JP2019115001A - Table conversion apparatus, table conversion method, and gateway apparatus - Google Patents

Abstract

To increase versatility of a table that can convert input values into multiple output values.SOLUTION: The table conversion method is a method for converting a table in which a plurality of output IDs is allocated to one input ID, which makes an input ID an input ID column, makes a plurality of output IDs assigned to the input ID an output ID column, and makes a line number of an output value column a pointer column, so that a plurality of output IDs assigned to the input ID is designated one by one in order.SELECTED DRAWING: Figure 7

Description

  The present invention relates to a table conversion device, a table conversion method, and a gateway device.

  Patent Document 1 discloses an address conversion system which converts a given virtual address into a real address. This system has a memory that stores a virtual address and a real address in association with each other. The system refers to the memory and translates the virtual address into a real address.

U.S. Pat. No. 4,453,230

  By the way, in a table for converting values such as addresses, there are cases where a plurality of output values are associated with one input value. In such a case, it is desirable to increase the versatility of a table that can convert an input value into a plurality of output values.

  Other problems and novel features will be apparent from the description of the present specification and the accompanying drawings.

  According to one embodiment, the table conversion method uses a plurality of output values assigned to input values as output value columns, and sequentially designates a plurality of output values assigned to the input values one by one. Thus, let the row number of the output value column be a pointer column.

  According to the embodiment, it is possible to increase the versatility of the table capable of converting the input value into a plurality of output values.

It is a figure which shows the CPU system which carries out multicast transmission of a message using a multicast table. FIG. 1 illustrates a system for multicasting messages using a multicast table. It is a figure which shows a multicast table. It is a figure which shows the system at the time of adding transmission destination ID. It is a figure which shows the problem at the time of adding transmission destination ID. FIG. 2 is a diagram showing a multicast table according to the first embodiment. It is a figure which shows the procedure in the case of producing | generating the multicast table concerning this embodiment. It is a figure which shows the procedure in the case of producing | generating the multicast table concerning this embodiment. It is a figure which shows the procedure in the case of producing | generating the multicast table concerning this embodiment. It is a figure which shows the procedure in the case of producing | generating the multicast table concerning this embodiment. It is a figure which shows the procedure in the case of producing | generating the multicast table concerning this embodiment. It is a figure which shows the example which adds a transmission destination address to the multicast table shown in FIG. FIG. 16 is a diagram showing an example of optimizing a multicast table in Embodiment 2; It is a figure for demonstrating the concrete size of a multicast table. It is a figure for demonstrating the concrete size of a multicast table. FIG. 16 is a diagram showing a multicast table according to the third embodiment. FIG. 18 is a diagram showing a multicast table according to a fourth embodiment. FIG. 18 is a diagram showing a multicast table according to the fifth embodiment. FIG. 18 is a diagram illustrating an example of a multicast table according to a fifth embodiment. It is a figure which shows the example which divided | segmented the multicast table shown in FIG. 18 into two tables. It is a figure which shows the example which divided | segmented the multicast table shown in FIG. 19 into two tables. FIG. 18 is a diagram showing a multicast table according to a sixth embodiment. It is a figure which shows the table conversion apparatus which converts a table. It is a figure which shows the gateway apparatus which performs ID conversion using a table.

  The following description and drawings are omitted and simplified as appropriate for clarification of the explanation. In addition, each element described in the drawing as a functional block that performs various processing can be configured by a CPU, a memory, and other circuits in terms of hardware, and in terms of software, a program loaded into a memory And so on. Therefore, it is understood by those skilled in the art that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof, and is not limited to any of them. In the drawings, the same elements are denoted by the same reference numerals, and the redundant description is omitted as necessary.

  Also, the programs described above can be stored and provided to a computer using various types of non-transitory computer readable media. Non-transitory computer readable media include various types of tangible storage media. Examples of non-transitory computer readable media are magnetic recording media (eg flexible disk, magnetic tape, hard disk drive), magneto-optical recording media (eg magneto-optical disk), CD-ROM (Read Only Memory) CD-R, CD -R / W, semiconductor memory (for example, mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, RAM (Random Access Memory)) is included. The programs may also be supplied to the computer by various types of temporary computer readable media. Examples of temporary computer readable media include electrical signals, light signals, and electromagnetic waves. The temporary computer readable medium can provide the program to the computer via a wired communication path such as electric wire and optical fiber, or a wireless communication path.

Embodiment 1
First, the table used in the present embodiment is a table in which a plurality of output values are associated with input values. Specifically, a multicast table in which an input ID (input value) is associated with a plurality of output IDs (output values) is used. When the input ID is input as the input address, by referring to the multicast table, a message or the like is multicast transmitted using the plurality of output IDs corresponding to the input ID as the output address.

  An in-vehicle system to which a multicast table is applied will be described. FIG. 1 is a diagram showing an in-vehicle system to which a multicast table is applied. The in-vehicle system 200 includes a transfer unit 201, a gateway 202, a multicast table 204, a CPU interface 203, and a plurality of CPUs 211 to 214.

