WO2000076132A1 - Performance management calculating circuit, block length judging circuit, and statistics calculating circuit - Google Patents

Abstract

A performance management calculating circuit (9) comprises an element calculating unit (10) provided with a calculation data selectively outputting section (11) for selectively outputting calculation data according to an external signal, a performance management calculating section (12) for performing performance management calculation by using the outputted calculation data, and selectively outputting performance evaluation results according to an external signal, and a control section (13) for receiving a user cell and feeding a scheduling signal to the calculation data selectively outputting section (11) and the performance management calculating section (12). Therefore, the performance management calculating section (14) is shared, and the circuit scale and the power consumption are reduced.

Description

 Description Performance management arithmetic circuit, block length judgment circuit, and editing arithmetic circuit

 The present invention relates to a performance management arithmetic circuit, a block length determining circuit, and a finite arithmetic circuit suitable for use in, for example, an ATM communication system. Background art

 A TMCA Synchronous Transfer Mode) communication system has a function of transmitting and receiving user cells, and a function of measuring the quality of the above-mentioned user cells, such as the user's iron cells, erroneous insertion of user cells, and user cell errors. This function is implemented as a 0 AM (Operation, Administration and Maintenance) function, and one of the 0 AM functions is a performance management function.

 This performance management is a function that performs S ^ of a user cell after selection, erroneous insertion of a user cell, or user cell error. This performance management function is implemented by allowing one terminal to insert a performance monitoring cell having a dragon value, such as a steel cell, between user terminals while transmitting it to the other terminal. Is realized. There are two methods for calculating the value for this purpose: performance management calculation (intermediate calculation) and compilation calculation.

 Here, the performance management operation is an operation when a user cell is received, and is reflected in the payload of the performance monitoring cell to be transmitted. The compilation operation is an operation when a performance monitoring cell is received, and is used to manage the performance of the ATM communication system.

In other words, 0 AM is a function for monitoring the normality and performance of the user information transfer function. To realize this function, the maintenance information is added to the transmission side information and the reception is performed. This is done by analyzing its performance on the part. Specifically, the performance monitoring cell is used to evaluate the performance of the selected VP and VC, such as user cell loss, user cell erroneous insertion, and user cell error in the payload. FIG. 17 is a sequence diagram for conceptually explaining the performance management method. The data transmitted by the terminal 60a shown in FIG. 17 is received by the terminal 60e via the ATM switch 60b, the VP cross-connect 60, and the ATM switch 60d. Here, in performance monitoring using these, the performance monitoring cells are automatically transmitted every time the transmitting side transmits a certain number of user cells called block length (block size). .

 That is, the ATM switch 6 Ob power, user cell, and performance monitoring senor shown in Fig. 17 are transmitted (step 61a), and the ATM switch 60d receives the cells (step 6lb).

 Here, as shown in a box 62, after the performance monitoring cell 62b is transmitted, the user cells of the block size are transmitted, and then the performance monitoring cell 62a is transmitted strongly. I have. The block size is a value that can be set between the terminals 60a and 60e, and is a power of 2, for example, 65536. In addition, the performance monitoring cells 62a and 62b map (store) the number of user cells measured from the performance monitoring cell inserted last time, the BIP calculation result, and the like. The BIP calculation result is the result obtained by calculating the bit error value of the payload.

 On the other hand, in the receiving-side ATM switch 60d, the performance management operation and the statistical operation using the received user cell and the performance monitoring cell are performed, and the erroneous insertion of the user cell, ^ of the user cell, and the BIP calculation result are performed. The error is detected, and the performance monitoring cell 61d in which the result of the performance Hffi is stored is transmitted to the ATM switch 60d, and the ATM switch 6 Ob is notified (step 6 1 c). Further, the ATM switch 6 Ob performs quality control based on the performance degradation result included in the notification (step 61e). After that, the next user cell group is transmitted and received between the terminals 60a and 60e.

 Next, the format of the performance monitoring cell and the meaning of each area are listed. FIG. 18 is a diagram showing the format of the performance monitoring cell. In FIG. 18, the horizontal direction indicates bits 0 to 7, and the vertical direction indicates octet numbers.

The VP I (Virtual Path Identifier) and VC I (Virtual Chann el Identifier) and PTI (Payload Type Identifier) indicate the level of the OAM cell, respectively. Here, 16 bits are allocated to the VC I, and among the possible values 0 to 65535, 32 types 0 to 31 are specially set among the numbers reserved for the system. If this number is set as a special number, system calculations are performed.

 Fig. 19 shows the possible values of VP I, VC I, and PTI. For example, if VCI in FIG. 19 is 3, it indicates that the connection of the receiving cell is VP, and the force, the point, and the point are intermediate. Connections and points will be described later.

 In FIG. 18, “CLP Cell Loss Priority” stores the superior fe † green information. That is, when congestion occurs in the network, this area is discarded from unimportant cells so that important cells can be rescued preferentially. In case of conflict, the value is set to 0.

 HEC (Header Error Control) is for error correction, OAMType represents 0 AM type, and FUNCType represents function type. The MCSN (Monitoring Cell Sequence Number) is stored as a sequential number for detecting loss or erroneous insertion of the performance monitor user cell. The value is represented by modulo 256.

 Further, TUC-0 + 1 indicates transmission 1 of all user cells of CLPs 0 and 1, and is represented by modulo 256. This ^? _ 0 means a value for CLP 0, and 0 +1 means a value for both CLP 0 and CLP 1. That is, with this area, the loss of the user cell or the erroneous insertion of the user cell is detected for all the user cells including both the good CLP 0 and the poor CLP 1. The use of such subscripts will be used in the same manner in the following description.

BEDC-0 + 1 is a 16-bit even parity (BIP-16) for detecting a bit error in the information area of the user cell, and includes all bits after the performance monitoring cell transmitted immediately before. BIP—16 total for ^ [blue information area of user cell The result of the calculation is stored.

 Further, TUC-0 indicates the transmission of the user senor of CLP 0, if 、, and is represented by the module port 65536. That is, as in the case of TUC-0 + 1, this region allows the user cell loss or user cell erroneous insertion to be detected for “!” And CLP0. The values stored in TUC-0 + 1 and TUC-0 are each incremented by 1 for each cell, and this value increases even if the block size is exceeded.

 Also, TSTP stores the time at which the performance monitoring cell whose usage is unspecified is inserted. TRCC-0 indicates the number of cells of CLP 0 received at the receiving side to the transmitting side. This is an area for notification, and the transmitting side can detect the loss of the user cell or the erroneous insertion of the user cell by this correction. And B L E R— 0

+1 is an area for notifying the number of errors detected on the receiving side to the transmitting side by BEDC-0 + 1, and TRCC-0 + 1 is all user cells of CLP 0 and 1 received on the receiving side. This area is used to notify the number to the transmitting side, and the transmitting side can detect the loss of the user cell or the erroneous insertion of the user cell based on this value. In CRC-10, a parity bit subjected to CRC calculation is inserted.

 Using these values, the performance management calculation and the “computation calculation” are performed. In order to calculate these values, there is a problem in that the resources (circuit size, power consumption) used become large.

 FIG. 20 is a functional block diagram of a conventional performance management arithmetic circuit. The performance management calculation circuit 66 shown in FIG. 20 performs performance management calculation. The performance management calculation circuit 66 includes a VC calculation unit 62a, an NVP calculation unit 62b, and a UVP calculation unit 62c. It is composed of 63 a, 63 b, 63 c, arithmetic and control units 64 a, 64 b, 64 c, and cell type determination means 65.

Here, the VC, NVP, and UVP connection-specific balance data are calculated in the performance management difficulties 63a, 63b, and 63c, respectively, and the results are respectively calculated and calculated. 64a, 64b, and 64c are stored in the calculation result holding units 64a, 64b, and 64c. According to the cell type, the data is stored.

 In this performance management calculation, as many as 30 types of performance management performances are required for the received user cells. In other words, for each connection (UVP, NVP, VC) and point (between End-to-End and SEGMENT) of the received user cell, a plurality of operations (MCSN, TUC 0, TUC 0/1, BEDC, transmission) (Measurement of the number of cells), so 30 (3 X 2 X 5) types of performance management difficulties are required. In addition, the plurality of operations are to calculate the values of the above MCSN, TUC O, TUC 0/1, and BED C, and to measure the number of transmitted user cells (measure the number of transmitted cells). Means that.

 Next, the differences between VC, NVP, and UVP connections will be briefly described. As is well known, a path is described by a VC that identifies a channel (circuit) and a VP that identifies a path. In other words, VC is for translating a terminal through a network, and VP is for identifying an ATM switch. Then, one ATM exchange sees the senor VC I coming from the other ATM exchange via the network and handles it to the corresponding terminal.

 Therefore, VC represents a VC connection accommodated in a VP, NVP represents a VP connection that bundles a plurality of VC connections, and UVP represents a VP connection that does not perform VC handling. It is open to users.

 FIG. 21 is a diagram showing a configuration of a conventional performance window calculation unit. The performance management function 167 shown in FIG. 21 performs a performance management calculation for each point, and includes a SEGMENT T value storage 68 a, an End value storage 68 b, and caro parts 69 a, 69 b. , SEGMEN T connection ¾ ^ part 70a, End consolidation ¾¾70b.

