AU2011100349B4 - Traceback packet transport protocol - Google Patents

Abstract

Tokens identifying all of the physical routing devices, i.e., network nodes, through which a packet travels are recorded in a limited amount of space reserved in the header of the packet for such tokens. When insufficient space remains in the header of the packet for all tokens required to identify 5 all physical routing devices through which the packet travels, sequences of multiple tokens are replaced with an abbreviation token representing the sequence. The sequence of tokens represented by an abbreviation token can also be abbreviation tokens, supporting recursive abbreviation of the token sequence in the header of the packet as needed to record the entire route of the packet through the network regardless of the limited space in the header for tracking the route of the packet.

Description

Regulation 3.2 AUSTRALIA PATENTS ACT 1990 COMPLETE SPECIFICATION FOR AN INNOVATION PATENT ORIGINAL Name of Applicant: Uniloc USA, Inc. Actual Inventors: Craig S Etchegoyen Dono Harjanto Address for Service: C/- MADDERNS, GPO Box 2752, Adelaide, South Australia, Australia Invention title: TRACEBACK PACKET TRANSPORT PROTOCOL The following statement is a full description of this invention, including the best method of performing it known to us.

2 BACKGROUND OF THE INVENTION FIELD OF THE INVENTION The present invention relates generally to computer networking and, more particularly, to 5 methods of and systems for transporting packets through a network while supporting packet traceback. Description of the Related Art It is advantageous to trace a particular route by which a packet is transported through a net work. However, packets that are transported through networks have fixed lengths while the number of 10 hops each packet can take through a network vary widely. Allocating insufficient space to record the route of a packet within the packet defeats proper tracing of the route. Often, there are no limits on the number of hops a packet can take through a network and so there is no amount of space that can be re served in a packet to guarantee accuracy for route tracing. Even in situations in which the number of hops a packet may take are limited, allocating space to record the maximum packet route in each packet 15 will waste precious bandwidth for all packets taking less than the maximum packet route. SUMMARY OF THE INVENTION In accordance with the present invention, tokens identifying all of the physical routing devices, ie, network nodes, through which a packet travels are recorded in a limited amount of space reserved in 20 the header of the packet for such tokens. When insufficient space remains in the header of the packet for all tokens required to identify all physical routing devices through which the packet travels, sequences of multiple tokens are replaced with a single token representing the sequence. The single token is sometimes referred to herein as an abbreviation token. The sequence of 25 tokens represented by an abbreviation token can also be composed of abbreviation tokens, supporting recursive abbreviation of the token sequence in the header of the packet as needed to record the entire route of the packet through the network regardless of the limited space in the header for tracking the route of the packet. 30 To identify the physical nodes through which the packet travels, the tokens are derived from hardware features of each node, much the way a digital fingerprint is derived. Accordingly, identification of the physical nodes through which the packet travels cannot be defeated by spoofing easily reconfigurable attributes such as network addresses.

