPv6)", RFC 2461, December 1998. [7] R. Hinden, S. Deering "IP Version 6 Addressing Architecture" , Internet Draft October 2003. [8] C. Huitema, B. Carpenter "Deprecating Site Local Addresses" , Internet Draft, November 2003. [9] C.E. Perkins et al. "IP Address Autoconfiguration for Ad Hoc Networks" , Internet draft, November 2001. [10] M. Rantonen, J. Keisala "IP Address Autoconfiguration with DAD minimization for Ad Hoc Networks" , Internet draft, August 2003. [11] R. Wakikawa et al. "Global Connectivity for IPv6 Mobile Ad Hoc Networks" Internet draft, November 2002. [12] C.E. Perkins, E.M. Belding Royer, S. Das "Ad hoc On-Demand Distance Vector (AODV) Routing" RFC 3561. [13] R. Hinden, B. Haberman "Unique Local IPv6 Unicast Addresses" , Internet draft, September 2003. [14] R. Droms "DHCPv6 Prefix Delegation for NEMO" , Internet draft, June 2003. [15] J. Manner, M. Kojo "Mobility Related Terminology" , Internet draft April 2003. [16] T. Ernst, H. Lach "Network Mobility Support Terminology" , Internet draft, May 2003. Author's addresses Pekka P„„kk÷nen VTT Technical Research Centre Of Finland (VTT Electronics) Kaitov„yl„ 1 P„„kk÷nen et al. Expires July 2004 [Page 10] INTERNET-DRAFT End-to-end connectivity December 2003 90571 Oulu Finland email: pekka.paakkonen@vtt.fi Mika Rantonen VTT Technical Research Centre Of Finland (VTT Electronics) Kaitov„yl„ 1 90571 Oulu Finland email: mika.rantonen@vtt.fi Juhani Latvakoski VTT Technical Research Centre Of Finland (VTT Electronics) Kaitov„yl„ 1 90571 Oulu Finland email: juhani.latvakoski@vtt.fi Appendix A: IPv6/MIPv6 addressing example This appendix describes how Proxy Neighbor Discovery and Redirect ICMP messages are related to the ad-hoc mode of the IEEE WLAN 802.11b access technology. A.1. Proxy Neighbor Discovery (PND) in the AR PND MAY be used with certain access technologies, when the ON-link addressing model is used in the AR. Figure 4 describes a situation, in which the peers cannot communicate directly with each other, but the AR is able to provide routing for the nodes (the circles represent the wireless communication ranges of the nodes). The PND function means that the AR answers on behalf of the destination with a Neighbor Advertisement (NA) to the Neighbor Solicitation (NS) message sent by the source (steps 1 and 2). The NA contains the link-layer address of the AR instead of the destination. This causes the source to send packets for the destination to the AR, which routes the packets to the final destination (steps 3 and 4). This means that the AR accepts packets not explicitly addressed to it. In the NA sent by the AR the Override-flag is set to zero, so that the NA sent by the real destination with the Override-flag set overrides the NA sent by the AR. This feature causes the source to send packets to the real destination when the peers are within each other's communication range. ________________________________________________________________ ( [ ) ] ( [ ) ] ( 4. IPv6 packet [ ) 1.NS 3.IPv6 packet ] ( |-----------------------[---AR<--)----------------------| ] ( \/ [ | ) | ] ( Node2 [ |---)------------------->Node1 ] (___________________________[________) 2.NA ] [____________________________________] Figure 4. Proxy Neighbor Discovery (PND). P„„kk÷nen et al. Expires July 2004 [Page 11] INTERNET-DRAFT End-to-end connectivity December 2003 Consider the local mobility of the nodes in figure 4. The nodes may be mobile in such a way that they may or may not be in each other's communication range during a UDP/TCP session. In this case the IP addresses don't change, because only one AR is present which sends RAs. Reachability of the neighbor (destination) is dependent on the upper layer information and Neighbor Unreachability Detection (NUD) as defined in [6]. A downside of the PND function is that the AR has to know the IP and link-layer addresses of the peers. Also security problems are present, because a malicious AR might execute PND for the destination. The advantage of the PND in the ON-link addressing model is that the traffic between the local nodes are sent directly between the nodes, when the nodes are in each other's communication range. In the OFF-link addressing model the traffic would always goes via the AR. A.2. Redirect ICMP message Another thing related to some access technologies is the sending of the Redirect ICMP-messages. When the source sends packets via the AR to the destination, the AR discovers that the destination is in fact on the same link as the source, and sends a Redirect ICMP message to the source. The Redirect-message contains the link-layer address of the destination, which causes the source to send packets to the real destination