RFC 4728
The Dynamic Source Routing Protocol (DSR) for Mobile Ad Hoc Networks for IPv4
Network Working Group D. Johnson Request for Comments: 4728 Rice University Category: Experimental Y. Hu UIUC D. Maltz Microsoft Research February 2007 The Dynamic Source Routing Protocol (DSR) for Mobile Ad Hoc Networks for IPv4 Status of This Memo This memo defines an Experimental Protocol for the Internet community. It does not specify an Internet standard of any kind. Discussion and suggestions for improvement are requested. Distribution of this memo is unlimited. Copyright Notice Copyright (C) The IETF Trust (2007).Abstract
The Dynamic Source Routing protocol (DSR) is a simple and efficient routing protocol designed specifically for use in multi-hop wireless ad hoc networks of mobile nodes. DSR allows the network to be completely self-organizing and self-configuring, without the need for any existing network infrastructure or administration. The protocol is composed of the two main mechanisms of "Route Discovery" and "Route Maintenance", which work together to allow nodes to discover and maintain routes to arbitrary destinations in the ad hoc network. All aspects of the protocol operate entirely on demand, allowing the routing packet overhead of DSR to scale automatically to only what is needed to react to changes in the routes currently in use. The protocol allows multiple routes to any destination and allows each sender to select and control the routes used in routing its packets, for example, for use in load balancing or for increased robustness. Other advantages of the DSR protocol include easily guaranteed loop- free routing, operation in networks containing unidirectional links, use of only "soft state" in routing, and very rapid recovery when routes in the network change. The DSR protocol is designed mainly for mobile ad hoc networks of up to about two hundred nodes and is designed to work well even with very high rates of mobility. This document specifies the operation of the DSR protocol for routing unicast IPv4 packets.
Table of Contents
1. Introduction ....................................................5 2. Assumptions .....................................................7 3. DSR Protocol Overview ...........................................9 3.1. Basic DSR Route Discovery .................................10 3.2. Basic DSR Route Maintenance ...............................12 3.3. Additional Route Discovery Features .......................14 3.3.1. Caching Overheard Routing Information ..............14 3.3.2. Replying to Route Requests Using Cached Routes .....15 3.3.3. Route Request Hop Limits ...........................16 3.4. Additional Route Maintenance Features .....................17 3.4.1. Packet Salvaging ...................................17 3.4.2. Queued Packets Destined over a Broken Link .........18 3.4.3. Automatic Route Shortening .........................19 3.4.4. Increased Spreading of Route Error Messages ........20 3.5. Optional DSR Flow State Extension .........................20 3.5.1. Flow Establishment .................................21 3.5.2. Receiving and Forwarding Establishment Packets .....22 3.5.3. Sending Packets along Established Flows ............22 3.5.4. Receiving and Forwarding Packets Sent along Established Flows ..................................23 3.5.5. Processing Route Errors ............................24 3.5.6. Interaction with Automatic Route Shortening ........24 3.5.7. Loop Detection .....................................25 3.5.8. Acknowledgement Destination ........................25 3.5.9. Crash Recovery .....................................25 3.5.10. Rate Limiting .....................................25 3.5.11. Interaction with Packet Salvaging .................26 4. Conceptual Data Structures .....................................26 4.1. Route Cache ...............................................26 4.2. Send Buffer ...............................................30 4.3. Route Request Table .......................................30 4.4. Gratuitous Route Reply Table ..............................31 4.5. Network Interface Queue and Maintenance Buffer ............32 4.6. Blacklist .................................................33 5. Additional Conceptual Data Structures for Flow State Extension ......................................................34 5.1. Flow Table ................................................34 5.2. Automatic Route Shortening Table ..........................35 5.3. Default Flow ID Table .....................................36 6. DSR Options Header Format ......................................36 6.1. Fixed Portion of DSR Options Header .......................37 6.2. Route Request Option ......................................40 6.3. Route Reply Option ........................................42
6.4. Route Error Option ........................................44
6.4.1. Node Unreachable Type-Specific Information .........46
6.4.2. Flow State Not Supported Type-Specific
Information ........................................46
6.4.3. Option Not Supported Type-Specific Information .....46
6.5. Acknowledgement Request Option ............................46
6.6. Acknowledgement Option ....................................47
6.7. DSR Source Route Option ...................................48
6.8. Pad1 Option ...............................................50
6.9. PadN Option ...............................................50