3 Various nodes of the network can learn the tokens of adjacent nodes of the network through interior gateway routing protocols such as RIP packets or Hello packets found in OSPF or similar protocols. Nodes can also encrypt the packet for secure hops using the token of the next node as an encryption key. 5 The abbreviation tokens can be produced in a manner that is consistent throughout all nodes of the network such that expansion of abbreviation tokens to reconstruct the route of the packet can be achieved by any device that knows the abbreviation token generation method. 0 The abbreviation tokens can also be generated by each node using its own particular method. In such cases, the node that generates an abbreviation token ensures that its own token immediately follows the abbreviation token to thereby identify itself as the node that can properly expand the abbreviation token. To reconstruct the route of a packet through the network, the node generating each abbreviation token is identified and asked to expand the abbreviation token. 5 Specifically, a first aspect of the present invention accordingly provides a method for routing a packet through a network from a source to a destination, the method comprising: storing data in a header of the packet to represent a complete route of the packet through the network, the data identifying at least one physical routing device of the network through which the packet travels. 0 In another form, the method further comprises replacing data identifying at least two physical routing devices with a single token in the header of the packet such that the single token identifies the at least two physical routing devices. In another form, the method further comprises storing data in the header of the packet that 5 identifies a particular physical routing device that performs the replacing. In another form, the storing is performed by node logic executing within each physical routing device that receives the packet along the rate. 30 In a second aspect, the present invention accordingly provides a method for identifying a particular route taken by a packet through a network, the method comprising: retrieving one or more tokens from a header of the packet, the tokens collectively identifying one or more physical routing devices through which the packet traveled; determining that at least a selected one of the tokens represents two or more other tokens; and 35 replacing the selected token with the two or more other tokens. BRIEF DESCRIPTION OF THE DRAWINGS 3a Other systems, methods, features and advantages of the invention will be or will become apparent to one with skill in the art upon examination of the following figures and detailed description. It is intended that all such additional systems, methods, features and advantages be included within this description, be within the scope of the invention, and be protected by the accompanying claims. 5 Component parts shown in the drawings are not necessarily to scale, and may be exaggerated to better illustrate the important features of the invention. In the drawings, like reference numerals may designate like parts throughout the different views, wherein: Figure 1 is a diagram showing computers, including a node manager, connected through a computer network that includes a number of nodes that transport packets in accordance with the present invention. 0 Figure 2 is a block diagram showing a node of Figure 1 in greater detail. Figure 3 is a logic flow diagram of the transport of a packet by the node of Figure 1 in accordance with the present invention. Figure 4 is a logic flow diagram showing a step of the logic flow diagram of Figure 3 in greater detail. Figure 5 is a logic flow diagram showing a step of the logic flow diagram of Figure 4 in greater detail. 5 Figure 6 is a block diagram of a token definition of the token database of Figure 2 in greater detail. Figure 7 is a block diagram of a packet that includes a number of token slots in its header in accordance with the present invention. Figures 8A-8E are block diagrams illustrating the recording of the route of WO 20091076232 PCT/US2008/085730 well-known structures and devices are shown in block diagram form in order to facilitate describing one or more embodiments. [0021] In accordance with one or more aspects of the embodiments described herein, there is provided a system and method for sending data via secured communication in a public key infrastructure. Figure 2 shows an example embodiment of how to bind the simple public key infrastructure of figure one to a specific computing device. The encryption system operates implements certain features of a simple public key infrastructure encryption system; however the requester's private key 22 is not stored locally but generated by producing a unique device ID 20 (also referred to herein as a unique device identifier) from the operating environment of the local computing device and combining it with a filler code 21 that together produce a comparable private key 22 that can, in turn, be used for decryption 27 of PKI encrypted data. The unique device ID is usually a string of data resulting from the compiling of unique identifying information from the devices computing environment. Examples of this are hard drive serial number and data storage damage locations in a computing device. Further examples of unique identifying information that be used to generate/compute the unique device ID are provided below. [0022] Figure 3 shows an exemplary way in which a recipient's private key can be generated by a sender. Initially a key pair may be generated (step 31) using standard PKI key generation known in the art. The two keys may then be set aside for use as a public key and a private key. The sender 29 may then request the requester 30 to send back to them a unique device ID 35 generated from the operating environment of the recipient's device. The sender 29 then compares the unique device ID with the private key 32 to be used by the requester 30 (step 32). The sender 29 may calculate the difference between the unique device ID and the private key 33 (step 33). This calculated difference may be stored as a filler code (step 34). [0023] Figure 4 illustrates an exemplary manner in which a device bound private key may be used in practice. The requester 37 asks for encrypted data to be sent to them and also publishes their public key 39 to be used as part of the data exchange (step 39). The sender 38 identifies the requester 37 (step 40), and encrypts the data 42 to be sent using their own private key 25 and the requester's public key 23 (step 42). Then (at step 44) the sender 38 sends to the requester 37 the encrypted data, the requester's filler code 21 that has been previously generated, and publishes their own public key 24. -4- 4 Figures 8A-8E are block diagrams illustrating the recording of the route of a packet, including the use of abbreviation tokens. Figure 9 is a logic flow diagram illustrating the reconstruction of the route taken by a packet through the network of Figure 1. 5 Figures 10A-IOE are block diagrams illustrating the reconstruction of the route recorded in Figures 8A 8E according to the logic flow diagram of Figure. 9. DETAILED DESCRIPTION In accordance with the present invention, nodes 108A-1 of network 106 transport packets in a 10 manner that records the full route of each packet through network 106 while requiring a relatively small portion of fixed size within each packet for the recording. The route identifies individual, specific ones of nodes 108A-I and not merely IP addresses or other easily configurable or modifiable characteristics of nodes 108A-I. 15 As described more completely below, each of nodes 108A-I has an identifying token that is unique among tokens of nodes within network 106. The identifier is derived from data specific to each of nodes 108A-1 such that the identifier identifies a specific node device. When a packet is routed to any of nodes 108A-1, the receiving node records its token within the packet to thereby record the receiving node as part of the route of the packet. When recorded tokens have filled the limited space of the packet 20 allocated for recording the route, a receiving node replaces multiple tokens in the recorded route with a single token that represents the sequence of replaced tokens, to thereby free space to record additional tokens of the packet's route. Before describing the recording of a packet's route through network 106 in accordance with 25 the present invention, some elements of node 108A (Figure 1) are briefly described. Nodes 108A-I are analogous to one another and the following description of node 108A is equally applicable to each of nodes 108B-1 except as noted herein. Node 108A is shown in greater detail in Figure 2 and includes one or more microprocessors 30 202 (collectively referred to as CPU 202) that retrieve data and/or instructions from memory 204 and execute retrieved instructions in a conventional manner. Memory 204 can include generally any com puter-readable medium including, for example, persistent memory such as magnetic and/or optical disks, ROM, and PROM and volatile memory such as RAM.