  The gateway 202 is configured by hardware such as a semiconductor circuit and includes a plurality of ports. The gateway 202 is connected to each in-vehicle device (not shown) through each port. The gateway relays communication according to a network standard such as CAN (Controller Area Network) or Ethernet (registered trademark). The gateway 202 enables communication between networks with different communication protocols.

  The gateway 202 is connected to the plurality of CPUs 211 to 214 via the CPU interface 203. The transfer unit 201 receives a message from the on-vehicle device via one port of the gateway 202 (arrow A). The transfer unit 201 transfers the received message to the on-vehicle device connected to another port via the gateway 202 (arrow B).

  Furthermore, the transfer unit 201 adds a label to the received message. This label contains the input ID. Therefore, the label is converted to the CPU address of the CPUs 211 to 214 in the multicast table 204. A message can be multicast transmitted to the CPUs 212 and 213 corresponding to the CPU address via the CPU interface 203 (arrow C). That is, based on the multicast table 204, the message is transferred to two or more CPUs corresponding to the CPU address.

  In the sub-systems of the CPUs 211 to 214, addresses based on the CPUs have redundancy. Thus, messages may be sent to the same CPU (e.g., CPU 212) even if they have different labels. In such a case, it is desirable to create a highly versatile table that can change the message output ID as appropriate.

  Next, multicast transmission using a multicast table will be described using FIG. 2 and FIG. FIG. 2 is a diagram illustrating a system 100 in which multicast transmission is performed. FIG. 2 is a diagram showing a multicast table used in the system 100 of FIG. In addition, the table shown below is simplified and shown. Although an actual multicast table is provided with several thousands of IDs, the number of IDs is reduced for simplification of the description.

  As shown in FIG. 2, the system 100 comprises a plurality of subsystems 10-14. The subsystems 10 to 14 correspond to, for example, the CPU systems 211 to 214 in FIG. 1. The subsystem 10 is assigned ID0, ID1, and ID2. The subsystem 11 is assigned ID3 and ID4. The subsystem 12 is assigned ID5 and ID6. The subsystem 14 is assigned an ID 7 and an ID 8. The subsystem 13 is assigned an ID 9. The ID is fixed in each subsystem. Each ID indicates the address of the message.

  As shown in FIG. 2, when the message is transmitted to ID0, the case where the message is transferred to ID3, ID5, and ID7 will be described. The multicast table in such a system 100 is as shown in FIG. Three output IDs of ID3, ID5, and ID7 are associated with the input ID of ID0. Therefore, when a message specifying ID 0 as an input ID is transmitted, ID 3, ID 5, and ID 7 become output IDs.

  The ID conversion table 20 is a multicast table in which an input ID is associated with a plurality of output IDs. Then, the ID conversion table 20 is referred to in order to perform ID conversion among the subsystems 11 to 14.

  In the above example, three output IDs ID3, ID5, and ID7 are associated with the input ID ID0. Therefore, when the subsystem 10 receives a message having ID 0 as an input ID, the subsystem 10 transmits a message with ID 3, ID 5, and ID 7 as output IDs. As a result, the message is transferred to the subsystem 11 to which the ID 3 is assigned, the subsystem 12 to which the ID 5 is assigned, and the subsystem 14 to which the ID 7 is assigned. The subsystem 10 may add a label or the like to the message to be transferred.

  The ID stored in the cell of the output ID is an address indicating the transfer destination of the message. Thus, a typical multicast table as shown in FIG. 3 includes an input ID column (IInput ID Column) and a plurality of output ID columns (Output ID column). The number of multicasts is determined according to the number of IDs stored in the output ID column.

  In such a multicast table, the number of output IDs serving as transfer destination addresses may be limited. For example, in the multicast table shown in FIG. 3, the number of columns (the number of columns) of the multicast table is four. That is, up to three output IDs are associated with one input ID. The number of output IDs serving as transfer destination addresses is three at maximum, and it is not possible to transfer a message to four or more output IDs.

  The case of increasing the output ID corresponding to the input ID will be described with reference to FIGS. 4 and 5. 4 and 5 show an example in which the ID 9 is to be added as a transfer destination in the multicast table shown in FIGS. 2 and 3 (arrow A0 in FIG. 4). FIGS. 4 and 5 show an example in which when a message is transmitted to ID0, the message is transferred to ID3, ID5, ID7, and ID9.

  If the number of columns in the multicast table is limited, ID 9 can not be added as an output ID. For example, as shown in FIG. 5, ID3, ID5, and ID7 are stored in the three output ID columns. Therefore, when the memory size and the number of columns of each cell of the table are limited, ID9 can not be added as an output ID to a row (row) whose input ID is ID0. If the number of columns of the output ID is increased in advance, some of the columns are wasted and the table size becomes large.