7. Additional Header Formats and Options for Flow State
Extension ......................................................51
7.1. DSR Flow State Header .....................................52
7.2. New Options and Extensions in DSR Options Header ..........52
7.2.1. Timeout Option .....................................52
7.2.2. Destination and Flow ID Option .....................53
7.3. New Error Types for Route Error Option ....................54
7.3.1. Unknown Flow Type-Specific Information .............54
7.3.2. Default Flow Unknown Type-Specific Information .....55
7.4. New Acknowledgement Request Option Extension ..............55
7.4.1. Previous Hop Address Extension .....................55
8. Detailed Operation .............................................56
8.1. General Packet Processing .................................56
8.1.1. Originating a Packet ...............................56
8.1.2. Adding a DSR Options Header to a Packet ............57
8.1.3. Adding a DSR Source Route Option to a Packet .......57
8.1.4. Processing a Received Packet .......................58
8.1.5. Processing a Received DSR Source Route Option ......60
8.1.6. Handling an Unknown DSR Option .....................63
8.2. Route Discovery Processing ................................64
8.2.1. Originating a Route Request ........................65
8.2.2. Processing a Received Route Request Option .........66
8.2.3. Generating a Route Reply Using the Route Cache .....68
8.2.4. Originating a Route Reply ..........................71
8.2.5. Preventing Route Reply Storms ......................72
8.2.6. Processing a Received Route Reply Option ...........74
8.3. Route Maintenance Processing ..............................74
8.3.1. Using Link-Layer Acknowledgements ..................75
8.3.2. Using Passive Acknowledgements .....................76
8.3.3. Using Network-Layer Acknowledgements ...............77
8.3.4. Originating a Route Error ..........................80
8.3.5. Processing a Received Route Error Option ...........81
8.3.6. Salvaging a Packet .................................82
8.4. Multiple Network Interface Support ........................84
8.5. IP Fragmentation and Reassembly ...........................84
8.6. Flow State Processing .....................................85
8.6.1. Originating a Packet ...............................85
8.6.2. Inserting a DSR Flow State Header ..................88
8.6.3. Receiving a Packet .................................88
8.6.4. Forwarding a Packet Using Flow IDs .................93
8.6.5. Promiscuously Receiving a Packet ...................93
8.6.6. Operation Where the Layer below DSR
Decreases the IP TTL ...............................94
8.6.7. Salvage Interactions with DSR ......................94
9. Protocol Constants and Configuration Variables .................95
10. IANA Considerations ...........................................96
11. Security Considerations .......................................96
Appendix A. Link-MaxLife Cache Description ........................97
Appendix B. Location of DSR in the ISO Network Reference Model ....99
Appendix C. Implementation and Evaluation Status .................100
Acknowledgements .................................................101
Normative References .............................................102
Informative References ...........................................102
1. Introduction
The Dynamic Source Routing protocol (DSR) [JOHNSON94, JOHNSON96a] is a simple and efficient routing protocol designed specifically for use in multi-hop wireless ad hoc networks of mobile nodes. Using DSR, the network is completely self-organizing and self-configuring, requiring no existing network infrastructure or administration. Network nodes cooperate to forward packets for each other to allow communication over multiple "hops" between nodes not directly within wireless transmission range of one another. As nodes in the network move about or join or leave the network, and as wireless transmission conditions such as sources of interference change, all routing is automatically determined and maintained by the DSR routing protocol. Since the number or sequence of intermediate hops needed to reach any destination may change at any time, the resulting network topology may be quite rich and rapidly changing. In designing DSR, we sought to create a routing protocol that had very low overhead yet was able to react very quickly to changes in the network. The DSR protocol provides highly reactive service in order to help ensure successful delivery of data packets in spite of node movement or other changes in network conditions. The DSR protocol is composed of two main mechanisms that work together to allow the discovery and maintenance of source routes in the ad hoc network: - Route Discovery is the mechanism by which a node S wishing to send a packet to a destination node D obtains a source route to D. Route Discovery is used only when S attempts to send a packet to D and does not already know a route to D. - Route Maintenance is the mechanism by which node S is able to detect, while using a source route to D, if the network topology has changed such that it can no longer use its route to D because a link along the route no longer works. When Route Maintenance indicates a source route is broken, S can attempt to use any other route it happens to know to D, or it can invoke Route Discovery again to find a new route for subsequent packets to D. Route Maintenance for this route is used only when S is actually sending packets to D. In DSR, Route Discovery and Route Maintenance each operate entirely "on demand". In particular, unlike other protocols, DSR requires no periodic packets of any kind at any layer within the network. For example, DSR does not use any periodic routing advertisement, link status sensing, or neighbor detection packets and does not rely on these functions from any underlying protocols in the network. This