Here, the SEGMENT value holding unit 68a performs measurement of SEG-MCSN, SEG-TUC 0, SEG-TUC 0/1, SEG-BED C, SEG—the number of transmitted cells when the point is set between Nakahiro. This is a memory that holds values. In addition, the end value holding unit 68b is used to measure the EE—MCSN, EE-TUCO, EE-TUC 0/1, EE—BED EE—measured value of the number of transmitted cells when the point force ends. This is the memory that holds Further, the adder section 69a actually uses the data held in the SEGMENT value holding section 68a to actually execute SEG-MCSN, SEG-TUC 0, SEG- TUC 0/1, SEG-BEDC, SEG- It performs each calculation of the number of cells to be sent out, and consists of five types of calories. Then, the difficulty part 69 b actually uses the data held in the End comfort 68 b to actually execute EE-MCSN, EE-TUC 0, EE-TUC 0/1, EE-BEDC, EE It is used to calculate numbers and consists of five types of adders.

 Further, the SEGMENT result (呆 0a stores the operation result in the adder unit 69a, and the End result holding unit 70b stores the operation result in the adder unit 69b. Each of these functions is provided by five types of memory.

 Thus, for each of the two types, neutral and end point, 10 additional operations are required to perform the operation of 5.

 In addition, in FIGS. 20 and 21, instead of performing one operation on one cell, two operations may be performed depending on the cell type, or no operation may be performed. Similarly, when the performance ^ S operation item 力 is set, the difficulty also increases, and the circuit test and the power consumption increase. However, there is a problem that the amount of memory becomes enormous because there are connections for each connection.

 Next, a counter of the number of reception user cells when performing the above-described performance management calculation will be described with reference to FIG.

 FIG. 22 is a functional block diagram of a conventional block size determination circuit. The block size determination circuit 71 shown in FIG. 22 determines whether the number of received user cells has reached the block size or not, and includes a cell transmission number counting section 71a and a block size setting section 71b. , Match detection section 71cF-PM cell transmission request section 71d.

Here, the cell transmission number counting section 71a counts the number of user cells transmitted from the transmission side, and is specifically composed of a counter. The block size setting section 7 lb holds the set value of the block size. And- The output unit 71c compares the value of the cell transmission number counting unit 71a with the value of the block size setting unit 71b, and when the values match, inputs a match detection signal to the F—PM cell transmission requesting unit 71d. Is what you do. The F-PM seno I output request unit 71d requests transmission of a performance monitoring cell based on the match detection signal from the match detection unit 71c.

 When the block size determination circuit 71 is used to transmit and receive cells separately for the path of を (meaning the path determined by VCI and VPI), it is determined whether the receiving user cell has reached the block size for each path. It needs to be measured (or counted). That is, for example, when there are 4096 paths and the block sizes are 1 024 (2 to the 10th power), the memory for 4096 X 10-bit memory is used, and the operator for processing these Memory is required. Note that this operator corresponds to a circuit element for realizing an adder and a multiplier, and is used in the same meaning in the following description. Therefore, the cell transmission number counting section 71a has a very large scale and has a problem that the power consumption increases.

 Next, the knitting calculation performed when the performance monitoring cell is received will be described.

 In the computation, the computation is performed using two types of values, TUCD0 and TUCD0 Z1, from the received performance monitoring cell. In other words, one is to find the difference (TUCD0) between the number of cells scheduled to be transmitted and the number of cells actually transmitted for CLP 0 (highly superior cells) and determine the +/- sign. The other is the difference between the number of cells scheduled to be transmitted and the number of cells actually transmitted (TUCD0 / 1) for CLP 0Z1 (all cells including cells with high priority and cells without priority). ), And the sign of +/− is semi-IJ-determined. For example, for CLP0, if the number of cells scheduled to be transmitted is 100 and the number of cells actually transmitted is 98, TUCD0 is 12. Similarly, if the number of cells to be transmitted is 100 and the number of cells actually transmitted is 101, TUCD0 is +1.

Then, based on the determination result of +/−, it is determined whether the threshold value is i: or less than Mlost (cell loss) and Mmisinserted (erroneous cell insertion). For example, if Mlost is set to -2, TUCD0 -2 is If Mmisinserted is set to +2, and TUCD 0 is +1 when the threshold setting value is matched, the threshold setting value is full.

 After that, using the above judgments, SECB Lost (cell fl ^ above threshold value), SECB Misinserted (error insertion above threshold), CLP0 + 1 misinsertion (threshold CLP0 + 1 number of erroneous insertions, CLPO + 1 Loss number (CLP0 + 1 cell loss number of thresholds), Total CLPO + 1 Loss number (cell loss of CLPO + 1 cells, CLPO Loss number (threshold The calculation is performed on all of the CLP0 cell (the number of irons) and the total number of CLP0 Loss (the number of CLPO cell losses).

 FIG. 23 is a functional block diagram of a conventional editing circuit. The statistical operation circuit 72 shown in FIG. 23 performs the tenth operation, and specifically, the seven types of tenth operation results ( It outputs Total CLPO + 1 Loss, CLPO + 1 Loss, SECB Lost, CLPO + 1 Misinsertion, SECB Misinserted, Total CLPO Loss, CLPO Loss. The meanings of these operations are the same as those described above, and the description thereof is omitted.

 The edit operation circuit 72 includes the above-described edit operation result output section 77, a memory group 73 for storing edit information for performing seven types of edit operations, and A memory group 7 4 that stores threshold setting values for performing the same kind of calculation, and the difference between the number of received user cells and the number of transmitted user cells for performing 7 @ ϋ And a calculation unit 76 that actually performs the calculation.

 The memory group 73 stores the editing information i¾f, and has memories 73a, 73b, 73c, 73d, 73e, 73f, 73g. . The memory group 74 has a threshold setting ^ !, and has memories 74a, 74b, and 74c. Further, the memory group 75 holds the difference between the number of received user cells and the number of transmitted user cells, and has memories 75a and 75b. The calculation unit 76 performs ten operations, and includes seven types of difficulty.

In this case, as in the case of the performance management operation (intermediate operation), the memory capacity for holding individual values is necessary, so the circuit scale and power consumption increase, and resources become more powerful. If not used, there is a ^ ヽ ぅ challenge.

 In addition, when statistical calculation items are introduced, the number of performances must be increased in the same manner, and there is a problem that circuit power and power consumption are reduced.

 As described above, in the above-described performance management calculation, block size determination calculation, and reduction calculation, the circuit scale and power consumption are large and are not used for resource power, and the performance management calculation items and statistical calculations are not used. There is a problem that the difficulty must be increased when the item power is {E}.

 The present invention has been made in view of such a problem, and performs pre-processing for determining the connection or the point of a received user cell, and then performs an operation by scheduling using a count value synchronized with the head of the cell. The purpose of the present invention is to provide a performance management arithmetic circuit capable of improving the circuit scale and power consumption by sharing the performance management performance by performing the above, and capable of coping with an increase in the number of calculation items. I do. Also, the present invention utilizes the fact that the number of user cells transmitted and received above is a power of two, so that a block length that does not require a memory area for measuring the number of user cells during connection is sufficient. It is an object to provide a determination circuit. In addition, the present invention can reduce the circuit scale and power consumption by sharing multiple memories by performing multiple evaluation operations by scheduling using the counter value synchronized with the cell head, It is an object of the present invention to provide a knitting operation circuit capable of coping with an increase in the number of calculation items. Disclosure of the invention

 Therefore, the performance management calculation circuit of the present invention performs a performance management calculation for Nffi quality of β¾ for each connection path or for each termination point and each of the intermediate identification information, and outputs a plurality of performance results. An operation circuit, which is capable of calculating at least one performance management operation relating to a connection path and a performance management operation relating to a termination point and a neutral identification information within one cell processing time to calculate a performance result of a diploid class. It is characterized by having a calculation unit.

Therefore, in this way, by sharing the operator memory, the circuit scale and power consumption can be reduced, and resources can be used effectively. Also, the operation Even when the number increases, it becomes possible to respond by changing the operation schedule, and therefore, there is an advantage that the increase in power consumption and the increase in the circuit scale accompanying the change can be suppressed.

 The block length determination circuit of the present invention further comprises: a block length setting unit that sets the number of user cells transmitted and received above, the block length indicating the number of user cells that is a power of 2, and the number of user cells to be received. The number of transmitted cells that can measure and display the value in multiple bits, and the lower bit extraction that reads the bit at ^ in the stored number of transmitted cells from the setting value set by the block length setting unit And a match detector that can detect that the number of user cells transmitted from one path matches the block length based on the result read by the lower bit extractor. It is a sign.

 Therefore, in this way, there is an advantage that a memory area for measuring the number of user cells can be used in the connection, the circuit scale and power consumption can be reduced, and resources can be used for the crane.

 The editing circuit of the present invention is a editing circuit for receiving a performance monitoring cell having information on the quality of the signal and outputting a plurality of values for the transmission path. A calculation unit capable of selectively outputting 1¾5 values, a value of the number of user cell losses, a value of the number of user cells incorrectly inserted, a deviation value, and a value of the total number of user cells. A control unit capable of inputting a scheduling signal to the editing unit using the value of the value of the user senor with built-in furnace and the value of the deviation, and the transmission path within one cell processing time It is characterized in that the evaluation value of の for is calculated in a time-sharing manner.