  Furthermore, since the number of output IDs corresponding to the input ID is different, there are unused cells. That is, when there are output IDs smaller than the maximum number of output IDs, a part of the memory area can not be used. For example, the number of output IDs corresponding to the input ID 0 is four, and the number of output IDs corresponding to the input ID 7 is one. Therefore, in the row of the input ID 7, two cells are unused. One cell has a bit length of 32 bits. In a hardware configuration in which memory must be secured even for unused cells, the memory size becomes larger.

  A multicast table for solving the above problems will be described with reference to FIG. FIG. 6 is a diagram showing the table structure of the multicast table according to the present embodiment.

  The multicast table includes four columns: an input ID (Input ID) column, a multicast number (multicast number) column, an identical ID (same ID) column, and a pointer (Pointer) column. The first column, the input ID column, is a column indicating the input ID of the received message. The second column, ie, the multicast number column, is a column indicating the multicast number of the corresponding input ID. That is, when a message is transmitted to the corresponding input ID, the number of output IDs designated as the transfer destination address is the number of multicasts.

  The same ID column, which is the third column, is a column indicating whether or not to transfer the message to the same ID as the corresponding input ID. For example, when the same ID as the input ID is the output ID, the value of the cell in the same ID column is the first value, for example, "1", and the same ID as the input ID is not the output ID. The value of the cell of the same ID column is the second value, for example, "0". Thus, depending on the value of the cell of the same ID column, it is determined whether or not to transfer the message to the same ID.

  The pointer column, which is the fourth column, is a column that indicates a row number so that a plurality of output IDs assigned to one input ID are sequentially designated one by one. That is, in each cell of the pointer column, a row number indicating an output ID associated with one input ID is stored as a pointer. In each column of the table, an ID, a pointer, the number of multicasts, and a value indicating whether or not the same ID is associated.

  Next, a procedure for generating a multicast table according to the present embodiment will be described. FIGS. 7 to 11 are diagrams for explaining the procedure for converting the multicast table (hereinafter, table T1) shown in FIG. 2 into the multicast table (hereinafter, table T2) according to the present embodiment. The table T2 is assigned line numbers 0 to 13. As described above, the tables T1 and T2 are multicast tables to be referred to for multicast transmission of a message in the gateway device or the like. Thus, the table T2 can be applied to gateway ID conversion.

  First, the input ID of the table T1 is set to the ID column of the table T2 (arrow A1 in FIG. 7). Since the table T1 is composed of five rows, the output ID is stored in the five cells from the top of the ID column of the table T2. That is, ID0, ID9, ID6, ID7, and ID5 are copied to cells of row numbers 0 to 4 in the ID column of the table T2, respectively.

  Next, the output ID of the table T1 is set to the ID column of the table T2 (arrow A2 in FIG. 7). Here, in the ID column of the table T2, the output ID is stored in order from the next cell in which the input ID is stored. Further, in the row in which the output ID is stored, the multicast number column and the same ID column become unused. Since five input IDs are stored in the table T1 as described above, the output IDs are stored in order from the cell of row number 5. Further, since eight output IDs are stored in the table T1, the output IDs are stored in the cells of row numbers 5 to 12 in the ID column of the table T2.

  Specifically, ID3, ID5, and ID7 are copied to cells of row numbers 5 to 7 in the ID column of the table T2 as output IDs associated with the input ID of ID0. ID8 is copied to the cell of row number 8 of the ID column of the table T2 as an output ID associated with the input ID of ID9. In cells of row numbers 9 and 10 in the ID column of the table T2, ID2 and ID5 are copied as output IDs associated with the input ID of ID6. In the row number 11 of the ID column of the table T2, ID0 is copied as an output ID associated with the input ID of ID7. In the cell of the row number 12 of the ID column of the table T2, ID6 is copied as an output ID associated with the input ID of ID5. As described above, one input ID or one output ID is stored in one cell of the ID column.

  Next, as shown in FIG. 8, the multicast number corresponding to each input ID is stored in the multicast number column of the table T2 (box A3 in FIG. 8). Here, in the table T1, the number of multicasts (the number of output IDs) corresponding to each input ID is input to cells of row numbers 0 to 4 in the column of the number of multicasts of the table T2. For example, in the table T1, since the number of output IDs associated with the input ID 0 is 3, the number of multicasts is 3 in the cell of the row number 0 of the table T2. Similarly, in the row numbers 1 to 4 of the table T2, the number of multicasts is 1, 2, 1, and 1, respectively. In addition, for the row in which the output ID is input, that is, for the row numbers 5 to 12, cells in the multicast number column are unused.

  Then, as shown in FIG. 9, the value is stored in the same ID column of the table T2 according to whether or not the same ID as the input ID is the output ID (frame A4 in FIG. 9). Here, in all input IDs, since the same ID as the input ID is not the output ID, all 0s are stored in the same ID column. For example, the output ID corresponding to the input ID0 of the table T1 does not include ID0. Therefore, in the row number 0 of the table T2, the value of the same ID column is 0. Similarly, in the row numbers 1 to 4 of the table T2, the value of the same ID column is zero. In the row in which the output ID is input, that is, the row numbers 5 to 12, cells of the same ID column are unused. By providing the same ID column, the same ID can be designated as the output destination address.