entirely on-demand behavior and lack of periodic activity allows the number of overhead packets caused by DSR to scale all the way down to zero, when all nodes are approximately stationary with respect to each other and all routes needed for current communication have already been discovered. As nodes begin to move more or as communication patterns change, the routing packet overhead of DSR automatically scales to only what is needed to track the routes currently in use. Network topology changes not affecting routes currently in use are ignored and do not cause reaction from the protocol. All state maintained by DSR is "soft state" [CLARK88], in that the loss of any state will not interfere with the correct operation of the protocol; all state is discovered as needed and can easily and quickly be rediscovered if needed after a failure without significant impact on the protocol. This use of only soft state allows the routing protocol to be very robust to problems such as dropped or delayed routing packets or node failures. In particular, a node in DSR that fails and reboots can easily rejoin the network immediately after rebooting; if the failed node was involved in forwarding packets for other nodes as an intermediate hop along one or more routes, it can also resume this forwarding quickly after rebooting, with no or minimal interruption to the routing protocol. In response to a single Route Discovery (as well as through routing information from other packets overheard), a node may learn and cache multiple routes to any destination. This support for multiple routes allows the reaction to routing changes to be much more rapid, since a node with multiple routes to a destination can try another cached route if the one it has been using should fail. This caching of multiple routes also avoids the overhead of needing to perform a new Route Discovery each time a route in use breaks. The sender of a packet selects and controls the route used for its own packets, which, together with support for multiple routes, also allows features such as load balancing to be defined. In addition, all routes used are easily guaranteed to be loop-free, since the sender can avoid duplicate hops in the routes selected. The operation of both Route Discovery and Route Maintenance in DSR are designed to allow unidirectional links and asymmetric routes to be supported. In particular, as noted in Section 2, in wireless networks, it is possible that a link between two nodes may not work equally well in both directions, due to differing transmit power levels or sources of interference. It is possible to interface a DSR network with other networks, external to this DSR network. Such external networks may, for example, be the Internet or may be other ad hoc networks routed with
a routing protocol other than DSR. Such external networks may also be other DSR networks that are treated as external networks in order to improve scalability. The complete handling of such external networks is beyond the scope of this document. However, this document specifies a minimal set of requirements and features necessary to allow nodes only implementing this specification to interoperate correctly with nodes implementing interfaces to such external networks. This document specifies the operation of the DSR protocol for routing unicast IPv4 packets in multi-hop wireless ad hoc networks. Advanced, optional features, such as Quality of Service (QoS) support and efficient multicast routing, and operation of DSR with IPv6 [RFC2460], will be covered in other documents. The specification of DSR in this document provides a compatible base on which such features can be added, either independently or by integration with the DSR operation specified here. As described in Appendix C, the design of DSR has been extensively studied through detailed simulations and testbed implementation and demonstration; this document encourages additional implementation and experimentation with the protocol. 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 [RFC2119].2. Assumptions
As described here, the DSR protocol is designed mainly for mobile ad hoc networks of up to about two hundred nodes and is designed to work well even with very high rates of mobility. Other protocol features and enhancements that may allow DSR to scale to larger networks are outside the scope of this document. We assume in this document that all nodes wishing to communicate with other nodes within the ad hoc network are willing to participate fully in the protocols of the network. In particular, each node participating in the ad hoc network SHOULD also be willing to forward packets for other nodes in the network. The diameter of an ad hoc network is the minimum number of hops necessary for a packet to reach from any node located at one extreme edge of the ad hoc network to another node located at the opposite extreme. We assume that this diameter will often be small (e.g., perhaps 5 or 10 hops), but it may often be greater than 1.
