Classifications

• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L45/00—Routing or path finding of packets in data switching networks
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L45/00—Routing or path finding of packets in data switching networks
  • H04L45/74—Address processing for routing
  • H04L45/745—Address table lookup; Address filtering
  • H04L45/74591—Address table lookup; Address filtering using content-addressable memories [CAM]

Definitions

• the present invention relates to associative memories, and in particular memories of the "TRIE" type (from the English verb "reTRIEve”).
• ACM Communications of the ACM, Vol. 3, No. 9, September 1960, pages 490-499). It consists of cutting the bit strings to be recognized into successive slices of fixed length (of K bits) and integrating them into a two-dimensional table T. Each row of the table constitutes a register of 2 K elementary cells.
• a register (R) is assigned to each section of the chain and a cell in the register is associated with the value (V), between 0 and 2 K -1 of this section.
• the register assigned to the first section of the chain which is also the entry point of the table, is called a porter.
• the data to be analyzed in the form of bit strings, that is to say to compare with the content of the TRIE memory, will also be called routes below.
• the succession of chained cells associated with a route will be called path in the table.
• Each register of the table will be said to be of order i> 0 if it is assigned to the (i + 1) th section of one or more stored routes.
• the gatekeeper register is therefore of order 0.
• the memory ⁇ TRîE ⁇ associates with each of its registers of order i> 0 a unique sequence of iK bits corresponding to the first iK bits of each route whose path in the table passes through a cell of the register in question.
• TRIE can be presented as shown in Figure 1, where the underlined data is status.
• the patterns 45A4, 45AB, 67AB, 788A and 788BD are respectively represented in the table of FIG. 1 by the paths: T [0.4] -> T [1, 5] - »T [2, A] • T [ 3.4] T [0.4] - T [1, 5] -> T [2, A] -> T [3, B] T [0.6] - T [4.7] - »T [ 5, A] - * T [6, B] T [0.7] - * T [7.8] - »T [8.8] -> T [9, A] T [0.7] - T [7.8] - T [8.8] - * T [9, B] - »T [10, D].
• the analysis rank i is set to 0 and the gate register R 0 is selected as register r.
• the content C of cell T [r, V j ] designated by the (i + 1) -th tranche V s of the route in the selected order register i is read in step 2. If this cell contains a further analysis pointer, as indicated in test 3 by the value 1 of a bit FP (C) stored in the cell, the order register i + 1 designated by this pointer Ptr (C) is selected as register r for the next iteration in step 4, and rank i is incremented.
• the status Ref (C) read in the cell concerned is returned to step 5 as a result of consulting the table.
• This algorithm allows the analysis of routes comprising any number of sections.
• the same table can be used for several types of analyzes by managing the data from different gatekeepers.
• it allows to control the time of data analysis: the analysis of a number
• N of K-bit slices will last at most N times the duration of an iteration.
• the algorithm of FIG. 2 can be implemented very quickly by a hardware component managing access to the memory table.
• a hardware component managing access to the memory table.
• the packet header is analyzed on the fly by the component, and the status associated with a route designates, for example, an output port of the router to which the packets carrying a destination address conforming to this route must be routed.
• Such a router can be multi-protocol.
• the reference in question can also trigger timers or jumps of a determined number of bits in the analyzed header in order to be able to choose which portion of the header should then be analyzed.
• a certain number of analyzes are generally executed successively, to trigger the operations required by the supported protocols according to the content of the headers.
• One of these analyzes will relate to the destination address to perform the routing function proper.
• the fact of being able to chain several elementary analyzes with programmable hops between them gives great flexibility to the process, particularly for the treatment of protocols encapsulated according to several layers of the ISO model.
• the on-the-fly analysis of the header slices as they arrive also provides great speed.
• TRIE tables Another advantage of the TRIE tables is that it allows routing constraints to be taken into account on the basis of the longest recorded path corresponding to a prefix of the route to be recognized, a constraint encountered in particular in the context of IP routing (see EP -A-0 989 502).
• EP-A-1 030 493 describes a TRIE memory, the content of which includes, in addition to the references proper associated with the packet headers, a program consisting of the sequence of elementary analyzes to be carried out according to the different configurations taken into account by Memory. These sequences are fully programmable. The user can arbitrarily define, at each step of the process, which portion of the header should be examined and from which register of the TRIE memory, which provides great processing flexibility.
• a TRIE memory can also be described in tree form, with nodes distributed in several successive stages corresponding to the analysis orders i previously mentioned. Each node of a stage i represents a decision to be made during the analysis of the (i + 1) th section of a route.
• the root node of the tree corresponds to the gatekeeper register, the leaf nodes to the status, and the intermediate nodes to the registers designated by the further analysis pointers.