Therefore, in this way, by sharing a memory for calculating a plurality of values, the circuit scale and power consumption can be reduced, and resources can be used effectively. In addition, even if the number of operations increases, it is possible to respond by changing the operation schedule, so that there is an IJ point that can suppress an increase in power consumption and an increase in circuit scale accompanying the change.説明 brief description FIG. 1 is a functional block diagram of the performance management arithmetic circuit according to the first embodiment of the present invention.

 FIG. 2 is a diagram showing a time chart for explaining a time division method for each connection according to the first embodiment of the present invention.

 FIG. 3 is a detailed block diagram of the performance management calculation unit according to the first embodiment of the present invention.

 FIG. 4 is a diagram showing a time chart for explaining a time-division method for each point time-divided for each connection according to the first embodiment of the present invention.

 FIG. 5 is a root diagram for explaining sharing of performance management difficulties for each connection according to the first embodiment of the present invention.

 FIG. 6 is a function block diagram of another performance management calculation unit according to the first embodiment of the present invention.

 FIG. 7 is a diagram showing a time chart for explaining a time division method for each action S1 when time division is performed for each connection and for a point according to the first embodiment of the present invention.

 FIG. 8 is a diagram for explaining sharing of the performance management calculation unit for each connection and each point according to the first embodiment of the present invention.

 FIG. 9A is a functional block diagram of the block size determination circuit according to the first embodiment of the present invention.

 FIG. 9 (b) is an explanatory diagram of a method of reading the lower bits of the block size determination circuit according to the first embodiment of the present invention.

 FIG. 9C is an explanatory diagram of a second method for reading the lower bits of the block size determination circuit according to the first embodiment of the present invention.

 FIG. 10 is a flowchart of the performance management operation according to the first embodiment of the present invention. FIG. 11 is a flowchart of measurement calculation according to the first embodiment of the present invention. FIG. 12 is a flowchart of the calculation processing of the performance evaluation result according to the first embodiment of the present invention.

FIG. 13 is a functional block diagram of a statistical operation circuit according to the second embodiment of the present invention. FIG. 14 is a more detailed functional block diagram of the control unit according to the second embodiment of the present invention. is there.

 FIG. 15 is a diagram showing a time chart of the statistical operation according to the second embodiment of the present invention.

 FIG. 16 is a flowchart of the knitting operation according to the second embodiment of the present invention. FIG. 17 is a sequence diagram for explaining the performance management method.

 FIG. 18 is a diagram showing the format of the bioperformance monitoring cell.

 FIG. 19 is a diagram showing possible values of VPI, VCI, and PTI.

 FIG. 20 is a functional block diagram of a conventional performance management arithmetic circuit.

 FIG. 21 is a diagram showing a configuration of a conventional performance management calculation unit.

 FIG. 22 is a functional block diagram of a conventional block size determination circuit.

 FIG. 23 is a functional block diagram of a conventional editing circuit. FtS form for carrying out the invention

 (A) Description of the first embodiment of the present invention

 FIG. 1 is a functional block diagram of a performance management arithmetic circuit (intermediate arithmetic circuit) according to the first embodiment of the present invention.

 The performance management calculation circuit 9 shown in FIG. 1 performs a performance management calculation with a quality of β 接 続 for each connection path or for each of the terminal point and middle vertical point identification information, and outputs multiple types of performance evaluation results. And has an element operation unit 10. The performance management arithmetic circuit 9 is an arithmetic circuit provided inside a terminal (not shown) accommodated in the ATM network. The arithmetic circuit 9 is transmitted from the transmission side, and is transmitted from the ATM switch, VP cross-connect, and ATM. The received cell is processed through the exchange.

 Here, the connection path means the connection of VC, NVP, and UVP, and in the following description, it is referred to as a connection. Further, the terminating point “Nakahiro identification information” means an identifier of the force between the end points or the force between Nakahiro with respect to the connection of VC, NVP, and UVP described above. In the following description, a performance management operation is used to mean an intermediate operation.

The element operation unit 10 performs the performance management operation on the connection and the termination point Neutral performance The performance comfort calculation on identification information can be time-divided within one cell processing time to calculate a plurality of types of performance Hffi results. It comprises a unit 12 and a control unit 13.

 Here, the skin calculation data selection output section 11 can selectively output three types of skin calculation data (VC, NVP, UVP) by an external signal, and holds the VC skin calculation data. It has a section 11a, an NVP skin operation data storage unit, a UVP skin operation data storage section 11c, and a selection section 11d.

 The VC operated data storage 11 a holds VC extracted operation data, the N VP skin operation data holding unit 1 lb holds N VP skin operation data, and The VP operation data section 11c holds the UVP skin operation data. Here, the VC skin calculation data holding unit 11a, NVP¾ "calculation data storage i 1b, UVP¾ ~ calculation data holding unit 11c are each composed of a plurality of memory areas. The memory function is a flip-flop. It can be realized even in the event of a party or a registry.

 The selector 1 Id is connected to each of the three storages, and outputs three kinds of skin calculation data (VCfe calculation data, NVP skin calculation data, and UVP skin calculation data). It can be selectively output by a scheduling signal input from the control unit 13. The function of the selection unit 11d is realized by, for example, a selector.

 Thus, three kinds of skin calculation data of V VP and UVP are available, and the selection unit 11 d selectively outputs the data to a subsequent stage.

 Further, the control unit 13 receives the user cell, and sets a scheduling {word that can be input to the skin calculation data selection output unit 11 and the performance management calculation unit 12; the cell type determination unit 13a, the cell It comprises a counter 13b and a preprocessing scheduling section 13c.

Here, the cell type determining unit 13a can determine the connection of the received user cell and output a senor type determination result. Specifically, the cell type determination unit 13a reads the head area (see FIG. 18) of the received user cell and performs preprocessing for determining three types of connections, VC, NVP, and UVP. It is output as a cell type determination result. Note that this cell type has a connection-specific meaning.

 Further, the cell counter 13b measures (or counts) the number of received user cells and outputs a measured value, and can output a time-division signal, and functions as a counting unit. I have. The cell counter 13b is synchronized with the head of the received cell.

 Further, the preprocessing scheduling unit 13c is connected to the cell type determination unit 13a and the cell power counter 13b, and selects the extraction data from the cell type determination result and the time division signal. A scheduling signal is input to the output unit 11 and the performance management calculation unit 12. This scheduling ί symbol is a signal calculated from the cell counter 13 of the above, and is a signal used to time-divide one cell processing time in detail.

 As a result, the cell type determining unit 13a performs preprocessing on which of the three types of user cells, VC, NVP, and UVP, has been received for the received user cells, The cell judgment result is output. Then, the cell counter 13b measures the number of user cells received one after another and outputs a time-division signal. Here, the cell counter 13b is synchronized with the head of the received cell. Furthermore, pre-processing scheduling unit 13. In, the scheduling is performed in detail from the cell type determination result and the time division signal, and the scheduling signal is input to the operated data selection output unit 11. Then, in the selection section 11 d in the skin calculation data selection output section 11, one of the three kinds of skin calculation data of VC, NVP, and UVP is output by the scheduling signal. , And are selectively output to the subsequent stage.

 Therefore, first, preprocessing is performed, and the preprocessed signal is subjected to performance management calculation.

 Further, the performance management calculation unit 12 shown in FIG. 1 performs a performance management calculation using three types of calculation data output from the skin calculation data selection / output unit 11 and obtains multiple performance results. It can be selectively output by an external signal.

(Ρ ΜDance) 14 and memory areas 12 a, 12 b, and 12 c Have been.

 Here, the performance management exercise 14 is based on the MCSN, TUC O, TUC0 / 1, BED for each of the three connections of VC, NVP, and UVP or for each point (end point 'middle boat') for the received user cell. It performs 5 g¾l performance management calculation of the number of transmitted cells.

 The meaning of these values is as follows. That is, MCSN is the sequential number of the received performance monitoring cell, TUC O represents the transmission iim of the user cell of CLP 0, and TUC 0 Z1 is the total number of users of CLP 0 and 1. In addition, BEDC is the result of performing BIP-16 calculation on the ^ t blue information area of all user cells after the performance monitoring cell transmitted immediately before. The number of transmitted cells is stored as the number of transmitted cells .. CLP 0 is a value indicating a cell or a cell, and TUC 0 indicates a loss of CLP 0 cells or erroneous insertion of cells. Is detected, and TUC0Z1 detects cell loss and erroneous cell insertion for all cells of CLP0 and CLP1.

 The memory page areas 12a, 12b, and 12c hold the results of performance management computation performed for each connection of VC, NVP, and UVP, and represent memory areas. That is, in the memory area 12a, for the VC, EE-MCSN, EE-TUCO, EE-TUC 0/1, EE-BEDC, EE-the performance of measuring the number of cells to be transmitted 1¾ SEG—MCSN, SEG-TUC 0, SEG-TUC 0/1, SEG—BEDC, SEG measurement performance per VC Is maintained.