Packets may be lost or corrupted in transmission on the wireless network. We assume that a node receiving a corrupted packet can detect the error, such as through a standard link-layer checksum or Cyclic Redundancy Check (CRC), and discard the packet. Nodes within the ad hoc network MAY move at any time without notice and MAY even move continuously, but we assume that the speed with which nodes move is moderate with respect to the packet transmission latency and wireless transmission range of the particular underlying network hardware in use. In particular, DSR can support very rapid rates of arbitrary node mobility, but we assume that nodes do not continuously move so rapidly as to make the flooding of every individual data packet the only possible routing protocol. A common feature of many network interfaces, including most current LAN hardware for broadcast media such as wireless, is the ability to operate the network interface in "promiscuous" receive mode. This mode causes the hardware to deliver every received packet to the network driver software without filtering based on link-layer destination address. Although we do not require this facility, some of our optimizations can take advantage of its availability. Use of promiscuous mode does increase the software overhead on the CPU, but we believe that wireless network speeds and capacity are more the inherent limiting factors to performance in current and future systems; we also believe that portions of the protocol are suitable for implementation directly within a programmable network interface unit to avoid this overhead on the CPU [JOHNSON96a]. Use of promiscuous mode may also increase the power consumption of the network interface hardware, depending on the design of the receiver hardware, and in such cases, DSR can easily be used without the optimizations that depend on promiscuous receive mode or can be programmed to only periodically switch the interface into promiscuous mode. Use of promiscuous receive mode is entirely optional. Wireless communication ability between any pair of nodes may at times not work equally well in both directions, due, for example, to transmit power levels or sources of interference around the two nodes [BANTZ94, LAUER95]. That is, wireless communications between each pair of nodes will in many cases be able to operate bidirectionally, but at times the wireless link between two nodes may be only unidirectional, allowing one node to successfully send packets to the other while no communication is possible in the reverse direction. Some Medium Access Control (MAC) protocols, however, such as MACA [KARN90], MACAW [BHARGHAVAN94], or IEEE 802.11 [IEEE80211], limit unicast data packet transmission to bidirectional links, due to the required bidirectional exchange of request to send (RTS) and clear to send (CTS) packets in these protocols and to the link-layer acknowledgement feature in IEEE 802.11. When used on top of MAC
protocols such as these, DSR can take advantage of additional optimizations, such as the ability to reverse a source route to obtain a route back to the origin of the original route. The IP address used by a node using the DSR protocol MAY be assigned by any mechanism (e.g., static assignment or use of Dynamic Host Configuration Protocol (DHCP) for dynamic assignment [RFC2131]), although the method of such assignment is outside the scope of this specification. A routing protocol such as DSR chooses a next-hop for each packet and provides the IP address of that next-hop. When the packet is transmitted, however, the lower-layer protocol often has a separate, MAC-layer address for the next-hop node. DSR uses the Address Resolution Protocol (ARP) [RFC826] to translate from next-hop IP addresses to next-hop MAC addresses. In addition, a node MAY add an entry to its ARP cache based on any received packet, when the IP address and MAC address of the transmitting node are available in the packet; for example, the IP address of the transmitting node is present in a Route Request option (in the Address list being accumulated) and any packets containing a source route. Adding entries to the ARP cache in this way avoids the overhead of ARP in most cases.3. DSR Protocol Overview
This section provides an overview of the operation of the DSR protocol. The basic version of DSR uses explicit "source routing", in which each data packet sent carries in its header the complete, ordered list of nodes through which the packet will pass. This use of explicit source routing allows the sender to select and control the routes used for its own packets, supports the use of multiple routes to any destination (for example, for load balancing), and allows a simple guarantee that the routes used are loop-free. By including this source route in the header of each data packet, other nodes forwarding or overhearing any of these packets can also easily cache this routing information for future use. Section 3.1 describes this basic operation of Route Discovery, Section 3.2 describes basic Route Maintenance, and Sections 3.3 and 3.4 describe additional features of these two parts of DSR's operation. Section 3.5 then describes an optional, compatible extension to DSR, known as "flow state", that allows the routing of most packets without an explicit source route header in the packet, while the fundamental properties of DSR's operation are preserved.
3.1. Basic DSR Route Discovery
When some source node originates a new packet addressed to some destination node, the source node places in the header of the packet a "source route" giving the sequence of hops that the packet is to follow on its way to the destination. Normally, the sender will obtain a suitable source route by searching its "Route Cache" of routes previously learned; if no route is found in its cache, it will initiate the Route Discovery protocol to dynamically find a new route to this destination node. In this case, we call the source node the "initiator" and the destination node the "target" of the Route Discovery. For example, suppose a node A is attempting to discover a route to node E. The Route Discovery initiated by node A in this example would proceed as follows: ^ "A" ^ "A,B" ^ "A,B,C" ^ "A,B,C,D" | id=2 | id=2 | id=2 | id=2 +-----+ +-----+ +-----+ +-----+ +-----+ | A |---->| B |---->| C |---->| D |---->| E | +-----+ +-----+ +-----+ +-----+ +-----+ | | | | v v v v To initiate the Route Discovery, node A transmits a "Route Request" as a single local broadcast packet, which is received by (approximately) all nodes currently within wireless transmission range of A, including node B in this example. Each Route Request identifies the initiator and target of the Route Discovery, and also contains a unique request identification (2, in this example), determined by the initiator of the Request. Each Route Request also contains a record listing the address of each intermediate node through which this particular copy of the Route Request has been forwarded. This route record is initialized to an empty list by the initiator of the Route Discovery. In this example, the route record initially lists only node A. When another node receives this Route Request (such as node B in this example), if it is the target of the Route Discovery, it returns a "Route Reply" to the initiator of the Route Discovery, giving a copy of the accumulated route record from the Route Request; when the initiator receives this Route Reply, it caches this route in its Route Cache for use in sending subsequent packets to this destination.