  Next, as shown in FIG. 10, a value indicating the row number of the ID column is stored in the pointer column of the table T2 (frame A5 in FIG. 10). In the pointer column, row numbers of ID columns are stored so that a plurality of output IDs assigned to one input ID are sequentially designated one by one. For example, in the table T1, as the output ID, ID3, ID5, and ID7 are associated with the input ID of ID0. Thus, in the cell of row number 0 in table T2, row number 5 specifying the first output ID (ID3) associated with the input ID of ID0 is input as a pointer (arrow A6 in FIG. 10).

  Next, line number 6 specifying the second output ID (ID5) associated with the input ID of ID0 is input to the cell of line number 5 of the pointer column of table T2 (arrow in FIG. 10) A7). Furthermore, in the cell of row number 6 in the pointer column of table T2, row number 7 specifying the third output ID (ID7) associated with the input ID of ID0 is input. In the ID column of table T2, the row number where the last output ID corresponding to one input ID is stored, for example, in row number 7, the cell of the pointer column is not used

  In this way, a pointer specifying the output ID corresponding to the input ID = ID0 is held in the table T2. The pointer input to the pointer column has a value indicating the output ID corresponding to one input ID in order. Similarly, when the output ID corresponding to the input ID of the table T1 = ID9, ID6, ID7, and ID5 is stored in the table T2, the multicast table according to the present embodiment is completed. A table T2 obtained by converting the table T1 into the table format according to the present embodiment is as shown in FIG. The table T2 shown in FIG. 11 is equivalent to the table T1.

  For example, when a message whose input ID is ID0 is received, the gateway device searches for cells having ID0 in order from the top of the ID column. Since the ID column of the cell of row number 0 is ID 0, the gateway apparatus refers to the multicast number column of row number 0, the same ID column, and the pointer column. Since the number of multicasts is three, the output ID of the transfer destination is three. Further, since the value of the same ID column is 0, it can be seen that no message is transferred to ID0. Since the pointer of the line number 0 is 5, the gateway device refers to the ID of the line number 5 and sets the ID 3 as the transfer destination of the message. Further, since the pointer of the line number 5 is 6, the gateway apparatus refers to the ID column of the line number 6 and sets the ID 5 as the transfer destination of the message. Furthermore, since the pointer of the line number 6 is 7, the gateway apparatus refers to the ID column of the line number 7 and sets the ID 7 as the transfer destination of the message. Since the number of IDs to which the message is to be transferred is 3 and the number of multicasts has been reached, the search process is ended. The gateway apparatus can search for all output IDs corresponding to the input ID by tracing the line numbers indicated by the pointer in order by the number of multicasts.

  In the table T2, even when the memory size is set in advance, an output ID corresponding to one input ID can be added. For example, as shown in FIG. 12, new lines may be added. Specifically, when it is desired to add ID 9 as an output ID to the input ID of ID 0, the ID is input to row number 13 of table T2. Furthermore, while updating the multicast number of row number 0 to 4, 13 specifying the row number is input to the cell of row number 7 of the pointer column.

  As described above, by converting the table T1 into the table T2, it is possible to add a message transfer destination. For example, even when the number of rows in the table is determined in advance, the output ID associated with the input ID can be increased by adding a row to the table T2. Therefore, the number of multicasts can be set freely. Therefore, versatility can be enhanced.

  Furthermore, memory cost can be suppressed by changing the bit length for each column and designing. For example, in the cell of the multicast number column, the same ID column, and the pointer column, the number of bits can be smaller than that of the ID cell. As a result, the table size can be reduced, and memory costs can be reduced.

Second Embodiment
In the second embodiment, the order of the ID columns is optimized so as to reduce the number of rows as compared with the first embodiment. FIG. 13 shows a table T2 and a table T3 in which the order of the ID columns is optimized. That is, in the table T2, the table T3 is obtained by optimizing the order of the rows in which the output ID is stored. Optimization can increase the number of unused lines. Therefore, the number of rows can be reduced, and the table size can be reduced.

  Specifically, two or more lines having a common ID are unified into one line. As a result, the duplication of rows is eliminated, and the number of rows can be reduced. For example, in the table T2, ID5 exists in the row numbers 4, 6, 10 of the ID column. Therefore, these three lines are unified into one line of line number 4. In the ID column of the table T3, only the row number 4 is ID5.

  Further, in the table T2, since ID0 is present in the row numbers 0 and 11, these two rows are unified to the row number 0. Similarly, in the table T2, since the ID 6 is present in the row numbers 2 and 12, these two rows are unified into one row of the row number 2. Thus, the number of lines can be reduced by unifying two or more lines whose IDs overlap with each other into one line. For example, while the row numbers 0 to 13 are used in the table T2, the row numbers 0 to 8 are used in the table T3.