Similarly, in the memory area 12b, for the NVP, for the termination point, E E-MCSN, EE-TUCO, EE-TUC 0/1, EE-BEDC, EE-Performance management of the measurement of the number of cells to be transmitted For the result of the calculation and the NVP, the performance management calculation of the measurement of the number of cells transmitted per SEG-MCSN, SEG-TUC 0, SEG-TUC 0/1, SEG-BEDC, SEG for the middle point Forced with results. In the memory 12c area, for the UVP, the performance management calculation of the end point, EE-M CSN, EE-TUC 0, EE-TUC 0/1, EE-BED EE, measurement of the number of transmitted cells For the UVP, for the neutral power, the results of the performance management calculation of SEG-MCSN, SEG-TUC 0, SEG-TUC 0/1, SEG—BED C, SEG—number of cells transmitted It is forced. Note that these memory functions can be realized by flip-flops, registers, and the like. As described above, the operation for performing the performance management operation for each connection is not prepared, and the connection type is determined by the cell type determination unit 13a, and the operation is performed in a time-division manner by the cell power counter 13b. By doing so, one performance performance II 4 can be shared.

 FIG. 2 is a diagram showing a time chart for explaining a time division method for each connection according to the first embodiment of the present invention. In Fig. 2, there are three stages, the upper stage, the middle stage, and the lower stage. The upper row of cells represents the receiving cell, and the cells with U (UVP calculation target cell) and those with N (NVP Calculation target cells) and those with N and VC (NVP and VC calculation target cells) are received strongly. Note that the cells that have been received are also omitted from the blank section.

 Then, the middle cell row is a time-expanded part of the upper cell row. Further, it shows the time allocation of the operation within the time of the lower middle cell row. Then, as shown in the lower part, for the UVP calculation target cell, the time for the calculation processing is assigned a time half of the reception time. Similarly, for the NVP calculation target cell, the time for processing is allocated half the time of the received time.

Here, for the NVP and VC calculation target cells shown in the middle cell row, not only VC but also NVP is calculated using ^^. The reason for performing this operation is as follows. That is, when a VC cell is received, it is necessary to know not only this VC cell but also information about the VP to which the VC cell belongs by calculation. Also, the reason why you need to know the information about the VP to which the VC cell belongs is 1 ^? Because it accommodates VC connections. Therefore, for the operation processing of VC, after performing the operation of NVP, the operation of VC It is done force.

 Thus, after the connection processing is determined by the preprocessing, performance management difficulties are performed between UVP, NVP, and NVP, VC. Is performed. Further, after the connection determination, a calculation based on a difference between the neutral point and the terminal point is performed, so that a calculation of five views is performed.

 FIG. 3 is a detailed block diagram of the performance management calculation unit 12 according to the first embodiment of the present invention. The performance management calculation section 12 shown in FIG. 3 includes a performance management performance section 14 and calculation sections {bearing sections 16a, 16b, 16c, 16d, 16e. And this performance management performance 14 has the 1st ring part 14a, 14b, 14c, 14d, 14e, and the calorie calculation part 15.

 Here, the first security ## 14a, 14b, 14c, 14d, 14e is composed of five types of skin operation data (MCSN, TUC0, MCN) selected by the scheduling signal from the control unit 13. TUC 0Z1, BEDC, and five types of measurement of the number of transmitted cells). Specifically, the first holding unit 14a stores the MCSN, the first holding unit 14b stores the TUC 0, and 4 c stores the TUC 0Z1, the first hold 4 d stores the BE DC, and the first hold 14 e stores the measured value of the number of transmitted cells. Further, the adder 15 is connected to each of the first Hogos 14a, 14b, 14c, 14d, and 14e, and performs the performance i-ri operation using these skin operation data. It is possible to output each of the performance results.

It consists of 5b, 15c, 15d, and 15e. The calculation results 16a, 16b, 16c, 16d, and 16e are the five types of performance results that are output from the calorie deficiencies 15a, 15b, 15c, 15d, and 15e, respectively. It is something you can do.

As a result, MCSN, TUC 0, TUC 0/1, BEDC, and the number of cells to be transmitted are obtained for a total of 6 types for each of the two points (termination point and media) for the three types of connections: VC, NVP, and UVP. The five types of performance management calculations are performed. That is, the performance management calculation unit 12 is one unit of calculation. Then, for example, when a performance management operation is performed on a termination point relating to the VC, first, five types of operation target data selected by the scheduling signal from the control unit 13 are respectively stored in the first holding unit 14 a , 14 b, 14 c, 14 d, and 1 e are read out, and the adder 15 performs a performance management operation using these operand data. The calculation results are stored in the holding units 16a, 16b, 16c, 16d, and 16e.

 Next, in the performance management operation circuit 9 shown in FIG. 1, the operator in the performance management calculation unit 12 is determined by scheduling the operator for each point using the cell counter 13 b. The sharing method will be described with reference to FIGS.

 FIG. 4 is a diagram showing a time chart for explaining a time-division method for each point time-divided for each connection according to the first embodiment of the present invention. In FIG. 4, there are two stages, an upper stage and a lower stage. The upper row of cells indicates the received user cells, and the UVP operation target cell, the NVP operation target cell, and the NVP and VC operation The image cell has been received. The lower part shows the time allocation of the arithmetic processing within one cell processing time of the upper user cell row.

 As shown in the lower part of FIG. 4, the time for the operation of the u V p operation target cell is allocated half the time of the received time, and the time is further divided into the middle and end points. And shared with, half the time is allocated. The same is true for the NVP calculation target cell.

 Furthermore, for the NVP and VC calculation target cells, the time for the NVP operation and the time for the VC operation are respectively assigned before and after the reception time, Time is further shared by Nakahiro and the endpoint, with half the time being allocated. The reason why the calculation is performed between the VC and the NVP is the same as that described above, and further description will be omitted.

In this way, different calculations for each point are time-divisionally calculated by scheduling. For this reason, when performing 5 operations such as MCSN, TUCO, TUC 0/1, BEDC, and measurement of the number of cells to be transmitted, performance management performance 1 can be shared between Nakahiro and the endpoint. The three types of connections are aggregated. Therefore, the circuit scale can be greatly improved. FIG. 5 is a conceptual diagram for explaining sharing of the performance management performance 14 for each connection according to the first embodiment of the present invention.

 VC operation data holding unit 11a shown in Fig. 5, NVP skin operation data

 The data to be processed (MCSN, TUCO, TUC0 / 1, BEDC, measurement of the number of cells to be transmitted) stored in each memory area of the lb and UVP skin calculation data holding unit 11c are controlled by It is selected by the scheduling signal from the section 13 and input to the calorie calculation section 15. Then, in the adding unit 15, the five types of performance results calculated are stored in each of the memory areas 12a, 12b, and 12c.

 In this way, the time division of each connection is further divided in a time division of each point, and the adder 15 is shared, so that an increase in the circuit scale is prevented, and the performance of the performance management operation is reduced. Therefore, the operation can be performed without increasing the difficulty, and the memory can be used efficiently.

 Next, the time-division data for each connection and for the middle and end ^ 演算 are further calculated between a plurality of skin calculation data (MCSN, TUC O, TUC0Z1, BEDC, ΐ10 measurements of the number of cells to be transmitted). The method of consolidating is explained using FIG. 6 and FIG. FIG. 6 is a function block diagram of another performance calculation unit according to the first embodiment of the present invention. The performance management calculation section 19 shown in FIG. 6 includes a first holding section 19a, an addition section 19, and a calculation and holding section 19c.

 The first holding unit 19a stores five kinds of skin calculation data selected by the scheduling signal from the control unit 13 (MCSN, TUC0, TUC0 / 1, BEDC, 5 mil for measuring the number of transmitted cells). The addition unit 19b is connected to the first observation unit 19a, performs a performance management operation using the math calculation data, and outputs each of the five types of performance results. It is a good thing. Further, the calculation result 1 holding unit 19c can display the five performance Hffi results output from the adding unit 19b.

With these, the total of 6 types for each 2 points of view (termination point / neutral) regarding 3 types of connections of VC, NVP and UVP, and MCS Ν, UCUC0, TUC0 / 1, BEDC, transmission Five types of performance management calculations for measuring the number of cells The calculation is performed in one performance management calculation unit 19.

 For example, when a performance management operation is performed on a termination point relating to a VC, first, five types of operation target data selected by the scheduling signal from the control unit 13 are respectively stored in the first holding unit 19a. , And a performance management operation is performed in the power calculation unit 19b using the extracted operation data, and each of the five types of performance evaluation results is output to the operation result mm 9c. It is preserved.

 FIG. 7 is a diagram showing a time chart for explaining a time-sharing method for each difficulty class when time-sharing is performed for each connection and for each point according to the first embodiment of the present invention.

 The time-sharing method shown in FIG. 7 is also a method of scheduling operators for each class using the cell counter 13b in the performance prosperity calculation circuit 9 shown in FIG. This is a method of sharing operators inside the circuit.

 In Fig. 7, there are two stages, the upper stage and the lower stage.The upper cell line indicates the received user cells, and the lower line indicates the time allocation of the arithmetic processing within the time of the upper cell line. It is. Also, as shown in the upper part of FIG. 7, there are three types of connection power in the reception cell. As shown in the lower part, half of the received time is allocated to the UVP calculation target cell for the time required for the operation, and the time is further divided into , And half of the time is allocated to the operation related to the end point.