5 CPU 202 and memory 204 are connected to one another through a conventional interconnect 206, which is a bus in this illustrative embodiment and which connects CPU 202 and memory 204 to network access circuitry 208. Network access circuitry 208 sends and receives data through a network 106 (FIGURE 1) and includes ethernet circuitry or fiber optic circuitry in some embodiments. 5 Node 108A is shown without user input and output, ie, user-interface devices. While many nodes of a network do not have user-interface devices, some nodes are computers intended to be used by a person and therefore do include user-interface devices. 10 A number of components of node 108A are stored in memory 204. In particular, routing logic 210 is all or part of one or more computer processes executing within CPU 202 from memory 204 in this illustrative embodiment but can also be implemented using digital logic circuitry. As used herein, "logic" refers to (i) logic implemented as computer instructions and/or data within one or more com puter processes and/or (ii) logic implemented in electronic circuitry. Routing table 212 and token table 15 214 are data stored persistently in memory 204. In this illustrative embodiment, routing table 212 and token table 214 are each organized as a database. Logic flow diagram 300 (Figure 3) illustrates the processing of packets by routing logic 210 of node 108A to thereby transport the packets through network 106. In step 302, routing logic 210 re 20 ceives the packet through network access circuitry 208 (Figure 2). In step 304 (Figure 3), routing logic 210 determines the next node to which the packet should be sent toward the packet's ultimate destination through network 106 by reference to routing table 212 (Figure 2). Steps 302 (Figure 3) and 304 are con ventional and are known and are not described further herein. 25 In step 306, routing logic 210 tags the packet with the token of node 108A in a manner de scribed more completely below in conjunction with logic flow diagram 306 (Figure 4). In alternative embodiments, routing logic 210 can tag the packet with the token of the next node determined in step 304 in addition to or instead of tagging the packet with the token of node 108A. In such alternative em bodiments, node 108A at least knows the token of each adjacent node among nodes 108A-I and stores 30 all known tokens in routing table 212 (Figure 2). In embodiments in which nodes 108A-I know each other's tokens, nodes 108A-1 inform one another of their tokens in a manner described below during net work configuration.