  Since the number of rows can be reduced by performing the optimization, the table size can be reduced. In the table T3, since the number of multicasts is provided, it becomes possible to designate the same line number as the pointers which designate output IDs corresponding to different input IDs. If the pointers are different in the same line, the transfer order may be changed. For example, in the table T2, since the pointers are 12 and 7 in the row numbers 4 and 6 in which the IDs are common, the transfer order is optimized. The output ID of the row to be shared may be specified last by the pointer. Such optimization can be performed by a computer program.

  The table size of the table subjected to the above optimization will be described. FIG. 14 is a diagram showing the table T4 before the conversion process. That is, the table T4 is a table having no pointer column. FIG. 15 is a diagram showing the table T5 after the table conversion and optimization processing.

  In the table T4 of FIG. 14, a maximum of four output IDs are set for the input ID. In addition, five input IDs are set. Therefore, the number of cells in the table is 25 = (5 * 5). And the cell of each ID is 32 bits. Therefore, the table size of the table T4 is 800 bits (= 32 * 25).

  The table T5 includes an ID column, a multicast number column, an identical ID column, and a pointer column. Similar to the cells of the input ID and the output ID of the table T4, the cell of the ID column is 32 bits. The cell of the multicast number column, the cell of the same ID column, and the cell of the pointer are 3 bits, 1 bit and 4 bits, respectively. Further, the number of rows of the table T5 is nine. Therefore, the table size of the table T5 is (32 + 3 + 1 + 4) * 9 = 360 bits. Therefore, the table T5 equivalent to the table T4 can be created with a small table size.

The generalized table size will be described. The table size Size before conversion is expressed by the following equation (1).
Size = ID_size * input_ID_nbr * (1 + max_mul_nbr) (1)

  Here, ID_size is the bit length of the cell of the input ID column, input_ID_nbr is the total number of input IDs, and max_mul_nbr is the maximum value of the multicast number.

The table size worst_size after conversion is expressed by the following equation (2).
worst_size = (ID_size + ([log2 (max_mul_nbr)] + 1) +1+ [log2 (input_ID_nbr + output_ID_nbr)]) * (input_ID_nbr + output_ID_nbr) (2)

  Here, ID_size is the bit length of the cell of the ID column, max_mul_nbr is the maximum value of the multicast number, input_ID_nbr is the total number of input IDs, and output_ID_nbr is the total number of output IDs. The table size worst_size after conversion is a value before optimization, and duplicate IDs are considered. The table size can be reduced by performing the above optimization.

Third Embodiment
The multicast table according to the third embodiment will be described with reference to FIG. In the table T6 shown in FIG. 16, the same ID column is omitted. The second embodiment is the same as the first embodiment except that the same ID column is omitted. Even when the table T6 is used, by following the pointer, the output ID associated with the input ID can be set as the transfer destination of the message. Therefore, the same effects as in the first and second embodiments can be obtained.

Fourth Embodiment
The multicast table according to the fourth embodiment will be described with reference to FIG. In the table T7 shown in FIG. 17, a pointer valid column is provided instead of the multicast number column. The pointer valid column is associated with the ID column. Then, when the cell of the pointer valid column has a first value, eg, 1, the ID of the row number indicated by the pointer is taken as the output ID to be the transfer destination. On the other hand, when the cell of the pointer valid column has a second value, for example, 0, designation of the next output ID is ended. Thus, the pointer valid column stores a value indicating whether to stop designation of the next output ID. Since the optimization shown in the second embodiment is difficult in the table T7 of the fourth embodiment, it is suitable for a table with little duplication.

Embodiment 5
The multicast table according to the fifth embodiment will be described with reference to FIG. In the table T7 shown in FIG. 18, a hash control information (Hash Control Info) column and a hash pointer (Hash Pointer) column are added to the table T2.

  Hash control information associated with the input ID is stored in the hash control information column. The hash control information includes, for example, a hash value. The hash pointer column stores a hash pointer for correcting an input ID which is erroneously searched when a hash collision occurs. That is, the hash pointer is information for correcting the designation in the multicast pointer (corresponding to the pointer column of the table T2) when the hash collides.

  In the present embodiment, hash control information is used to search for an input ID. For example, when a message is received, a hash value that matches the received message data hash value is searched from the hash control information column. The search time can be shortened by searching using hash control information. Then, when the input ID corresponding to the searched hash value matches the input ID of the message, the output ID is searched as in the first embodiment and the like. That is, the output ID corresponding to the input ID is searched by tracing the pointer in order.

  On the other hand, when a hash collision occurs, the input ID corresponding to the searched hash value does not match the input ID of the message. In this case, the correct input ID is corrected by referring to the hash pointer. Then, from the corrected input ID, the output ID corresponding to the input ID is searched by tracing the line numbers designated by the pointer in order.