instead of the routing via the AR, because the source updates the link-layer address of the destination's IP address in the neighbor cache [6]. If the peers are not within the radio communication range of each other (figure 4) and the AR routes the packets between the peers, communication ceases between the peers. This situation occurs when the IPv6 prefix is advertised as OFF-link or the PND function is used with the ON-link model in the AR. Figure 5 describes a situation in which the AR sends a Redirect-message with the destination's link-layer-address (LL-address), which causes the source to update its neighbor cache. AR-----Destination | | | || Redirect ICMP-message | || Target link-layer address | \ / Destination's LL-address | \/ | IPv6/MIPv6-node Neighbor cache: IPv6 address -> AR's LL-address/ Destination's LL-address Figure 5. Redirect ICMP messages. Appendix B: Global connectivity via next-hop routing P„„kk÷nen et al. Expires July 2004 [Page 12] INTERNET-DRAFT End-to-end connectivity December 2003 This appendix describes how global connectivity could be achieved with AODV by using next-hop routing, and is similar to the one described in [11]. This mechanism has been tested in a testbed and has been described in figure 6. AR /|\ RT: Int.node -> eth0 1.RREQ (I-flag set) | 5. AODV-node -> Int.node 2.RREP (I-flag set) | 4.RREQ | 6.RREP | \|/ Int.node RT: 5.AODV-node -> eth0 /|\ 3.MR ->eth0 1.RREQ (I-flag set) | 3.::/0 -> AR 2.RREP (I-flag set) | 4.RREQ | 6.RREP | \|/ RT: 3. AR -> Int.node AODV-node 3. ::/0 -> Int.node Int.node -> eth0 Figure 6. Global connectivity via next-hop routing. The global connectivity is enabled with IPv6 prefix discovery (steps 1-3) and route refresh sequences (steps 4-6). IPv6 prefix discovery is used for the address autoconfiguration purposes to enable global connectivity for the AODV-node, and the route refresh procedure maintains AODV-node reachability from the AR i.e. from the Internet. The IPv6 Prefix discovery sequence begins when the AODV-node sends a RREQ with the I-flag set to the ALL_MANET_GW_MCAST-multicast address [11] from its configured site-local MANET-address (step1 in figure 6). The site-local addresses's uniqueness has been tested in the DAD procedure. The AR returns the IPv6 Prefix by sending a RREP with the I-flag set to the AODV-node's MANET-address (step2). When the AODV-node and intermediate nodes receive this RREP, a RT entry for the default router and AR towards the next-hop towards the AR are created (step3). The AR's address is a global address configured on the ingress interface of the AR. After this the AODV-node attaches the IPv6 Prefix to the interface-ID configured during the MANET-related DAD to create a unique global IPv6 address. This enables global connectivity for the AODV-node. In the route refresh sequence the AODV-node sends a RREQ to the default router's address from its configured global IPv6 address (step4). The default router's address is a global address configured on the ingress interface of the AR. When the AR and intermediate nodes receive the RREQ, they create an entry for the AODV-node's global address towards the next-hop towards the AODV-node (step5). This enables AODV-node's reachability from the Internet. Finally the AR sends a RREP in response to the RREQ (step6). The AODV-node has to execute the route refresh sequence in certain time periods in order to maintain reachability from the AR [11]. If the route cannot be refreshed i.e. the AODV-node loses P„„kk÷nen et al. Expires July 2004 [Page 13] INTERNET-DRAFT End-to-end connectivity December 2003 reachability to the AR, the AODV-node has to return to using the site-local MANET-address, because the AR might not be present anymore to provide default routing and the validity of the IPv6 prefix has expired. In [11] it has been proposed to implement default routing via the global address of the AR. In that kind of configuration the furthermost AODV-node of figure 6 would require three RT entries (::/0 -> AR ; AR -> next-hop towards AR ; next-hop -> ethx) for default routing. This isn't implementable on the Linux platform, because the next-hop router must have a RT entry towards the interface as described in figure 6. This is the reason default routing had to be provided via a next-hop as describes in this appendix. Full Copyright Statement Copyright (C) The Internet Society (2003). All Rights Reserved. This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into languages other than English. The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assigns. This document and the information contained herein is provided on an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.