6 In step 308 (Figure 3), routing logic 210 sends the packet as tagged to the next node toward the ultimate destination of the packet through network 106. The sending of a packet to a next node is conventional and known and not described further herein. However, in some embodiments in which each of nodes 108A-1 knows the tokens of at least its adjacent nodes, the sending node can encrypt pay 5 load 706 of the packet for the next node, thereby preventing interception of the packet by an unauthor ized node or other nefarious logic that might be injected into either node 108A or the node to which node 108A sends the packet. Logic flow diagram 306 (Figure 4) shows step 306 in greater detail. In step 402, routing logic 10 210 finds the first empty token slot of the packet. An example of a packet transported by node 108A is shown as packet 702 (Figure 7). Packet 702 includes a header 704 and a payload 706. Payload 706 is that portion of packet 702 that is data intended to be transported through network 106 and is conventional. Header 704 is 15 mostly conventional except that header 704 includes a number of token slots 708A-C, each of which can store a single token. In this illustrative embodiment, each token is 16 octets in length and header 704 can store three (3) tokens, one in each of token slots 708A-C. Of course, tokens can be of different lengths and header 704 can include different numbers of token slots in alternative embodiments. 20 In this illustrative embodiment, an empty token slot stores all zeros. Accordingly, 16 octets of all zeros is not a valid token. Thus, routing logic 210 finds the first empty token slot by identifying the first of token slots 708A-C that stores all zeros. In test step 404 (Figure 4), routing logic 210 determines whether an empty token slot was 25 found in step 402. If so, processing transfers to step 414 in which routing logic 210 stores the token of node 108A into the first empty token slot of the packet. As noted above, routing logic 210 can store the token of the next node in the first empty slot of the packet, first ensuring that the token of node 108A is stored 30 in the immediately preceding token slot. Conversely, if routing logic 210 did not find an empty token slot in step 402, processing by routing logic 210 transfers from test step 404 to step 406.

7 In step 406, routing logic 210 retrieves the tokens from the packet, e.g., retrieves the tokens stored in token slots 708A-C. In step 408, routing logic 210 determines an abbreviation for a sequence of at least two tokens 5 in a manner described more completely below with respect to logic flow diagram 408 (Figure 5). The abbreviation is itself a token that is unique among all tokens known to node 108A and all tokens for any nodes of network 106. In step 410, routing logic 210 replaces the two or more tokens represented by the abbreviation 10 with the abbreviation itself. Since the abbreviation is a single token replacing at least two other tokens, at least one token slot will be freed and therefore empty. Routing logic 210 marks the token slots freed by replacement with the abbreviation as empty by storing all zeros therein. In step 412, routing logic 210 identifies the first empty token slot after the abbreviation substi 15 tution of step 410. Processing transfers from step 412 to step 414 in which routing logic 210 stores the token of node 108A into the first empty token slot of the packet as described above. After step 414, processing according to logic flow diagram 306 ends, and therefore step 306 (Figure 3), completes. 20 Step 408 (Figure 4) is shown in greater detail as logic flow diagram 408 (Figure 5). In step 502, routing logic 210 retrieves, from token table 214 (Figure 2), an abbreviation token that represents the full sequence of tokens retrieved from token slots 708A-C of the subject packet. In 25 other embodiments, routing logic 210 can also retrieve an abbreviation token that represents a contigu ous sub-sequence of the full sequence of retrieved tokens. Token table 214 includes one or more token definitions such as token definition 602 (FIGURE 6). Token definition 602 includes a token 604 and a definition 606. Definition 606 is a sequence of two 30 or more tokens. Token 604 is a token that is an abbreviation of the sequence of tokens represented in definition 606.