• the tree representation makes it easy to visualize the paths.
• the tree in FIG. 3 thus shows the paths recorded in the table in FIG. 1, the root and the intermediate nodes being represented by circles (registers) and the leaves by rectangles (status).
• the tree representation makes it possible to design compression methods aimed at reducing the memory size required to implement a TRIE table. This reduction is particularly useful for rapid implementations of large tables using static memory circuits (SRAM).
• SRAM static memory circuits
• a hardware implementation in the form of a table where each register contains 2 K cells is ineffective in terms memory occupation because such a table has many empty cells, as shown in Figure 1.
• the nodes close to the root have a number of descendants valid close the number of possible descendants (2).
• the average number of valid descendants of a given node decreases considerably and tends towards 1 (or 2 if we take into account a default status). In this case, there are only between 10% and 15% of useful cells in the memory.
• path compression consists in aggregating at a node Y a stage i the non-empty nodes of stages i + 1 to i + j-1 (j ⁇ 2) which are descendants of this node Y when each of these nodes of stages i to i + j-1 has a single non-empty descendant (register or status). See also US-A-6,014,659 or US-A-6,505,206.
• level compression consists in aggregating at a node Z of a stage i the non-empty nodes of stages i + 1 to i + j-1 (j ⁇ 2) which are descendants of this node Z when each of these nodes of stages i + 1 to i + j-1 itself has at least one non-empty descendant (register or status).
• the length of the slice to be analyzed in relation to the compressed node Z is multiplied by j; - compression in extent ("width compression” or "pointer compression”) consists in eliminating the empty descendants of a given node, see also US-A-5,781,772, EP-A-0 458 698 or WO 00/75804 . It is useless to reserve a register of 2 K cells to analyze a slice having only L ⁇ 2 K valid values in paths recorded in the TRIE memory: we can be satisfied with a compressed area of L cells, associated with map data showing values valid of the range.
• This cartographic data typically takes the form of a bitmap vector of 2 K bits set to 1 at the L positions corresponding to the L valid values of the slice and at 0 at the 2 K -L other positions.
• the modification during analysis of the length of the slices cut out in the data to be analyzed does not lend itself to rapid implementation in a specific hardware component. It essentially makes it possible to reduce the memory size required by a software implementation, which by nature is slower.
• the extended compression method does not suffer from this limitation. However, it requires that the actual analysis of a section be preceded by the analysis of the associated bitmap to validate the value of the section and locate the corresponding cell.
• This method can be implemented in a fast hardware module, but a limitation to this speed is that its complexity increases strongly with the width K of the slice. Indeed, the function used to obtain the address of the successor of a given node requires resources (size of the nodes) of complexity proportional to 2.
• a TRIE table is suitable for parallel processing in pipeline mode, as mentioned in the article "Putting Routing Tables in Silicon", by TB. Pei et al., IEEE Network Magazine, January 1992, pages 42-50.
• M the maximum number of stages in the tree
• the available memory space can be divided into N memory planes, with N ⁇ M.
• Each memory plan P: of level j (0 ⁇ j ⁇ N) is reserved for the nodes of one or more consecutive stages of the tree.
• N operators operate in parallel with each a respective buffer containing a data chain to analyze.
• This pipeline processing by the N operators increases the maximum processing rate of the device.
• An object of the present invention is to propose an efficient method for compressing a TRIE memory, which facilitates high speed processing of strings of data to be analyzed and can be implemented by a hardware component of limited complexity.
• the invention thus provides a TRIE memory device, comprising means for storing binary patterns associated with respective references, and means for analyzing data strings by successive K-bit slices to extract one of the references during a agreement between an analyzed data string and a stored binary pattern associated with this reference (K> 1).
• the storage means comprise several successive stages, each including several memory cells.
• Each non-empty memory cell of a stage i ⁇ 0 contains a cell type indicator and data comprising: - a pointer designating another memory cell when the cell type indicator is in a first or a second state , the pointer being accompanied by a test value on K bits when the cell type indicator is in the second state; - a reference associated with a binary pattern stored when the cell type indicator is in a third state.
• the analysis means include: - means for reading a cell of a stage i ⁇ 0 in connection with the analysis of the (i + 1) th section of a data chain; - means for selecting a cell of stage i + 1, to be read in relation to the analysis of a (i + 2) -th slice of the data chain, in response to the first state of the indicator in said cell of stage i, the selected cell being located in relation to the cell designated by the pointer contained in said cell of stage i as a function of the value of the (i + 1) th tranche of the data string; means for selecting the cell designated by the pointer contained in said cell of stage i, to be read in relation to the analysis of a (i + 2) -th slice of the data chain, in response to the second state of the indicator in said cell of stage i when the value of the (i + 1) -th slice of the data chain coincides with the test value contained in said cell of stage i; and - means for extracting the reference contained in said cell of stage i in response to the