 Then, the operation time related to the operation is allocated to the time for the five types of operations of MCSN, TUCO, TUC 0/1, BEDC, and the measurement of the number of transmitted cells, and the same performance management is performed between these operations. The operation unit 19 is shared. The same applies to the NVP calculation target cell.

Further, for the NVP and VC calculation target cells, the time for the NVP operation and the time for the VC operation are respectively assigned before and after the reception time. In addition, half of the time is allocated to the operation for Nakahiro and the operation for the endpoint. Then, the operation time related to the intermediate power is allocated to the time for five types of operations of MCSN, TUC 0, TUC 0/1, BEDC, and the measurement of the number of transmitted cells, and the same performance is obtained among these operations. Management operation Part 19 Powerfully shared. The reason for calculating VC, NVP, and ^^ # is the same as that described above, and further description will be omitted.

 FIG. 8 is a conceptual diagram for explaining the sharing of the performance management calculation unit 19 for each connection and each point according to the first embodiment of the present invention.

 In the memory area of the VC skinned data storage area 11a, the NVP processing data storage section 1lb, and the UVP processing data storage section 11c shown in FIG. The calculated data is stored. That is, in the VC operation data holding unit 11a area, SEG-MCSN, SEG-TUC 0, SEG-TUC 0/1, SEG-BEDC, SEG_number of cells to be transmitted and EE-MCSN relating to the end point are stored. , EE-TUC 0, EE-TUC 0/1, EE-BE DC, EE—Stores the number of transmitted cells and the number of transmitted cells.

 And VC skin calculation data storage ί ^ Ι 1 a, NVP skin calculation data storage unit 1 lb,

The data is selected by a data scheduling signal stored in each memory area of the UVP storage data holding unit 11c and input to the addition unit 19b. Then, performance calculation is performed in the adder 19b, and the results are stored in the respective memory areas 12a, 12b, and 12c. As described above, by the scheduling, the calorie calculation unit 19b performs the performance management calculation time-divided for each of the three connections, further performs the performance management calculation time-divided between the middle sheet and the terminal point, and Performance management calculations are performed in a time-sharing manner for each of the five views.

 In this way, the functions can be shared without preparing a plurality of different operators, so that it is possible to prevent an increase in the circuit scale. Calculations can be performed without increasing difficulty, and efficient use of memory can be achieved.

 Further, as a method for reducing the circuit scale, a block size determination circuit will be described with reference to FIGS. 9 (a) to 9 (c).

FIG. 9A is a functional block diagram of the block size determination circuit according to the first embodiment of the present invention. The block size determination circuit 21 shown in FIG. 9 (a) determines whether the number of received cells has reached the block size, and stores the operation result. Holding unit (Operation result holding unit for TUC 0 Z 1) 16 c Lower bit extraction unit 21 a, block size setting unit 21 b, match detection unit 21 c, F—PM cell transmission request unit 21 It is configured with d.

 Here, the arithmetic and listening unit 16c can measure the number of received user cells and display the value in a plurality of bits, and functions as a transmission cell number holding unit. Also, specifically, the arithmetic and control unit 16. Is a memory that stores the value of TUC 0/1.

 Further, the block size setting unit (block length setting unit) 2 lb is a number of user cells transmitted and received on ί¾¾, and sets a block size representing the number of user cells as a power of two.

 In addition, the lower bit extracting unit 21a reads a bit out of the 16 bits of the operation result 16c from the block value setting unit 21b. Thus, for example, the function is exhibited by performing an AND operation with the value of the lower 10 bits.

 The match detecting unit 21c can detect that the number of user cells transmitted from one path matches the block size based on the result read by the lower bit extracting unit 21a.

 The F-PM cell transmission requesting unit 21d outputs a signal requesting transmission of a performance monitoring cell when a coincidence is detected by the coincidence detecting unit 21c.

As a result, in operation result 6c, the number of received user cells is constantly measured, and in block size setting section 21b, a block size representing the number of user cells consisting of powers of 2 is set. If the value is set to “1”, the lower 10 bits of the 16 bits of the arithmetic result section 16c are read from the block size setting value in the lower bit extracting section 21a. Then, based on the result read by the lower-order bit extraction unit 21a, the coincidence detection unit 21c detects that the number of user cells transmitted from one path matches the block size, and detects the F- In the PM cell transmission request section 21d, when a match is detected, a signal requesting transmission of a performance monitoring cell is output. Here, for example, the flow of processing when the number of paths determined by VCI and VPI is 4,096 and the block size is 10 2 4 (2 to the 10th power) is shown in FIG.

This will be described using 9 (b).

 FIG. 9 (b) is an explanatory diagram of a method of reading the lower bits of the block size determination circuit according to the first embodiment of the present invention. The operation result 16c shown in FIG. 9 (b) is such that the value is incremented every time a cell is received, and is continuously incremented without being reset, and the value is 16 bits. It is displayed as a button. Note that the value for the upper 6 bits is unnecessary for the explanation, so d, which represents d'n't carre, is inserted.

 Here, if there is no loss of the user cell, when the head of the cell is received, the display of the value of the lower 10 bits of this operation result 6c becomes all 0, and the number of 10 2 3 When the cell of (1) is received, the values of the lower 10 bits are all 1. further,

When the 104th cell is received, the values of the lower 10 bits are all 0, and the next leading cell is received. Therefore, the detection of the reception of cells of 1204 block sizes is detected by reading the lower 10 bits.

 In this way, in the lower bit extracting section 21a, the lower 10 bits of the 16 bits of the arithmetic result 6c are read from the set value set by the block size setting section 21. At 2 1 c, it is detected whether the lower 10 bits of the power have all become 1 or not.

 In this way, by using the operation surplus portion (operation result for TUC 0/1 | observation portion) 16c in the circuit, for example, a memory area of 4096 × 10 bits can be obtained. Even without securing, the number of user cells for the block size can be measured. That is, for example, there is no need to prepare a 10-bit memory for each of the 496 paths. This also makes it possible to determine that a block-sized cell has been received without requiring an operator memory such as an adder for performing or processing those memories, greatly simplifying the circuit scale. You will be able to dagger.

Furthermore, even if there is a loss of a user cell or the like, and the value of the arithmetic and listening unit 16c deviates, it is possible to make the same determination using the following method. FIG. 9C is an explanatory diagram of a second method for reading the lower bits of the block size determination circuit according to the first embodiment of the present invention. For example, 1001100110 is stored in the lower 10 bits of the arithmetic and logic unit 19c shown in FIG. 9 (c), and the values for the upper 6 bits are unnecessary for explanation. d is inserted.

 Then, when the first cell is received, the lower 10 bits (for example, 10011001 10) of the lower 6 bits of the operation result section 16c shown in FIG. 9 (c) are input to the lower bit extracting section 21a. Read well. Each time a cell is received, the match detection unit 21c monitors the display of the calculation result holding unit 16c, and when the display becomes 1001100110 again, the 1024th The fact that a cell has been received is detected.

 In this way, in the case of, even in the case of a user cell ^, user cell erroneous insertion or user cell error, by reading the lower 10 bits, for example, cells of a block size of 102 4 can be obtained. It becomes the power of detection of the reception. In addition, a block-sized cell was received without the need for an operator memory such as a memory area of 496 X 10 bits or an additional memory for operating those memories. Can be determined, and the circuit scale can be greatly reduced. With such a configuration, the performance management operation circuit 9 processes the user cell from the transmitting side and performs the performance operation.

FIG. 10 is a flowchart of the performance management operation according to the first embodiment of the present invention. The flowchart shown in FIG. 10 is started when a cell is received. First, in step A1, the connection strength of the receiving cell is determined. Here, when the connection of the receiving cell is UVP, a UVP route is taken, and steps after step A2 are executed. Similarly, if the connection of the receiving cell is NVP, the NVP route is taken, the steps after step A6 are executed fi ¹ , and if the connection of the receiving cell is VC, the VC route is taken, Step A7 and subsequent steps are executed.

First, when the connection of the receiving cell is determined to be UVP, the connection information is set to UVP (step A 2). Then, in step A3, it is determined whether or not the received cell is a user cell. If the received cell is a user cell, a YES route is taken, and in step A4, a measurement operation is performed. Note that this measurement The operation will be described later. In step A3, if the received cell is not the user's cell, the NO route is taken. In step A5, it is checked whether a special number (special number) is set, and if it is set, the YES route is set. In step A4, a measurement calculation is performed. If it is not set, the program takes N0 root and the program ends.

 In step A1 of FIG. 10, when the connection of the receiving cell is NVP, the NVP route is taken, the connection is determined to be NVP, and the connection information is set to NVP (step A6). A5 and subsequent steps are executed.

 FIG. 11 is a flowchart of the measurement calculation according to the first embodiment of the present invention. The flowchart shown in FIG. 11 is called at step A4, step A11, step A12, and step A14 in FIG.