8 To identify an abbreviation for a sequence of two or more tokens, routing logic 210 searches token table 214 for a token definition whose definition 606 is that sequence. The corresponding token 604 is the abbreviation. 5 In test step 504 (Figure 5), routing logic 210 determines whether an abbreviation was found within token table 214 in step 502. If so, routing logic 210 determines, in step 506, that the retrieved abbreviation is the appropriate abbreviation for the sequence of tokens and processing according to logic flow diagram 408, and therefore step 408 (Figure 4) completes. 10 In alternative embodiments, routing logic 210 can use a hashing function to map multiple to kens to a single abbreviation token. In such embodiments, the look-up and test of steps 502 and 504 can be replaced with a single hashing step. In some embodiments, the hashing function is designed to avoid producing hashed tokens that can be confused with a node token. Processing from such a hashing step would transfer to step 506, which is described above. 15 Returning to logic flow diagram 408 (Figure 5), if an abbreviation was not found within token table 214 in step 502, processing by routing logic 210 transfers from test step 504 to step 508. In step 508, routing logic 210 creates a token that is unique from all abbreviation tokens 20 stored in token table 214 (Figure 2) and from all tokens of nodes in network 106 (Figure 1). In this illus trative embodiment, the total range of values for legitimate tokens is divided into a range reserved for nodes of network 106 and a range that each of nodes 108A-I can use in their respective token tables 214. For example, for a token 16 octets in length, all tokens in which the most significant octet is zero can be reserved for tokens of nodes of network 106. That would leave roughly seven times as many tokens for 25 representing sequences of tokens within each node. Similar rules may be applied in other embodiments where the token has a length other than 16 octets. For example, in a system having token lengths of 4 octets, all tokens in which the most significant bit is zero can be reserved for tokens of nodes of network 106, while all other 4-octet tokens may represent token sequences. 30 Each of nodes 108A-I is informed of the reserved token range during the registration process by which each node is assigned its own token. During system configuration, each of nodes 108A-1 regis ters with node manager 110, identifying itself with a digital fingerprint in this illustrative embodiment. Digital fingerprints are known and are described, e.g., in U.S. Patent 5,490,216 and that description is 9 incorporated herein by reference. In general, a digital fingerprint uniquely identifies the physical device (e.g., a computer or router) based on a sampling of user-configurable and/or non-user-configurable ma chine parameters readable from the device, wherein each parameter may represent a particular hardware or a software configuration associated with the device. 5 It should be appreciated that the token created by routing logic 210 in step 508 need not be unique with respect to tokens used by nodes 108B-1 in their respective token tables. As a result, routing logic 210 of node 108A need not consult any other node or computer to create an adequately unique new token in step 508, thus avoiding significant delay in transport of the subject packet to its ultimate desti 10 nation through network 106. However, in such embodiments, routing logic 210 of node 108A should take care to not replace the last token of node 108A with an abbreviation as the last token of node 108A identifies node 108A as the particular one of nodes 108A-1 that is capable of reversing the abbreviation. In alternative embodiments, creation of an abbreviation token for multiple other tokens can be per formed in a way that is both deterministic and global within nodes 108A-I such that an abbreviation 15 used by any of nodes 108A-1 can be properly reversed by any of nodes 108A-1. In these alternative em bodiments, routing logic 210 of node 108A can replace the last token of node 108A with an abbrevia tion. In step 510, routing logic 210 creates a new token definition that associates the token created 20 in step 508 with the sequence of tokens to be replaced with the new token. In step 512, routing logic 210 returns the token created in step 508 as the abbreviation token and ends processing according to logic flow diagram 408, and therefore step 408 (Figure 4) completes. 25 To illustrate the packet transportation described above, the recording of a route of a packet transported through network 106 (Figure 1) from computer 102 to computer 104 is described. As shown in Figure 1, computer 102 connects to network 106 through node 108A. Accordingly, node 108A is the first to process a packet from computer 102. 30 As shown in Figure 8A, token slots 708A-C of the packet are initially all empty. Performance of step 306 by node 108A results in node 108A storing its token in the first empty token slot, i.e., in to ken slot 708A in this illustrative example. Node 108A forwards the packet so tagged to node 108C.

10 As shown in Figure 8B, when received by node 108C, token slot 708A stores the token of node 108A and token slots 708B-C of the packet are initially empty. Performance of step 306 by node 108C results in node 108C storing its token in the first empty token slot, i.e., in token slot 708B in this illustrative example. Node 108C forwards the packet so tagged to node 108F. 5 As shown in Figure 8C, when received by node 108F, token slot 708A stores the token of node 108A, token slot 708B stores the token of node 108C, and token slot 708C is empty. Performance of step 306 by node 108F results in node 108F storing its token in the first empty token slot, i.e., in to ken slot 708C in this illustrative example. Node 108F forwards the packet so tagged to node 108H. 10 As shown in Figure 8D, when received by node 108H, token slot 708A stores the token of node 108A, token slot 708B stores the token of node 108C, and token slot 708C stores the token of node 108F. None of token slots 708A-C is empty. Performance of step 306 by node 108H results in (i) re placement of the sequence of tokens for nodes 108A, 108C, and 108F with an abbreviation 802 and (ii) 15 node 108H storing its token in the first empty token slot, i.e., in token slot 708B in this illustrative ex ample. Replacing three (3) tokens with one (1) frees up two token slots in the subject packet. In addi tion, since the token of node 108H immediately follows abbreviation 802, node 108H is marked as the author of abbreviation 802. Such is used in reconstructing the route of the subject packet in the manner described below. Node 108H forwards the packet so tagged to node 1081. 20 Node 1081 processes the packet in an analogous manner and stores its own token in token slot 708C as shown in Figure 8E. It should be appreciated that another node can replace abbreviation 802, the token of node 108H, and the token of node 1081 with another abbreviation token. Node 1081 for ward the packet tagged with the token of node 1081 to computer 104, to thereby effect delivery of the 25 packet to computer 104. Tracing a route taken by a particular packet is illustrated by logic flow diagram 900 (Figure 9). The steps of logic flow diagram can be performed by logic in any of nodes 108A-1, node manager 110, and computers 102 and 104. For the purposes of current discussion, the logic is referred to as 30 "traceback logic".