• the sons of a node of the TRIE tree can in particular be located, preferably in a reversible manner, either in a register comprising all the possible successors of a given node (register of 2 K cells), or in a zone constituting a minimum register (of one or two cells) as soon as the conditions of "path compression" are met (the node has only one valid node-child).
• a pointer then designates either the first cell of the 2 K cell register, or the cell or the first cell of the zone constituting the minimum register.
• the device thus implements an intermediate process between compression in extent and compression of path.
• Compression results from the fact that a tree node through which only one memorized path passes does not need to point to another register of 2 K cells. It just needs to point to a reduced area of one or two cells, which occupies less memory space.
• An advantage of this device is that it lends itself to a simple hardware implementation given that the size of the slices analyzed remains fixed and that the analysis of a slice can be carried out in a clock cycle. However, like the extended compression method, it does not suffer from the exponential increase in the size of the nodes with the size of the slices analyzed. With a given component technology, it is then possible to increase the processing speed of the device by increasing the size of the slices without being penalized by too great a complexity.
• the analysis means preferably return a default reference.
• This default reference can be common to the entire TRIE tree. It is however advantageous that it depends on the node to which said cell of stage i corresponds. The default reference can then be the one associated with the longest recorded path corresponding to a prefix of the route to be recognized ("iongest match").
• the storage means are divided into N distinct memory areas of levels 0 to N-1, N being less than a maximum number of stages of the storage means, and the analysis means are organized in pipeline in relation to the N memory areas.
• FIG. 2 is a flow diagram of a conventional analysis procedure executed to consult the TRIE memory
• - Figure 3 is a tree representation of the TRIE memory having the content illustrated in Figure 1
• - Figure 4 is a block diagram of a packet router incorporating a device according to the invention
• - Figure 5 is a diagram of a circuit forming a device according to the invention
• - Figure 6 is a tree representation of a TRIE memory organized according to the invention
• - Figure 7 shows in three diagrams an example of the content of a memory cell in one embodiment of the invention
• - Figure 8 is a tree representation of a TRIE memory organized according to another embodiment of the invention
• - Figure 9 is a flowchart illustrating a software-based embodiment of a device functionally similar to that of Figure 5.
• ATM Asynchronous Transfer Mode
• the router 10 shown in FIG. 4 operates with a host computer 11.
• the host computer 11 can send and receive packets, in particular for managing the routing process. For this, it has a virtual channel (VC) at the input and output of router 10.
• VC virtual channel
• the router 10 comprises a forwarding module 12 which routes the received packets according to instructions, hereinafter called “routing references” or "final status", obtained by an analysis module 13 from of a memory 14 organized as a TRIE memory table.
• the routing module 12 can essentially carry out a translation of the virtual path identifiers VPI / VCI (Virtual Path Identifier / Virtual Channel Identifier), the fusion of the virtual channels according to the virtual paths, and the delivery of the packets on the output ports of the router. For this, it needs to know the VPI / VCI pairs of outgoing packets, which can constitute the routing references stored in the TRIE memory 14.
• Each ATM cell containing the header of a packet to be routed passes through a buffer memory 15 to which the analysis unit 13 has access to analyze portions of these headers by means of the TRIE memory 14.
• Configuring the router 10 consists in recording the relevant data in the TRIE memory 14. This operation is carried out by a unit (not shown) for managing the TRIE memory under the control of the computer. host 11.
• the configuration commands can be received in packets transmitted over the network and intended for the router 10.
• EP-A-0 989 502. In the example of router shown in FIG. 4, the analysis unit
• the router 13 cooperates with a PLC 16 programmed to perform certain checks and certain actions on the packet headers, in a manner dependent on the communication protocols supported by the router. Apart from this automaton 16, the operation of the router 10 is independent of the packet transport protocols.
• FIG. 5 shows a TRIE memory device according to the invention.
• each elementary cell of the TRIE memory occupies
• FIG. 5 Found in FIG. 5 is the analysis unit 13, the TRIE memory 14, as well as the buffer memory 15 intended to receive a data chain to be analyzed by K-bit slices.
• the TRIE memory comprises a memory plane 14, advantageously produced in SRAM ("Static Random Access Memory") technology.
• This memory plane comprises a data bus D of width 32 bits, as well as an address bus AD, the width of which depends on the quantity of data to be stored in the memory TRIE.
• Each register includes one or more memory cells of 32-bit size, addressable by the AD bus.
• FIG. 6 shows a way in accordance with the invention of organizing the cells of a TRIE tree from a root node of a stage ⁇ .
• the register 100 is the gatekeeper register of the memory.