 In FIG. 11, first, in step B1, it is determined whether the point is between relay points, and if it is medium, the MCSN, TUC0, TUC0 / 1, BEDC, and the number of cells to be transmitted in step B2 are determined. The five types of computations for measurement are executed. Also, if the connection is not neutral in step B1, the force between the connection endpoints is checked in step B3, and the MCSN, TUC O, TUC 0/1 in step B4 are checked. , BEDC, and the measurement of the number of cells to be transmitted are executed.

 FIG. 12 is a flowchart of the calculation processing of the performance evaluation result according to the first embodiment of the present invention. The flowchart shown in FIG. 12 corresponds to the routine in step B2 and step B4 in FIG.

 Then, in step C1, the setting of the calculation enable is sewn. If the setting is not possible, the routine is terminated by taking the Fa1 selt. Using the route, in step C2, the calculation of the expression (1) is performed on the number-of-transmitted memory.

 Transmission number memory = Transmission number memory + 1… (1)

Subsequently, in step C3, the operation of equation (2) is performed on TUC0Z1. TUC 0/1 = TUC 0/1 + 1-(2)

 Further, in step C4, the operation of equation (3) is performed on BEDC0Z1. Here, exor represents an exclusive OR operation.

 BEDC 0/1 = BEDC exor BIP 16 operation. "(3)

 Subsequently, in step C5, the calculation of equation (4) is performed for TU CO.

4 signs.

 TUC 0 = if (CLP = 0) {TUC 0 + 1}… (4)

 That is, the TUC 0 is incremented for the user cell having a high degree of merit. In step C6, the block size is determined. If the number of cells to be transmitted has not reached the block size, a NO list is taken here and the program ends. On the other hand, if the number of cells to be transmitted has reached the block size, the YES route is taken, the transmission number memory is reset in step C7, and in step C8, the calculation of equation (5) is performed on the MCSN. Done.

 MCSN = MCSN + 1… (5)

 Finally, in step C9, the performance monitoring cell to be inserted is calculated, and the calculation is completed for MCSN, TUC 0, TUC 0/1, BEDC, and 5 @ ¾ | Then, the measurement calculation in steps B1 to B4 in FIG. 11 ends.

 Again, in step A1 of FIG. 10, the connection of the received cell is determined, and if it is determined that the received cell is VC, the connection information is set to VC (step A7). It is determined whether the cell is a user cell. Here, if the received cell is a user cell, a YES route is taken, a measurement operation is performed in step A12, and in step A13, the connection information is changed to NVP in step A13. Measurement calculation is performed, and the program force ends. That is, by switching the connection information, the performance management operation is switched within one cell processing time.

In step A8, if the received cell is not a user cell, the N0 route is taken.In step A9, it is checked whether a special number is set, and the special number is checked. If YES, take the YES route and Calculation is performed. If not set, the NO route is taken, the connection information is set to NVP in step A10, the measurement calculation is performed in step A11, and the program ends.

 As described above, before the measurement calculation in step A4, step A11, step A12, and step A14, the steps shown in the frame 25 are performed as preprocessing.

 Also, by sharing the operators in the circuit for performing the performance management operation in a time-sharing manner, the circuit scale can be reduced. In addition, by scheduling, the eclipse management operation can be performed in a time-division manner by connection, by the neutral end point J¾lj, and by the function ^ S, and one arithmetic unit can be shared. Can be prevented. In this way, the substrate area can be reduced, and the cost of a printed circuit board (PCB) can be reduced. In addition, the circuit scale can be used to promote low power consumption, and resources can be used for cranes.

 Also, in this way, even if the power of the performance calculation is {1}, it can be handled by changing the calculation schedule, so that the calculation can be performed without increasing the performance, and the power consumption associated with it This has the advantage that the increase in the number of circuits and the circuit size can be suppressed, and the memory can be used efficiently.

 (B) Description of the second embodiment of the present invention

 FIG. 13 is a functional block diagram of the knitting operation circuit according to the second embodiment of the present invention. The edit operation circuit 30 shown in FIG. 13 is a edit operation circuit that receives a performance monitoring cell having information on the quality of β¾ and outputs a plurality of evaluation values for a transmission line. It is composed of a ten arithmetic unit 32 and a control unit 33. The editing unit 3 2 is capable of selectively outputting a plurality of ffi values for fei ^, and the performance unit 3 la, 3 1b, 3 1c, 3 1d, 3 1 e, 31 f, 31 g, the first operation part 35, and a plurality of operation result holding parts 36 a, 36 b, 36 c, 36 d, 36 e, 36 f, It consists of 36 g.

Here, the performance words 3 1 a, 3 lb, 31 c, 31 d, 31 e, 31 f, 31 g hold a plurality of values for the ^ ¾ path, respectively. What you can do Then, the values stored in them are as follows. In other words, the performance evaluation result 1 SECB Lost indicating the loss of the threshold value of RJ: in the holding section 31a, and the performance evaluation: SECB Misinserted indicating the erroneous insertion of the threshold value JiLh in the H holding section 31b. In the performance evaluation result holding section 31c, the number of CLP0 + 1 erroneous insertions indicating the number of CLP0 + 1 erroneous insertions with a threshold value full, and the performance IHffi 結 1 In the holding section 31d, the CLP0 + 1 with the threshold l * full CLPO + 1 Loss, which represents the cell loss number, is stored.

 Further, the cell loss number of the CLP0 + 1 cell of the performance lffi-bearing unit 31e is shown in Table-TTotal CLPO + 1 Loss number, and the CLP0 cell of the threshold Nm is stored in the performance Nffi-bearing unit 31f. The CLPO cell loss number, which represents the number of irons, and the performance Hffi 31a, store the number of CLPO cell losses, "TTotal CLPO Loss number".

 Further, the first calculation unit 35 performs a tenth calculation, and includes a first selection unit 35a, an addition unit W ^ 35b, and a force. The first selection unit 35a can selectively output a plurality of l¾ values for ^ in a time-sharing manner according to the scheduling signal input from the control unit 33, and The adder 35b is capable of performing a tenth operation using a plurality of values for the output from the first selector 35a and outputting a plurality of performance results.

 The operation results 36 a, 36 b, 36 c, 36 d, 36 e, 36 f, and 36 g are obtained by performing {¾ ^ on a plurality of types of performance results from the first operation unit 35. These include SECB Lost, SECB isinserted, CLPO + 1 misinsertion count, CLPO + 11 oss count, Total CLPO + 1 Loss count, CLPO Loss count, Total CLPO Loss count, and tl¾ each. ing.

 As a result, the numbers stored in the performance dragon concoction units 3 la, 31 b, 31 c, 31 d, 31 e, 31 f, and 31 g are stored in the first selection unit 35 a in the first calculation unit 35. The output is scheduled and output in detail, and the adding unit 35b performs a tenth operation, and the resulting performance Hffi result of the class S is calculated and stored in the operation result holding units 36a, 36b, 36 They are stored in c, 36 d, 36 e, 36 f, and 36 g.

In addition, the control unit 33 shown in FIG. 13 provides one of the values for the number of user cells lost—the value for the number of incorrectly inserted cells and the value for the total number of user cells. The scheduling signal can be input to the editing unit 32 by using L, the value of the end, and the value of the difference with respect to the number of user cells with liability.

 The control section 33 includes a cell counter (counting section) 33b, a threshold comparison result output section 39, and a time-division signal output section 33c. Here, the cell counter 33b can measure the number of received user cells, output a measured value, and output a time-division signal. The cell counter 33b is synchronized with the received cell.

 Then, the threshold comparison result output unit 39 outputs one of a threshold value allowed for the number of user cell losses and an insertion threshold value allowed for the number of erroneous insertions of the user cell, and ^ for the total number of user cells. {1} A threshold value comparison result is output by a time-division signal input from the cell counter 33b using any of the difference values regarding the number of user cells with priority.

 Further, the time-division signal output section 33 c can input a scheduling signal to the editing section 32 using the threshold value comparison result and the time-division signal. As a result, the time-division signal is input from the cell counter 33b synchronized with the reception cell to the harmful IJ signal output unit 33c, and the time-division signal output unit 33c is output from the cell counter 33b. The scheduling signal is input to the first processing unit 32 and the first selection unit 35a in the first calculation unit 35 in the processing unit 32. Ten operations are performed by scheduling. Then, a plurality of Hffi values of the editing unit 32 are output in a time-division manner.

 FIG. 14 is a more detailed functional block diagram of the control unit 33 according to the second embodiment of the present invention. The control unit 33 shown in FIG. 14 includes a cell counter 33 b, a time-division signal output unit 33 c, and a threshold comparison result output unit 39.

The threshold comparison result output unit 39 includes a ίΐ ^ threshold value holding unit 39 a, an incorrectly inserted threshold value holding unit 39 b, a second selection unit 39 c, a TUCD0 hogo 39 d, and a TUCD0 / It consists of 1 holding unit 39e, third selecting unit 39f, sign judging unit 39g, and comparing unit (comparator) 39h. The TUCDO holding unit 39 d holds a difference value (TUCDO) between the number of transmission cells (CLP0) and the scheduled transmission number (TUC0) for the number of priority cells (CLP0). Holds the value (TUCD0 / 1) of the expected number of transmissions and the number of transmissions (TUC0 / 0) for the number of all cells (CLP0 / 1). The third selection unit 39 f selects one of the TUCDO holding unit 39 d and the TUCD0 / 1 holding unit 39 e as a difference value according to the time division signal input from the cell counter 33 b. It can be output in a typical manner. The TUCDO holding unit 39d, the TUCD0 / 1 ring unit 39e, and the third selection unit 39f cooperate to function as a difference ^ t green information output unit.