I I The traceback logic stores a route 1002 (Figure IOA-E) and a token list 1004, both of which are lists of tokens. Initially, route 1002 is empty and token list 1004 includes the tokens stored in token slots 708A-C of the subject packet, as shown in Figure I0A. 5 Loop step 902 and next step 914 define a loop in which the traceback logic processes route 1002 and token list 1004 according to steps 904-912 until token list 1004 is empty. In step 904, the traceback logic pops the last token from token list 1004. As used herein, pop ping the token from token list 1004 means retrieving the token from the last position in token list 1004 10 and removing the retrieved token from token list 1004. As shown in Figure 10A, the last token of token list 1004 identifies node 1081. In test step 906, the traceback logic determines whether the popped token identifies a node. If so, processing transfers to step 908. Conversely, if the popped token does not identify a node, process 15 ing by the traceback logic transfers from test step 906 to step 910. In this illustrative example, the token popped from the last position in token list 1004 is the token of node 1081. Accordingly, processing by the traceback logic transfers to step 908 in which the traceback logic pushes the popped token onto the beginning of route 1002. The result is shown in FIG 20 URE 1OB in which the token for node 1081 is popped from the end of token list 1004 and is pushed on to the beginning of route 1002. After step 908, processing by the traceback logic transfers to next step 914, in which the loop of steps 902-914 are repeated if token list 1004 is not empty. 25 In the next iteration of the loop of steps 902-914 in this illustrative example, the traceback logic pops the token of node 108H from the end of token list 1004 and pushes the token on to the begin ning of route 1002 in an analogous manner. The result is shown in Figure lOC. 30 In the next iteration of the loop of steps 902-914 in this illustrative example, the traceback logic pops abbreviation 802 (Figure lOC) from the end of token list 1004. In test step 906 (Figure 9), the traceback logic determines that abbreviation 802 is not a node token and processing of the traceback logic transfers to step 910.