  A specific example of the table according to the present embodiment is shown in FIG. FIG. 19 is a diagram showing a table T9 having hash control information and a hash pointer column, and is a table corresponding to the table T4 of FIG. The table T9 includes an ID column, a multicast pointer column, an identical ID column, a multicast number column, a hash control information column, and a hash pointer column. The multicast pointer column corresponds to the pointer column of the table T2. A table provided with such hash control information and the like can be created by the same method as the method described in the first embodiment.

  Bit length of ID column, multicast pointer column, same ID column, multicast number column, hash control information column, and hash pointer column is 32 bits, 4 bits, 1 bit, 3 bits, 3 bits, 4 bits respectively ing. Since the table T9 includes nine rows of row numbers 0 to 8, the table size is 423 (= 47 * 9) bits.

  Furthermore, in the present embodiment, it is possible to divide the input ID and the output ID into separate tables. FIG. 20 is a diagram showing an example in which the input ID and the output ID are divided into different tables. T8 shown in FIG. 18 is divided into two tables T10 and T11. In FIG. 20, two tables, an input ID table T10 and an output ID table T11, are shown.

  The input ID table T10 has the same columns as the table T7 of FIG. On the other hand, the output ID table T11 comprises only an output ID column and a multicast pointer column. In the output ID table T11, the hash control information and the hash pointer are not used, so the hash control information column and the hash pointer column can be omitted. This makes it possible to reduce the table size as a whole.

  FIG. 21 is a diagram showing a specific example in which the input ID and the output ID are divided into different tables. FIG. 21 shows an input ID table T12 and an output ID table T13. FIG. 21 shows the table T9 shown in FIG. 19 divided into two tables. Such an input ID table T12 and an output ID table T13 can be created by the same method as that described in the first embodiment.

  When the input ID is searched in the input ID table T12, the gateway device refers to the multicast pointer corresponding to the input ID and designates the row number of the output ID table T13. Thereby, a line number to be an output ID corresponding to the input ID is designated. Then, in the same manner as described above, the gateway device refers to the pointer column of the output ID table T13, and sequentially designates the line numbers of the output ID corresponding to the input ID. Thus, all output IDs corresponding to one input ID can be set as the transfer destination of the message.

  The input table T12 shown in FIG. 21 includes a 32-bit input ID column, a 2-bit multicast pointer column, a 1-bit identical ID column, a 3-bit multicast number column, a 3-bit hash control information column, and a 3-bit hash. It has a pointer column. The number of rows of the input ID table T12 is five. Therefore, the table size of the input ID table T12 is 220 (= 44 * 5) bits.

  The output ID table T13 shown in FIG. 21 includes a 32-bit output ID column and a 2-bit multicast pointer column. The number of rows of the output ID table T13 is four. Therefore, the table size of the output ID table T13 is 136 (= 34 * 4) bits. Therefore, the total table size of the input ID table T12 and the output ID table T13 is 356 bits.

  The input ID and the output ID can be divided into different tables not only for this embodiment but also for the tables T2 to T7 shown in the first to fourth embodiments. In this case, in the output ID table, the same ID column and the multicast number column can be omitted.

  Furthermore, the input ID and the output ID are divided into separate tables, and the respective tables are stored in separate memories. A first memory holds an input ID table, and a second memory different from the first memory holds an output ID table. By doing this, the search for the input ID and the tracking of the output ID can be performed in parallel.

  For example, the case where two messages are received consecutively will be described. The gateway device searches the input ID for the first message in the input ID table, and then tracks the output ID using the pointer in the output ID table. The gateway device searches the input ID for the second message while tracking the output ID in the output ID table. That is, in the input ID table, since the search for the input ID of the first message is completed, the input ID of the second message can be searched. As described above, by storing two tables in different memories, the processing can be speeded up by performing the search of the input ID and the tracking of the output ID in parallel.

  In the input ID table, the input ID may be searched by binary search, or the input ID may be searched by linear search. In binary search, input IDs are sorted in descending or ascending order, and the search is started from the median value. In linear search, input IDs are searched in order from the top. This makes it possible to search at high speed.

Sixth Embodiment
The multicast table according to the present embodiment will be described with reference to FIG. FIG. 22 shows the table T14. In the present embodiment, routing information is added to the table T14. Except for the routing information column, the table is the same as the table T2, so the description is omitted.

  The routing information includes destination information such as contacts. In addition, in place of the routing information or in addition to the routing information, other meta information such as a filtering rule may be added to the table T14. Various meta information can be added by providing a meta information column. Thereby, versatility can be improved.

  The table converted by the above table conversion method is suitable for the multicast table of the on-vehicle gateway. In the on-vehicle system as shown in FIG. 1, it is required to transmit a message without delay while being restricted by memory capacity, CPU speed, power consumption and the like. In order to reduce the latency, it is desirable to configure the gateway shown in FIG. 1 by hardware. Therefore, as described in the present embodiment, by using a multicast table having a pointer, it is possible to rapidly multicast a message.