 The determination unit 39g can determine whether the difference value (TUCDO, TUCD0 / 1) is positive or negative and output the result as ^.

 As a result, first, for CLP0 having a high fem- ple, the difference (TUCDO) power between the number of user cells to be transmitted and the number of actually transmitted user cells <, is obtained in the TUCD0 holding unit 39d. For example, for CLP0, if the number of user cells scheduled to be transmitted is 100 and the number of user cells actually transmitted is 98, TUCD0 is -2. Similarly, if the number of user cells to be transmitted is 100 and the number of user cells actually transmitted is 101, TUCDO is +1.

 Similarly, for CLP0 / 1, the difference (TUCD0 / 1) between the number of user cells scheduled to be transmitted and the number of user cells actually transmitted is found in TUCD0 / 1 storage 39e. For example, for CLP0 / 1, if the number of user cells to be transmitted is 500 and the number of user cells actually transmitted is 498, TUCDO is 2. Similarly, if the number of user cells scheduled to be transmitted is 500 and the number of user cells actually transmitted is 501, TUCD0 is +1.

The TUCD0, TUCD0 / 1, and the third selector 39f are scheduled by the time division signal from the cell counter 33b, and are selectively output. Further, either one of the selected TUCDO and TUCD0 / 1 is input to the sign determination unit 39a and to the comparison unit 39h. For example, a TUCD0 force having a value of +1 is selected and output, and the value is input to the force determination unit 37a, where the sign of + Z— is determined, and the TUCDO value of +1 is compared. Entered in part 3 9h. Further, the loss threshold value holding unit 39a shown in FIG. 14 can hold the threshold value (Mlost) of the loss number of the user cell, and the erroneous insertion threshold value holding unit 39b stores the threshold value of the user cell. It can hold the threshold value (Mmisinserted) of the number of erroneous insertions, and the threshold value (Mlost) of the iron number of the user cell and the threshold value of the number of erroneous insertions of the user cell in the second selection unit 39 (Mmisinserted), which can be time-divided according to the code and selectively output as a threshold value to the comparing unit 39h.

 Further, the comparing section 39h can compare the threshold value (Mlost or Mmisinserted) with ^ Ht (TUCD0 or TUCD0 / 1) and output a threshold value comparison result. In addition, this]: the comparing section 39 h calculates the difference value (MCSND) between the sequential number of the user cell to be received and the sequential number of the actually received user cell, and calculates the MCSND based on this difference. In addition, for the threshold value and the ^ value, L, which is the number of irons in the user cell, is used to select whether to calculate the number of incorrect insertions of the user cell.

 Then, the threshold value 39 a is set to the threshold value (Mlost) of the number of loss of user cells, and the threshold value of misinsertion threshold value 39 b is set to the threshold value of the number of misinsertion of user cells (Mmisinserted). Force is set. For example, Mlost is set to -2 and Mmisinserted is set to +2. Then, in the second selection unit 39c, the difference between Mlost and Mmisinserted, the deviation force, is input from the sign judgment unit 37a using the judgment result of +/−, Is selected in a time-sharing manner in accordance with the difference between and is output to the comparing section 39h. For example, when the value of the difference value TUCD0 is _2, it is sufficient to compare with Mlost and not with Mmisinserted. Similarly, when the value of ^ MITUCDO is +1, it can be compared with Mmisinserted, not with Mlost. Therefore, according to the judgment result of +/−, only one of Mlost and Mmisinserted is output strongly and selectively.

Further, in the comparing section 39h, the TUCD0 or TUCD0 / 1 output from the third selecting section 39f is compared with Mlost or Mmisinserted output from the second selecting section 39c to set the threshold. Value i: is less than or equal to And the threshold value comparison result is input to the control function 33c. At this time, which of the number of user cell losses and the number of user cell erroneous insertions to perform is selected based on the MCSND.

 For example, if the MCSND force is 2 or less, Mraisinserted is calculated, and if MCSND is greater than 2, the Mlost force or calculation is performed. Here, if Mmisinserted is set to +2, when TUCD0 power + 1, it is within the allowable insertion error, but when TUCD0 is +3, it is not allowed. On the other hand, if Mlost is set to 1, when TUCD0 is 1, the force is within the allowable loss. TUCD0 / 1 is also compared in the same way.

 Then, in the control operation 33c, in synchronization with the cell counter 33b, based on the threshold value comparison result from the comparison unit 39h, the control unit 32 For the first selector 35a of the scheduler, a scheduling message is input, and performance discussions 31a, 31b, 31c, 31d, 31e, 31f are performed. , 31g, one value is selected, input to the addition unit 35a, and the knitting operation is performed. Then, the result of each operation is calculated or a predetermined calculation result holding unit 36a, 3 They are stored in 6b, 36c, 36d, 36e, 36f, 36g.

 The cell counter 33b can input a detailed schedule signal to the third selecting section 39f and the control difficulty 33c in synchronization with the receiving cell. FIG. 15 is a diagram showing a time chart of the statistical calculation according to the second embodiment of the present invention. In FIG. 15, the upper and lower stages have power. The upper cell is a received performance monitoring cell, which is subjected to performance management calculation. The lower part shows the time allocation of the statistical operation processing within the time of the upper performance monitoring cell.The first half of the cell time shown in the lower part of Fig. 15 is allocated to the operation of TUCD0 + 1. , The latter half are allocated to the operation of TUCD0. In other words, two types of operations are scheduled within a single cell time, and are therefore shared in a thorough manner.

First, in the first half of the TUCD0 + 1 operation time, the operations marked with Ρ and Q are performed. The calculation section marked with Ρ is the calculation section of Total CLP0 + 1 Loss number, and The calculated interval is the interval in which one of the SECB Lost, CLPO + 1 Loss, SECB Misinserted, and CLP0 + 1 error insertions is calculated. One of these is selected by TUCD0 + 1, Mlost, misinserted. More specifically, TUCD0 + 1 is selected according to its +/- sign, Mlost is determined by the threshold value i: or less, and Mmisinserted is selected by the threshold value h or less. ing.

 Furthermore, in the latter half of the TUCD0 calculation time, the calculations marked R and S are performed. Here, R is a calculation section of the total CLPO Loss number, and S is a calculation section of the CLPO Loss number.

 In this way, TUCD0 + 1, TUCD0 (difference between the number of transmitting user cells and the number of receiving user cells) calculated from TUC0 + 1, TUC0 in the performance monitoring cell pay mouth, and the threshold setting value (Mlost, Mmisinserted) ), Power, pre-processing, and scheduling, it is possible to share each operation unit in the 10 operations. Then, the result of the 7-warning compilation (Total CLPO + 1 Loss, CLPO + 1 Loss, SE CB lost, CLPO + 1 Misinsertion, SECB Misinserted, Total CLPO Loss, CLPO Loss) is 1 It is calculated and output in a time-division manner within the cell processing time, and 30 edits are performed, so that the circuit scale can be expanded.

 With such a configuration, when the receiving cell is a performance monitoring cell, the performance monitoring cell is processed in the edit operation circuit 30 to execute the edit operation.

 FIG. 16 is a flowchart of the knitting operation according to the second embodiment of the present invention. The flowchart shown in FIG. 16 is started when a performance monitoring cell is received.

First, in step D1, it is determined whether or not TUCD0 + 1 power is smaller than 0 for all user cells in which both the superior user cell and the other user cell are combined. Here, if the TUCD0 + 1 force is ϋθ: a NO route is taken, and in step D2, it is determined whether or not this TUCD0 + 1 is equal to or smaller than the threshold set value Mmisinserted. If it is equal to or less than the threshold value of the erroneous insertion, the YES route is taken, and in step D3, it is determined whether or not the MCSND force is larger than 2, and the MCSND force is determined. If the force is equal to or less than 2, the NO route is taken, and in step D4, the operation of equation (6) is performed on the number of CLP0 + 1 erroneous insertions.

 CLP0 + 1 error insertion fiber calculation = CLP0U error insertion fiber calculation + MCSND… (6)

 Further, in step D6, the positive / negative of TUCD0 is determined for the dominant user cell. If the TUCD0 is negative, a YES route is taken. Is determined. Then, if the loss is equal to or less than the threshold value of the loss, the YES route is taken, and in step D8, it is determined whether or not MCSND is greater than 2; if MCSND is 2 or less, the N0 route is taken, and In D9, the operation of Expression (7) is performed on the CLPO + I ^ number.

 CLP0Loss = CLP0Loss + TUCD0… (7)

 Further, in step D10, the calculation of Expression (8) is performed on the total CLPO loss number.

 TotalCLP0Loss = TotalCLP0Loss + TUCD0… (8)

 In step D2, if the threshold value _b of TUCD0 + 1 force misinsertion is NO, a NO route is taken, and in step D5, calculation of equation (9) is performed for SECB Misinserted.