• the cells marked S in FIG. 6 are of the status type and contain a reference associated with one of the saved paths.
• the cells marked R each contain a pointer in "register” mode, which designates a register of 2 K cells of the next stage. The pointer in register mode therefore designates, explicitly or implicitly, the first cell of this register of the next stage.
• the cells marked L each contain a pointer in "linear" mode, which designates a cell isolated from the next stage.
• the other cells, not marked in FIG. 6, do not contain useful information relating to recorded paths. It can be seen that the use of the linear mode makes it possible to reduce the number of these empty cells, and therefore to optimize the use of the available memory space.
• pointer In register mode, the operation of pointers is identical to that described in the introduction.
• the pointer locates the register in which the analysis will be to continue, and the value of the current section of K bits makes it possible to locate the cell of this register which will be read to continue the analysis.
• FIG. 7 shows the content of a memory cell in a particular exemplary embodiment.
• the first four bits of the cell represent an FP flag whose value indicates in particular the type (status, pointer in register mode or pointer in linear mode) of the data stored in the cell.
• the 28 remaining bits of the cell constitute the Ref reference used in the operation of routing packets and / or commands intended for the PLC 16.
• the address calculation logic 21 proceeds to concatenate the pointer PtrR and the value V j of the next slice to be analyzed to generate the address AD at which the next cell will be read in the memory 14.
• the address calculation logic 21 provides on the address bus AD of memory 14 an address corresponding to the pointer PtrL received from the detection 20 so that the isolated cell designated by the pointer is read in relation to the slice V j . If V j ⁇ Val, the analysis ends by indicating that the analyzed chain does not correspond to any path recorded in the TRIE memory. the above description, the unoccupied cells of the tree
• TRIE (or the cells eliminated in linear mode) give rise, when they are encountered during the analysis, to an indication of error interpreted as the absence of recorded pattern corresponding to the analyzed chain.
• these cells are associated with a default reference which the detection circuit 20 returns when such a cell is encountered.
• the default reference may, in certain applications, be the same for the entire TRIE tree. In this case, it is not necessary to store it in the nodes of the tree, so that one can use the data structure shown diagrammatically in figure 6.
• this default reference can vary depending on the mother cell whose pointer designates such a default reference. This advantageously makes it possible to integrate the constraints linked to the "longest match" into the analysis.
• the memory 14 can in particular receive a data structure such as that illustrated in FIG. 8.
• the TRIE tree represented in this FIG. 8 contains the same paths as that of FIG.
• FIG. 6 Each unoccupied cell in FIG. 6 is now occupied by a default reference, which is symbolized by the letter D in FIG. 8.
• cells marked D contain the same default reference, which depends on the mother cell pointing to this register.
• linear mode the cells are no longer isolated, but grouped in pairs.
• the first cell of the pair has the same content as the isolated cell according to Figure 6, while the other cell contains a default reference, returned when V j ⁇ Val during the analysis of the upstream node.
• the two cells of such a pair have consecutive addresses. Consequently, the analysis can be carried out using the circuit of the figure: it suffices that the address calculation logic 21 completes the pointer PtrL in linear mode by concatenating therein the bit v delivered by the comparator 22 to produce the address of the next cell to be read in linear mode.
• FIG. 9 A software-based embodiment of this latter embodiment of the invention is illustrated in FIG. 9.
• the flow diagram illustrates the operation of the circuit of FIG. 5, and can be used to produce an emulator.
• the preferred embodiments of the invention are based on hardware, but it can also be envisaged, for applications which are not too fast, to use software.
• the analysis rank i is set to 0 and the gate register R 0 is selected as register r.
• the variable v of K bits receives the value V s of the current slot.
• step 41 the algorithm returns to step 32 above. Its execution ends when a status is encountered (34), or on an error (43) if a cell type cannot be decoded (FP (C) ⁇ L in test 38).
• the mode of compression and storage of the data of the TRIE memory which has just been described makes it possible to perform the analysis of the data strings in pipeline mode in order to increase the rate of analysis of the data, the memory 14 being distributed. in N distinct memory zones of levels 0 to N-1, as indicated above.

Landscapes

• Engineering & Computer Science (AREA)
• Computer Networks & Wireless Communication (AREA)
• Signal Processing (AREA)
• Data Exchanges In Wide-Area Networks (AREA)

Applications Claiming Priority (1)

Publications (1)

Family

ID=34630219

Family Applications (1)

Country Status (4)

Families Citing this family (4)

Family Cites Families (8)

• 2003
  • 2003-10-28 WO PCT/FR2003/003215 patent/WO2005052812A1/fr active Application Filing
  • 2003-10-28 AU AU2003292288A patent/AU2003292288A1/en not_active Abandoned
  • 2003-10-28 EP EP03767854A patent/EP1678632A1/de not_active Ceased
• 2006
  • 2006-04-27 US US11/380,550 patent/US7444562B2/en active Active

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events