  In addition, when the gateway is configured by hardware, the table size and table format may not be able to be changed. Specifically, it is not possible to add a column to the table or to change the bit length of the column. Therefore, the transfer destination can be easily added by using the table converted by the above table conversion method.

(Table converter)
A table conversion apparatus for creating the tables shown in the first to sixth embodiments will be described with reference to FIG. FIG. 23 is a block diagram showing the table conversion apparatus 300. The table conversion device 300 includes a processor 301 and a memory 302.

  The memory 302 holds a table conversion program for performing table conversion according to the procedure shown in FIGS. The processor 301 is a CPU or the like and executes a program. When a pre-conversion table, for example, a table T1 is input, the processor 301 executes a program stored in the memory 302. By doing this, the table T1 is converted to the table T2 according to the procedure shown in FIGS. This makes it possible to convert to a highly versatile table.

  Furthermore, the memory 302 may hold a program for performing the optimization process as described in the second embodiment. Then, the processor executes the program to optimize the table. Thus, the table size can be reduced.

  FIG. 24 shows a gateway apparatus 400 that performs multicast transmission using the table 403. The gateway device 400 is connected to the plurality of subsystems 401 and 402. The gateway device 400 communicates with the subsystems 401 and 402 according to a protocol such as CAN or Ethernet. The gateway device 400 relays communication between different networks. The gateway device 400 is applicable to an in-vehicle system and the like.

  The gateway device 400 is configured by hardware. The gateway device 400 includes a circuit for searching for an input ID, and a circuit for specifying an output ID with reference to a pointer. Then, as described above, the gateway device 400 performs ID conversion by referring to the table 403. This makes it possible to perform multicast transmission of messages between multiple subsystems.

  Furthermore, the above table can be used for applications other than the ID conversion of the gateway device 400. For example, the table can be used to convert input values into multiple output values.

Although a part or all of the above-mentioned embodiment may be described as the following supplementary notes, it is not limited to the following.
Appendix (Supplementary Note 1)
A table conversion apparatus for converting a table in which a plurality of output values are assigned to one input value, the table conversion apparatus comprising:
A memory for holding the table;
The table conversion device is
Let the input value be an input value column,
Let a plurality of output values assigned to the input value be an output value column,
A table conversion apparatus, wherein a row number of an output value column is used as a pointer column so that a plurality of output values assigned to the input value are sequentially designated one by one.
(Supplementary Note 2)
The table conversion device according to appendix 1, wherein a value indicating whether or not the same value as the input value is assigned as the output value of the input value is a same value column.
(Supplementary Note 3)
The table conversion device according to any one of Appendices 1 or 2, wherein the number of output values corresponding to each of the input values is used as an output value number column.
(Supplementary Note 4)
The table conversion device according to appendix 3, wherein the order of the rows of the output value column is optimized to reduce the number of rows.
(Supplementary Note 5)
The table conversion device according to any one of appendices 1 to 4, wherein the input value column and the output value column are one column as an input / output value column.
(Supplementary Note 6)
The input value is an input ID, and the output value is an output ID indicating an output destination,
The table conversion device according to any one of appendices 1 to 5, wherein the table is a multicast table used for multicast transmission.
(Appendix 7)
The table conversion device according to claim 6, wherein meta information associated with the output ID is a meta information column.
(Supplementary Note 8)
Let hash control information associated with the input ID be a hash control information column,
6. The table conversion apparatus according to claim 6, wherein a pointer for correcting designation by the pointer column is a hash pointer column when a hash collision occurs.
(Appendix 9)
The input value column and the output value column are divided into separate tables,
The table conversion device according to any one of appendices 1 to 8, wherein in the first table having the input value column, the pointer column designates a row number of the second table having the output value column.
(Supplementary Note 10)
Storing the first table in a first memory;
10. The table conversion apparatus according to appendix 9, wherein the second table is stored in a second memory capable of tracking in a row number indicated by the pointer column in parallel with the search of the input value in the first memory.
(Supplementary Note 11)
The table conversion device according to appendix 9, wherein a binary search or a linear search is performed in the first table.
(Supplementary Note 12)
A table conversion method for converting a table in which a plurality of output values are assigned to one input value,
Let the input value be an input value column,
Let a plurality of output values assigned to the input value be an output value column,
A table conversion method, wherein a row number of an output value column is used as a pointer column so that a plurality of output values assigned to the input value are sequentially designated one by one.
(Supplementary Note 13)
15. The table conversion method according to appendix 12, wherein a value indicating whether or not the same value as the input value is assigned as the output value of the input value is a same value column.
(Supplementary Note 14)
15. The table conversion method according to appendix 12 or 13, wherein the number of output values corresponding to each of the input values is used as an output value number column.
(Supplementary Note 15)
15. The table conversion method according to appendix 13, wherein the order of the rows of the output value column is optimized to reduce the number of rows.
(Supplementary Note 16)
The table conversion method according to any one of Appendices 11 to 15, wherein the input value column and the output value column are one column as an input / output value column.
(Supplementary Note 17)
The input value is an input ID, and the output value is an output ID indicating an output destination,
The table conversion method according to any one of Appendices 11 to 16, wherein the table is a multicast table used for multicast transmission.
(Appendix 18)
24. The table conversion method according to appendix 16 or 17, wherein meta information associated with the output ID is a meta information column.
(Appendix 19)
Let hash control information associated with the input ID be a hash control information column,
The table conversion method according to any one of appendices 16 to 18, wherein a pointer for correcting designation by the pointer column is a hash pointer column when a hash collision occurs.
(Supplementary Note 20)
The input value column and the output value column are divided into separate tables,
The table conversion method according to any one of Appendices 11 to 19, wherein, in the first table having the input value column, the pointer column designates a row number of the second table having the output value column.
(Supplementary Note 21)
Storing the first table in a first memory;
20. The table conversion method according to appendix 19, wherein the second table is stored in a second memory capable of tracking in a row number indicated by the pointer column in parallel with the search of the input value in the first memory.
(Supplementary Note 22)
The table conversion device according to any one of appendices 21 to 21, wherein in the first table, a binary search or a linear search is performed.
(Supplementary Note 23)
The gateway apparatus which performs ID conversion using the table converted by the table conversion method as described in any one of supplementary notes 11-22.