Otherwise, if this node receiving the Route Request has recently seen another Route Request message from this initiator bearing this same request identification and target address, or if this node's own address is already listed in the route record in the Route Request, this node discards the Request. (A node considers a Request recently seen if it still has information about that Request in its Route Request Table, which is described in Section 4.3.) Otherwise, this node appends its own address to the route record in the Route Request and propagates it by transmitting it as a local broadcast packet (with the same request identification). In this example, node B broadcast the Route Request, which is received by node C; nodes C and D each also, in turn, broadcast the Request, resulting in receipt of a copy of the Request by node E. In returning the Route Reply to the initiator of the Route Discovery, such as in this example, node E replying back to node A, node E will typically examine its own Route Cache for a route back to A and, if one is found, will use it for the source route for delivery of the packet containing the Route Reply. Otherwise, E SHOULD perform its own Route Discovery for target node A, but to avoid possible infinite recursion of Route Discoveries, it MUST in this case piggyback this Route Reply on the packet containing its own Route Request for A. It is also possible to piggyback other small data packets, such as a TCP SYN packet [RFC793], on a Route Request using this same mechanism. Node E could instead simply reverse the sequence of hops in the route record that it is trying to send in the Route Reply and use this as the source route on the packet carrying the Route Reply itself. For MAC protocols, such as IEEE 802.11, that require a bidirectional frame exchange for unicast packets as part of the MAC protocol [IEEE80211], the discovered source route MUST be reversed in this way to return the Route Reply, since this route reversal tests the discovered route to ensure that it is bidirectional before the Route Discovery initiator begins using the route. This route reversal also avoids the overhead of a possible second Route Discovery. When initiating a Route Discovery, the sending node saves a copy of the original packet (that triggered the discovery) in a local buffer called the "Send Buffer". The Send Buffer contains a copy of each packet that cannot be transmitted by this node because it does not yet have a source route to the packet's destination. Each packet in the Send Buffer is logically associated with the time that it was placed into the Send Buffer and is discarded after residing in the Send Buffer for some timeout period SendBufferTimeout; if necessary for preventing the Send Buffer from overflowing, a FIFO or other replacement strategy MAY also be used to evict packets even before they expire.
While a packet remains in the Send Buffer, the node SHOULD occasionally initiate a new Route Discovery for the packet's destination address. However, the node MUST limit the rate at which such new Route Discoveries for the same address are initiated (as described in Section 4.3), since it is possible that the destination node is not currently reachable. In particular, due to the limited wireless transmission range and the movement of the nodes in the network, the network may at times become partitioned, meaning that there is currently no sequence of nodes through which a packet could be forwarded to reach the destination. Depending on the movement pattern and the density of nodes in the network, such network partitions may be rare or common. If a new Route Discovery was initiated for each packet sent by a node in such a partitioned network, a large number of unproductive Route Request packets would be propagated throughout the subset of the ad hoc network reachable from this node. In order to reduce the overhead from such Route Discoveries, a node SHOULD use an exponential back-off algorithm to limit the rate at which it initiates new Route Discoveries for the same target, doubling the timeout between each successive discovery initiated for the same target. If the node attempts to send additional data packets to this same destination node more frequently than this limit, the subsequent packets SHOULD be buffered in the Send Buffer until a Route Reply is received giving a route to this destination, but the node MUST NOT initiate a new Route Discovery until the minimum allowable interval between new Route Discoveries for this target has been reached. This limitation on the maximum rate of Route Discoveries for the same target is similar to the mechanism required by Internet nodes to limit the rate at which ARP Requests are sent for any single target IP address [RFC1122].3.2. Basic DSR Route Maintenance
When originating or forwarding a packet using a source route, each node transmitting the packet is responsible for confirming that data can flow over the link from that node to the next hop. For example, in the situation shown below, node A has originated a packet for node E using a source route through intermediate nodes B, C, and D: +-----+ +-----+ +-----+ +-----+ +-----+ | A |---->| B |---->| C |-->? | D | | E | +-----+ +-----+ +-----+ +-----+ +-----+ In this case, node A is responsible for the link from A to B, node B is responsible for the link from B to C, node C is responsible for the link from C to D, and node D is responsible for the link from D to E.