 SECBMisinserted = SECBMisinserted + MCSND… (9)

 Similarly, if the MCSND power is larger than 2 in step D3, a YES route is taken, and in step D5, the calculation of equation (9) is performed for SECB Misinserted.

 Then, in Step D 1, if all the user cells including both the user cell with excellent 5¾ and the user cell that does so are TUCD0 + 1, take YES route and in Step D 11, The equation (10) is calculated for TotalCLPO + 1 Loss.

 TotalCLPO + 1 Loss = TotalCLP0 + l Loss + TUCDO + 1… (10)

Further, in step D12, it is determined whether the total CLPO + 1 loss force is equal to or less than Mlost, and if it is equal to or less than Mlost, a YES route is taken, and in step D13, it is determined whether MCSND is greater than 2 or not. If MCSND is 2 or less, N 0 Then, in step D14, the operation of equation (11) is performed for CLPO + lLoss.

 CLP0 + 1LOSS = CLP0 + 1LOSS + TUCD0 + 1… (1 1)

 Then, steps after step D6 are executed.

 In step D12, when the total CLPO + 1 loss force is larger than Mlost, a NO route is taken, and in step D15, the calculation of equation (1 2) is performed on SECBLost.

 SECBLost = SECBLost + CSND-(1 2)

 Then, step D6 and subsequent steps are executed. If the NO route is taken in step D6 and step D7, or if the YES route is taken in step D8, the process of step D10 is performed and the total CLPO loss The operation of equation (8) for numbers is performed.

 In this way, the operators in the editing operation circuit 30 are shared in a time-sharing manner, and the operator memory is shared, thereby increasing the circuit scale.

 In addition, since the circuit scale can be reduced in this way, the size of the substrate can be reduced, and the cost of the PCB can be reduced. In addition, since the circuit size can be reduced, low power consumption is promoted and resources can be used effectively.

 Also, in this way, even if the item power of the editing operation can be increased by one, it is possible to respond by changing the calculation schedule, so that the operation can be performed without reducing the performance, thereby increasing power consumption. And IJ points that can suppress the increase in circuit size and enable efficient use of memory.

 (C) Other

 The present invention is not limited to the embodiments described above, and can be carried out in various modifications without departing from the spirit of the present invention.

In the first embodiment, the processing time for each of VC, NVP, and UVP is divided equally, but this time interval can be variously changed according to the design policy. Similarly, the processing time for Nakahiro and the end point and the multiple performances are divided into MCSN, TUC O, TUC 0/1, BEDC, and the number of transmitted cells. The measurement and the divided force The time interval can be changed variously according to the design policy.

 Furthermore, the block size determination circuit 21 can extend the power described for 16 bits to 32 bits or the like. It does not hinder.

 The values such as the threshold value in the second embodiment are merely examples, and are variously changed according to the design policy. In addition, the order in which the operations are performed can be changed.

 In addition, the function of the memory is realized by a register in addition to the memory. Production available

 i: As described above, according to the present invention, the circuit size can be reduced and power consumption can be reduced. Therefore, in the field of performance management using an ATM, the device can be made compact. And contribute to lower power consumption. In addition, since the introduction of the performance management arithmetic unit will be promoted, itH property of communication using the ATM network can be improved.

Claims

The scope of the claims

1. A performance management arithmetic circuit that performs a performance management operation to determine the quality of the road for each connection path or terminal point and relay point! ^ For each iJ information, and outputs multiple performance results.

 An element operation unit that is capable of time-dividing at least the performance management operation related to the connection and the end point / neutral performance 1-period operation related to the identification information within one cell processing time, and calculating the performance substitution result of Fiber 1 ^ 1 ( 10) Performance, characterized by being configured with

2. The element operation unit (10)

 A skin calculation data selection output unit (11) capable of selectively outputting a plurality of extraction calculation data by an external signal;

 It is output from the calculation data selection output section (11)! A performance calculation unit (12, 19) that performs a performance calculation using a small number of skinned calculation data and selectively outputs the performance result of (2) by an external signal;

 A control unit (13) for receiving a user cell and inputting a scheduling signal to the skin calculation data selection output unit (11) and the performance management calculation unit (12, 19). The performance management arithmetic circuit according to claim 1, characterized in that:

3. The skin calculation data selection output unit (11)

 A plurality of holding units (1 1 a, l i b, 11 c) capable of holding a number of β calculation data;

 It is connected to each of the number holding units (11a, lib, 11c), and can selectively output bell-number extraction data according to the scheduling signal input from the control unit (13). 3. The performance management arithmetic circuit according to claim 2, further comprising a selection unit (lid).

4. The control unit (13) Power \ A cell type determination unit (13a) capable of determining a connection route of the received user cell and outputting a cell type determination result;

 A counting unit (13b) capable of measuring the number of user cells to be received and outputting a measured value and outputting a time-division signal;

 The cell type determination unit (13) is connected to the cell type determination unit (13a) and the number unit (13b). A second aspect of the present invention, comprising: a management operation unit (12, 19); and a pre-processing scheduling unit (13c) capable of inputting the scheduling signal. 2. The performance management arithmetic circuit according to 1.

5. The performance management calculation unit (12, 19)

 The control unit (13) at least one first holding unit (14a, 14b, 14c, 14d, 14e, 19 a)

 Connected to each of the at least one (14a, 14b, 14c, 14d, 14e, 19a) of £ Lh to perform a performance management operation using the data to be operated; An adder (15, 19b) that can output each of the performance results;

 The performance of the acknowledgment output from the impeachment calculation unit (15, 19 b) Multiple operations that can ^^ the Hffi result; ¾ (parent unit (16 a, 16 b, 16 c, 16 d, 16 e , 19 c).

6. A block length setting unit (21b) for setting a block length that is the number of user cells transmitted and received in the transmission, the number of user cells being a power of two,

 A transmission cell number holding unit (16c) capable of measuring the number of user cells to be received and displaying the value in multiple bits;

A block length setting section (21b) a lower bit extracting section for reading a required number of bits of the transmission cell number holding section (16c) from the set value, and (21a) A match detection unit (21c) capable of detecting that the number of user cells transmitted from one path power matches the block length based on the result read by the lower bit extraction unit (21a). A block length determination circuit characterized by comprising:

7. An operation circuit for receiving a performance monitoring cell having information on the quality of fei ^ and outputting a plurality of evaluation values for the fe¾ path,

 A plurality of calculation units (32) capable of selectively outputting a plurality of values of the ^ ^, and a value of the difference between the value of the user cell loss and the number of erroneous insertions of the user cell. A control unit (33) capable of inputting a scheduling signal to the ten arithmetic unit (32) using the value and the value of the difference between the value of the user cell number with all the user cells and the value of the user cell number with priority )

 10. A knitting operation circuit, wherein a plurality of values for the path are time-divisionally operated within one cell processing time.

8. ¾10 arithmetic unit (32)

 In this case, a performance basket (31a, 31b, 31c, 31d, 31e, 31f, 31g) capable of respectively obtaining a plurality of fffi values,

 A first selection unit (35a) that can selectively output the fHffi value of the thigh for the 5 ° in a time-sharing manner according to the scheduling signal input from the control unit (33); The first selection unit (35a) performs an arithmetic operation using the output value of the force output from the adder unit (35b) that can output the result of class S, and outputs the result. (35b) Multiple operations that can hold the results of the multiple performance statements of force, etc. {Students (36 &, 36b, 36c, 36d, 36e, 36f, 36g) 8. The circuit according to claim 7, wherein the circuit comprises:

9. The control unit (33)

 A dozen or so parts (33b) capable of measuring the number of received user cells and outputting a measurement value and outputting a time-division signal,

Loss threshold allowed for the number of losses in the user cell and erroneous insertion of the user cell The threshold comparison result is obtained by using the value of the insertion threshold allowed for the number of cells, the value of the deviation, and one of the difference of the total number of user cells and the difference of the number of user cells with priority. A threshold comparison result output unit (39) that can be output by a time-division signal input from the

 Using the threshold comparison result and the time-division signal, the I-th arithmetic unit (32) is provided with a time-division signal output unit (33c) capable of inputting a scheduling signal. 10. The editing circuit according to claim 7, wherein:

10. The threshold comparison result output unit (39)

 Using one of a difference value between the scheduled transmission number and the transmitted number for the total number of user cells and a difference value between the scheduled transmission number and the transmitted number for the priority user cell number, 1¾110 Several parts (33b) a difference that can be selectively output as a difference report by the time-division signal input from the camera [blue report output section (39d, 39e, 39f);

 A ^ determination unit (39 g) which can determine whether the ^ green report is positive or negative and output it as ^ ^ ^;

An iron threshold (39a) capable of holding a threshold value of the broken fiber of the user cell, an erroneous insertion threshold holding unit (39b) capable of holding a threshold value of the number of erroneous insertions of the user cell, and A ^second selection unit (39c) capable of time-dividing one of the threshold value of the loss number of the current cell and the threshold value of the number of erroneous insertion of the user cell according to the time division and selectively outputting as threshold information. ) When,

 A comparing unit that compares the threshold information with the difference information and outputs the threshold comparison result (3

9. The statistical arithmetic circuit according to claim 9, wherein the statistical arithmetic circuit is configured to include: 9h).