  As mentioned above, although the invention made by the present inventor was concretely explained based on an embodiment, the present invention is not limited to the embodiment mentioned already, A various change in the range which does not deviate from the gist It goes without saying that it is possible.

100 system 200 in-vehicle system 201 transfer unit 202 gateway 204 multicast table 211 CPU
212 CPU
213 CPU
214 CPU

Claims (20)

A table conversion apparatus for converting a table in which a plurality of output values are assigned to one input value, the table conversion apparatus comprising:
A memory for holding the table;
The table conversion device is
Let the input value be an input value column,
Let a plurality of output values assigned to the input value be an output value column,
A table conversion apparatus, wherein a row number of an output value column is used as a pointer column so that a plurality of output values assigned to the input value are sequentially designated one by one.   The table conversion device according to claim 1, wherein a value indicating whether or not the same value as the input value is assigned as the output value of the input value is a same value column.   The table conversion apparatus according to claim 1, wherein the number of the output values corresponding to each of the input values is set as an output value number column.   4. The table conversion apparatus according to claim 3, wherein the order of the rows of the output value column is optimized so as to reduce the number of rows.   The table conversion apparatus according to claim 1, wherein the input value column and the output value column are one column as an input / output value column. The input value is an input ID, and the output value is an output ID indicating an output destination,
The table conversion apparatus according to claim 1, wherein the table is a multicast table used for multicast transmission.   7. The table conversion apparatus according to claim 6, wherein the meta information associated with the output ID is a meta information column. Let hash control information associated with the input ID be a hash control information column,
7. The table conversion apparatus according to claim 6, wherein a pointer that corrects designation by the pointer column is a hash pointer column when a hash collision occurs. The input value column and the output value column are divided into separate tables,
The table conversion apparatus according to claim 1, wherein, in the first table having the input value column, the pointer column designates a row number of the second table having the output value column. Storing the first table in a first memory;
10. The table conversion apparatus according to claim 9, wherein the second table is stored in a second memory capable of tracking in a row number indicated by the pointer column in parallel with the search for an input value in the first memory. . A table conversion method for converting a table in which a plurality of output values are assigned to one input value,
Let the input value be an input value column,
Let a plurality of output values assigned to the input value be an output value column,
A table conversion method, wherein a row number of an output value column is used as a pointer column so that a plurality of output values assigned to the input value are sequentially designated one by one.   The table conversion method according to claim 11, wherein a value indicating whether or not the same value as the input value is assigned as the output value of the input value is a same value column.   The table conversion method according to claim 11, wherein the number of the output values corresponding to each of the input values is set as an output value number column.   The table conversion method according to claim 13, wherein the order of the rows of the output value column is optimized to reduce the number of rows.   The table conversion method according to claim 11, wherein the input value column and the output value column are one column as an input / output value column. The input value is an input ID, and the output value is an output ID indicating an output destination,
The table conversion method according to claim 11, wherein the table is a multicast table used for multicast transmission. Let hash control information associated with the input ID be a hash control information column,
17. The table conversion method according to claim 16, wherein a pointer that corrects designation by the pointer column is a hash pointer column when a hash collision occurs. The input value column and the output value column are divided into separate tables,
The table conversion method according to claim 11, wherein, in the first table having the input value column, the pointer column designates a row number of the second table having the output value column. Storing the first table in a first memory;
The table conversion method according to claim 18, wherein the second table is stored in a second memory capable of tracking in a row number indicated by the pointer column in parallel with the search of the input value in the first memory. .   The gateway apparatus which performs ID conversion using the table converted by the table conversion method of Claim 11.

Images