An acknowledgement can provide confirmation that a link is capable of carrying data, and in wireless networks, acknowledgements are often provided at no cost, either as an existing standard part of the MAC protocol in use (such as the link-layer acknowledgement frame defined by IEEE 802.11 [IEEE80211]), or by a "passive acknowledgement" [JUBIN87] (in which, for example, B confirms receipt at C by overhearing C transmit the packet when forwarding it on to D). If a built-in acknowledgement mechanism is not available, the node transmitting the packet can explicitly request that a DSR-specific software acknowledgement be returned by the next node along the route; this software acknowledgement will normally be transmitted directly to the sending node, but if the link between these two nodes is unidirectional (Section 4.6), this software acknowledgement could travel over a different, multi-hop path. After an acknowledgement has been received from some neighbor, a node MAY choose not to require acknowledgements from that neighbor for a brief period of time, unless the network interface connecting a node to that neighbor always receives an acknowledgement in response to unicast traffic. When a software acknowledgement is used, the acknowledgement request SHOULD be retransmitted up to a maximum number of times. A retransmission of the acknowledgement request can be sent as a separate packet, piggybacked on a retransmission of the original data packet, or piggybacked on any packet with the same next-hop destination that does not also contain a software acknowledgement. After the acknowledgement request has been retransmitted the maximum number of times, if no acknowledgement has been received, then the sender treats the link to this next-hop destination as currently "broken". It SHOULD remove this link from its Route Cache and SHOULD return a "Route Error" to each node that has sent a packet routed over that link since an acknowledgement was last received. For example, in the situation shown above, if C does not receive an acknowledgement from D after some number of requests, it would return a Route Error to A, as well as any other node that may have used the link from C to D since C last received an acknowledgement from D. Node A then removes this broken link from its cache; any retransmission of the original packet can be performed by upper layer protocols such as TCP, if necessary. For sending such a retransmission or other packets to this same destination E, if A has in its Route Cache another route to E (for example, from additional Route Replies from its earlier Route Discovery, or from having overheard sufficient routing information from other packets), it can
send the packet using the new route immediately. Otherwise, it SHOULD perform a new Route Discovery for this target (subject to the back-off described in Section 3.1).3.3. Additional Route Discovery Features
3.3.1. Caching Overheard Routing Information
A node forwarding or otherwise overhearing any packet SHOULD add all usable routing information from that packet to its own Route Cache. The usefulness of routing information in a packet depends on the directionality characteristics of the physical medium (Section 2), as well as on the MAC protocol being used. Specifically, three distinct cases are possible: - Links in the network frequently are capable of operating only unidirectionally (not bidirectionally), and the MAC protocol in use in the network is capable of transmitting unicast packets over unidirectional links. - Links in the network occasionally are capable of operating only unidirectionally (not bidirectionally), but this unidirectional restriction on any link is not persistent; almost all links are physically bidirectional, and the MAC protocol in use in the network is capable of transmitting unicast packets over unidirectional links. - The MAC protocol in use in the network is not capable of transmitting unicast packets over unidirectional links; only bidirectional links can be used by the MAC protocol for transmitting unicast packets. For example, the IEEE 802.11 Distributed Coordination Function (DCF) MAC protocol [IEEE80211] is capable of transmitting a unicast packet only over a bidirectional link, since the MAC protocol requires the return of a link-level acknowledgement packet from the receiver and also optionally requires the bidirectional exchange of an RTS and CTS packet between the transmitter and receiver nodes. In the first case above, for example, the source route used in a data packet, the accumulated route record in a Route Request, or the route being returned in a Route Reply SHOULD all be cached by any node in the "forward" direction. Any node SHOULD cache this information from any such packet received, whether the packet was addressed to this node, sent to a broadcast (or multicast) MAC address, or overheard while the node's network interface is in promiscuous mode. However, the "reverse" direction of the links identified in such packet headers SHOULD NOT be cached.
For example, in the situation shown below, node A is using a source
route to communicate with node E:
+-----+ +-----+ +-----+ +-----+ +-----+
| A |---->| B |---->| C |---->| D |---->| E |
+-----+ +-----+ +-----+ +-----+ +-----+
As node C forwards a data packet along the route from A to E, it
SHOULD add to its cache the presence of the "forward" direction links
that it learns from the headers of these packets, from itself to D
and from D to E. Node C SHOULD NOT, in this case, cache the
"reverse" direction of the links identified in these packet headers,
from itself back to B and from B to A, since these links might be
unidirectional.
In the second case above, in which links may occasionally operate
unidirectionally, the links described above SHOULD be cached in both
directions. Furthermore, in this case, if node X overhears (e.g.,
through promiscuous mode) a packet transmitted by node C that is
using a source route from node A to E, node X SHOULD cache all of
these links as well, also including the link from C to X over which
it overheard the packet.