12 In step 910, the traceback logic queries the node whose token is at the top of route 1002 (Fig ure 1OC) for expansion of abbreviation 802. It should be appreciated that, in embodiments such as those described above in which an abbreviation is a hash of the multiple tokens and is produced in a manner 5 shared by all of nodes 108A-I, it is unnecessary to query the particular one of nodes 108A-I that stored abbreviation 802 in the packet and step 910 is therefore obviated. However, as described above in some embodiments, each of nodes 108A-1 maintains its own token table 214 (Figure 2) separately and inde pendently of other nodes of network 106. Accordingly, the only node of network 106 that can expand abbreviation 802 in this illustrative embodiment is the node that created abbreviation 802. Since the 10 node that created abbreviation 802 in a performance of step 408 (Figure 4) stored its own token in the subject packet in step 414, the token immediately following abbreviation 802 identifies the node that authored abbreviation 802. As shown in Figure 1OC, the token of node 108H is at the top of route 1002 and is therefore 15 the node that created abbreviation 802. The traceback logic therefore queries node 108H for expansion of abbreviation 802 in step 910. Each of nodes 108A-I is configured to receive requests for expansion of abbreviation tokens and to respond by returning the sequence of tokens associated with the received abbreviation token within token table 214. In this illustrative example, node 108H responds with the sequence described above, namely, tokens for nodes 108A, 108C, and 108F in order. 20 In step 912, the traceback logic appends the token sequence received in step 910 to token list 1004 (Figure IOD). As shown in Figure IOD, token list 1004 does not include abbreviation 802 (popped from token list 1004 in step 904) and includes the expansion thereof (appended in step 912). It should be appreciated that the token list resulting from expansion of abbreviation 802 can include other abbre 25 viations, including abbreviations authored by other nodes. Subsequent iterations of the loop of steps 902-914 (Figure 9) by the traceback logic result in moving the tokens of nodes 108F, 108C, and 108A from the end of token list 1004 to route 1002 in se quence until token list 1004 is empty, as shown in Figure 10E. 30 Thus, in accordance with the present invention, a route of five (5) hops was completely re corded and reconstructed from only three (3) slots available to record nodes to which the packet hopped. Given the recursive nature of the recording of the nodes as described above, i.e., that abbreviations of 13 token sequences can themselves include abbreviations of token sequences that can in turn include ab breviations of token sequences, the length of a packet route through a network that can be traced is unlimited, aside from the practical limitations of the collective capacity of token table 214 of all nodes of the network. 5 As described above, some embodiments require that each of nodes 108A-1 knows all the to kens of at least its adjacent nodes, i.e. those of nodes 108A-1 with which data is directly exchanged. Of course, all of nodes 108A-1 can know the tokens of all others of nodes 108A-l, but such is not always necessary. For example, since node 108A never directly exchanges data with node 108D, node 108A is 10 not always required to have the token of node 108D. As described briefly, each of nodes 108A-1 receives a token from node manager I10. Node manager 110 derives the token from a digital fingerprint of each node. Thus, the token identifies the particular physical device that acts as a node and not an easily reconfigurable attribute such as a network 15 address. Nodes 108A-I learn the tokens of others of nodes 108A-I during route configuration. Route configuration involves an exchange of information among nodes 108A-I to build routing table 212. In particular, each of nodes 108A-1 builds a routing table such that a given destination address of a packet 20 indicates to which node the packet should be forwarded. Conventional route configuration protocols include RIP (Routing Information Protocol), EIGRP (Enhanced Interior Gateway Routing Protocol), and OSPF (Open Shortest Path First). In each such route configuration protocol, one or more packets are exchanged between nodes 25 108A-1 to share various elements of information of nodes 108A-1 that can be used by each of nodes 108A-1 to properly identify a next node in routing a particular packet to its destination. In this illustra tive embodiment, nodes 108A-1 include their respective tokens assigned by node manager 1 10 in such route configuration packets, either by including the token as a field in a packet conveying other items of information (such as an additional field in a RIP packet) or as an additional packet such as a Hello 30 packet in EIGRP or OSPF. The above description is illustrative only and is not limiting. The present invention is defined solely by the claims which follow and their full range of equivalents. It is intended that the following 14 appended claims be interpreted as including all such alterations, modifications, permutations, and substi tute equivalents as fall within the true spirit and scope of the present invention. It will be understood that the term "comprise" and any of its derivatives (eg. comprises, 5 comprising) as used in this specification is to be taken to be inclusive of features to which it refers, and is not meant to exclude the presence of any additional features unless otherwise stated or implied. The reference to any prior art in this specification is not, and should not be taken as, an acknowledgement of any form of suggestion that such prior art forms part of the common general 10 knowledge.

Claims (5)

1. A method for routing a packet through a network from a source to a destination, the method comprising: 5 storing data in a header of the packet to represent a complete route of the packet through the network, the data identifying at least one physical routing device of the network through which the packet travels.

2. The method of claim 1 further comprising: 0 replacing data identifying at least two physical routing devices with a single token in the header of the packet such that the single token identifies the at least two physical routing devices.

3. The method of claim 2 further comprising: storing data in the header of the packet that identifies a particular physical routing device that 5 performs the replacing.

4. The method of claim I wherein the storing is performed by node logic executing within each physical routing device that receives the packet along the rate. 0

5. A method for identifying a particular route taken by a packet through a network, the method comprising: retrieving one or more tokens from a header of the packet, the tokens collectively identifying one or more physical routing devices through which the packet traveled; determining that at least a selected one of the tokens represents two or more other tokens; and 5 replacing the selected token with the two or more other tokens.