In the final case, in which the MAC protocol requires physical
bidirectionality for unicast operation, links from a source route
SHOULD be cached in both directions, except when the packet also
contains a Route Reply, in which case only the links already
traversed in this source route SHOULD be cached. However, the links
not yet traversed in this route SHOULD NOT be cached.
3.3.2. Replying to Route Requests Using Cached Routes
A node receiving a Route Request for which it is not the target
searches its own Route Cache for a route to the target of the
Request. If it is found, the node generally returns a Route Reply to
the initiator itself rather than forward the Route Request. In the
Route Reply, this node sets the route record to list the sequence of
hops over which this copy of the Route Request was forwarded to it,
concatenated with the source route to this target obtained from its
own Route Cache.
However, before transmitting a Route Reply packet that was generated
using information from its Route Cache in this way, a node MUST
verify that the resulting route being returned in the Route Reply,
after this concatenation, contains no duplicate nodes listed in the
route record. For example, the figure below illustrates a case in
which a Route Request for target E has been received by node F, and
node F already has in its Route Cache a route from itself to E:
+-----+ +-----+ +-----+ +-----+
| A |---->| B |- >| D |---->| E |
+-----+ +-----+ \ / +-----+ +-----+
\ /
\ +-----+ /
>| C |-
+-----+
| ^
v |
Route Request +-----+
Route: A - B - C - F | F | Cache: C - D - E
+-----+
The concatenation of the accumulated route record from the Route
Request and the cached route from F's Route Cache would include a
duplicate node in passing from C to F and back to C.
Node F in this case could attempt to edit the route to eliminate the
duplication, resulting in a route from A to B to C to D and on to E,
but in this case, node F would not be on the route that it returned
in its own Route Reply. DSR Route Discovery prohibits node F from
returning such a Route Reply from its cache; this prohibition
increases the probability that the resulting route is valid, since
node F in this case should have received a Route Error if the route
had previously stopped working. Furthermore, this prohibition means
that a future Route Error traversing the route is very likely to pass
through any node that sent the Route Reply for the route (including
node F), which helps to ensure that stale data is removed from caches
(such as at F) in a timely manner; otherwise, the next Route
Discovery initiated by A might also be contaminated by a Route Reply
from F containing the same stale route. If, due to this restriction
on returning a Route Reply based on information from its Route Cache,
node F does not return such a Route Reply, it propagates the Route
Request normally.
3.3.3. Route Request Hop Limits
Each Route Request message contains a "hop limit" that may be used to
limit the number of intermediate nodes allowed to forward that copy
of the Route Request. This hop limit is implemented using the Time-
to-Live (TTL) field in the IP header of the packet carrying the Route
Request. As the Request is forwarded, this limit is decremented, and
the Request packet is discarded if the limit reaches zero before
finding the target. This Route Request hop limit can be used to
implement a variety of algorithms for controlling the spread of a
Route Request during a Route Discovery attempt.
For example, a node MAY use this hop limit to implement a "non- propagating" Route Request as an initial phase of a Route Discovery. A node using this technique sends its first Route Request attempt for some target node using a hop limit of 1, such that any node receiving the initial transmission of the Route Request will not forward the Request to other nodes by re-broadcasting it. This form of Route Request is called a "non-propagating" Route Request; it provides an inexpensive method for determining if the target is currently a neighbor of the initiator or if a neighbor node has a route to the target cached (effectively using the neighbors' Route Caches as an extension of the initiator's own Route Cache). If no Route Reply is received after a short timeout, then the node sends a "propagating" Route Request for the target node (i.e., with hop limit as defined by the value of the DiscoveryHopLimit configuration variable). As another example, a node MAY use this hop limit to implement an "expanding ring" search for the target [JOHNSON96a]. A node using this technique sends an initial non-propagating Route Request as described above; if no Route Reply is received for it, the node originates another Route Request with a hop limit of 2. For each Route Request originated, if no Route Reply is received for it, the node doubles the hop limit used on the previous attempt, to progressively explore for the target node without allowing the Route Request to propagate over the entire network. However, this expanding ring search approach could increase the average latency of Route Discovery, since multiple Discovery attempts and timeouts may be needed before discovering a route to the target node.3.4. Additional Route Maintenance Features
3.4.1. Packet Salvaging
When an intermediate node forwarding a packet detects through Route Maintenance that the next hop along the route for that packet is broken, if the node has another route to the packet's destination in its Route Cache, the node SHOULD "salvage" the packet rather than discard it. To salvage a packet, the node replaces the original source route on the packet with a route from its Route Cache. The node then forwards the packet to the next node indicated along this source route. For example, in the situation shown in the example of Section 3.2, if node C has another route cached to node E, it can salvage the packet by replacing the original route in the packet with this new route from its own Route Cache rather than discarding the packet. When salvaging a packet, a count is maintained in the packet of the number of times that it has been salvaged, to prevent a single packet from being salvaged endlessly. Otherwise, since the TTL is
decremented only once by each node, a single node could salvage a packet an unbounded number of times. Even if we chose to require the TTL to be decremented on each salvage attempt, packet salvaging is an expensive operation, so it is desirable to bound the maximum number of times a packet can be salvaged independently of the maximum number of hops a packet can traverse. As described in Section 3.2, an intermediate node, such as in this case, that detects through Route Maintenance that the next hop along the route for a packet that it is forwarding is broken, the node also SHOULD return a Route Error to the original sender of the packet, identifying the link over which the packet could not be forwarded. If the node sends this Route Error, it SHOULD originate the Route Error before salvaging the packet.3.4.2. Queued Packets Destined over a Broken Link
When an intermediate node forwarding a packet detects through Route Maintenance that the next-hop link along the route for that packet is broken, in addition to handling that packet as defined for Route Maintenance, the node SHOULD also handle in a similar way any pending packets that it has queued that are destined over this new broken link. Specifically, the node SHOULD search its Network Interface Queue and Maintenance Buffer (Section 4.5) for packets for which the next-hop link is this new broken link. For each such packet currently queued at this node, the node SHOULD process that packet as follows: - Remove the packet from the node's Network Interface Queue and Maintenance Buffer. - Originate a Route Error for this packet to the original sender of the packet, using the procedure described in Section 8.3.4, as if the node had already reached the maximum number of retransmission attempts for that packet for Route Maintenance. However, in sending such Route Errors for queued packets in response to detection of a single, new broken link, the node SHOULD send no more than one Route Error to each original sender of any of these packets. - If the node has another route to the packet's IP Destination Address in its Route Cache, the node SHOULD salvage the packet as described in Section 8.3.6. Otherwise, the node SHOULD discard the packet.
3.4.3. Automatic Route Shortening
Source routes in use MAY be automatically shortened if one or more intermediate nodes in the route become no longer necessary. This mechanism of automatically shortening routes in use is somewhat similar to the use of passive acknowledgements [JUBIN87]. In particular, if a node is able to overhear a packet carrying a source route (e.g., by operating its network interface in promiscuous receive mode), then this node examines the unexpended portion of that source route. If this node is not the intended next-hop destination for the packet but is named in the later unexpended portion of the packet's source route, then it can infer that the intermediate nodes before itself in the source route are no longer needed in the route. For example, the figure below illustrates an example in which node D has overheard a data packet being transmitted from B to C, for later forwarding to D and to E: +-----+ +-----+ +-----+ +-----+ +-----+ | A |---->| B |---->| C | | D | | E | +-----+ +-----+ +-----+ +-----+ +-----+ \ ^ \ / --------------------- In this case, this node (node D) SHOULD return a "gratuitous" Route Reply to the original sender of the packet (node A). The Route Reply gives the shorter route as the concatenation of the portion of the original source route up through the node that transmitted the overheard packet (node B), plus the suffix of the original source route beginning with the node returning the gratuitous Route Reply (node D). In this example, the route returned in the gratuitous Route Reply message sent from D to A gives the new route as the sequence of hops from A to B to D to E. When deciding whether to return a gratuitous Route Reply in this way, a node MAY factor in additional information beyond the fact that it was able to overhear the packet. For example, the node MAY decide to return the gratuitous Route Reply only when the overheard packet is received with a signal strength or signal-to-noise ratio above some specific threshold. In addition, each node maintains a Gratuitous Route Reply Table, as described in Section 4.4, to limit the rate at which it originates gratuitous Route Replies for the same returned route.
3.4.4. Increased Spreading of Route Error Messages
When a source node receives a Route Error for a data packet that it originated, this source node propagates this Route Error to its neighbors by piggybacking it on its next Route Request. In this way, stale information in the caches of nodes around this source node will not generate Route Replies that contain the same invalid link for which this source node received the Route Error. For example, in the situation shown in the example of Section 3.2, node A learns from the Route Error message from C that the link from C to D is currently broken. It thus removes this link from its own Route Cache and initiates a new Route Discovery (if it has no other route to E in its Route Cache). On the Route Request packet initiating this Route Discovery, node A piggybacks a copy of this Route Error, ensuring that the Route Error spreads well to other nodes, and guaranteeing that any Route Reply that it receives (including those from other node's Route Caches) in response to this Route Request does not contain a route that assumes the existence of this broken link.
(page 20 continued on part 2)
