JPH0964889A - Data comparison / transfer device - Google Patents

Abstract

(57) Abstract: An object of the present invention is to improve the processing capability of an apparatus for transferring transfer source data to a transfer destination while comparing transfer source data and transfer destination data. A transfer source data storage unit 105 stores transfer source data 103 extracted from the transfer source device 101. The transfer destination data access unit 106 is the transfer destination device 1
Access 02. The transfer destination data storage means 107
The transfer destination data 10 extracted from the transfer destination device 102 by the access by the transfer destination data access unit 106
4 is stored. The comparison means 108 compares the transfer source data 103 stored in the transfer source data storage means 105 with the transfer destination data 104 stored in the transfer destination data storage means 107. The control unit 109 changes the transfer source device 101 to the transfer destination device 10 according to the comparison result of the comparison unit 108.
2 controls the transfer of the transfer source data 103 to the second transfer source.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for transferring transfer source data to a transfer destination while comparing transfer source data and transfer destination data.

[0002]

2. Description of the Related Art As a broadband connectionless data service constructed for the purpose of realizing a connection between LANs, for example, SMDS (switched multi-megabit) is used.
data service) is known.

Since a broadband connectionless data service processes connectionless data of a large number of subscribers, it is important to process subscriber information efficiently.

FIG. 8 shows a general configuration diagram of the SMDS system. SMDS is an L3-PDU (Level 3 Protoc
a variable-length message called "Ol Data Unit"
-A called PDU (Level 2 Protocol Data Unit)
It is a high-speed connectionless data service that transfers by dividing it into fixed-length packets similar to cells in the TM (asynchronous transfer) system. In addition to the user information that is actually transferred data, the L3-PDU is added with a source address SA (Source Address) and a destination address DA (Destination Address) as address information. Further, the same identification information called a message identifier is added to L2-PDUs forming the same L3-PDU.

The information from the subscriber is SNI (Subscriber
Network Interface). Then, the subscriber is provided with a DS1 subscriber line 801 having a transmission rate of 1.5 Mbps and a DS3 subscriber line 803 having a transmission rate of 45 Mbps, for example, as physical transmission paths. The subscriber terminal uses one or more L2-PDUs over SMDS for L2-PD.
It is divided into U and the L2-PDU is further divided into DS1 subscriber line 8
01 or the data format on the DS3 subscriber line 803, and then output to each subscriber line.

The DS1 subscriber line 801 is a DS1 termination unit 80.
2, the DS3 subscriber lines 803 are each terminated to a DS3 termination 804, where the L2-PDU is retrieved.
A semi-fixed VCI (virtual channel identifier) and VPI (virtual path identifier) corresponding to SMDS are added to the L2-PDU from the subscriber. The DS1 terminal unit 802 and the DS3 terminal unit 804 use the semi-fixed VCI + VPI.
Is extracted, and the L2-PDU is extracted.
The VCI + VPI of the PDU is replaced with a semi-fixed VCI + VPI corresponding to each terminal, and the L2-PD
U is output toward the switch unit 808.

The switch unit 808 is actually an AT.
It is often an M switchboard, and in this case, the L2-PD
U is converted into an ATM cell and then output to the switch unit 808. Since the L2-PDU and the ATM cell are fixed-length packets having almost the same data length, the AT
It can be said that the M switch and the SMDS have a high affinity.

The DS1 terminal unit 802 or the DS3 terminal unit 80
The L2-PDU (ATM cell) output from the multiplexer 4 is the multiplexer / demultiplexer 805, 806, or 8
It is input to the switch unit 808 via 07. The switch unit 808 is provided for each DS1 termination unit 802 or DS3 termination unit 8
L with semi-fixed VCI + VPI corresponding to 04
The 2-PDU is transferred to the SMDS line part (LP) 813.

The LP 813 is connected to the switch unit 808 by a highway 815 having a transmission speed of 622 Mbps. This LP 813 has a maximum of, for example, 64 subscriber interfaces (SNs) accommodated by the station including the LP 813.
I) can be handled. Then, as many LPs 813 as the number of stations are connected to the switch unit 808 via the highway 815.

The LP 813 can recognize from which DS1 terminating unit 802 or DS3 terminating unit 804 the L2-PDU is input by discriminating the semi-fixed VCI + VPI attached to the received L2-PDU. it can. Then, the LP 813 extracts the source address SA from the message (L3-PDU) transferred using the plurality of L2-PDUs, and calculates the degenerate source address corresponding to the address. Also, LP
813 extracts the destination address DA from the above message. Then, when the above-mentioned destination address DA is addressed to the inside of the station, the LP 813 determines the VCI + VPI of each L2-PDU in which the above-mentioned message is divided and stored.
Is a semi-fixed VCI + V of the DS1 terminal 802 or DS3 terminal 804 in the station corresponding to the destination address DA.
Change to PI. On the other hand, when the above-mentioned destination address DA is addressed to the outside of the station, the LP 813 calculates the LP of the destination station and then calculates the VCI + VPI of each L2-PDU in which the above-mentioned message is divided and stored. V corresponding to the degenerate source address and destination station LP813
Change to CI + VPI. After that, the LP 813 outputs each L2-PDU to the switch unit 808 again via the interface 814 and the highway 815.

Each switch module (not shown) in the switch unit 808 switches the input L2-PDU at high speed. In this case, L input from LP813
The 2-PDU is a switch unit 808, a multiplexer / demultiplexer 8 if it is destined for the station.
DS1 termination unit 80 via 07, 806, or 805.
2 or DS3 termination unit 804 and DS1 subscriber line 801
Alternatively, it is transferred to the subscriber in the station via the DS3 subscriber line 803. Also, the L2-PDU is transferred to the LP 813 of the destination station via the multiplexer / demultiplexer 810, the terminating circuit (FINF) 811, and the interoffice relay line 812 when it is addressed to the outside of the station. It

After the LP 813 of the destination station recognizes that the message is addressed to the station by discriminating the destination address DA of the message transferred by using the plurality of L2-PDUs received by the LP 813, VCI + VP of each L2-PDU in which the message is stored
I is the DS1 in the station corresponding to the above-mentioned destination address DA
Semi-fixed VCI of termination unit 802 or DS3 termination unit 804
It is replaced with + VPI and outputs each L2-PDU to the switch unit. Each of these L2-PDUs has a switch unit 80.
8, via the DS1 terminating unit 802 or the DS3 terminating unit 804, and the DS1 subscriber line 801 or the DS3 subscriber line 803 to the subscriber in the station.

As described above, the DS1 subscriber line 80 shown in FIG.
DS1 terminator 802 or DS3 subscriber line 8 terminating 1
The DS3 terminating unit 804 that terminates 03 does not process subscriber information individually, but the LP813 collectively processes all subscriber information, thereby reducing the hardware cost of the entire exchange and maintaining it. It can be done easily. Further, the control device 809 may access the LP 813 via the control line 817 in call processing or the like, and does not need to individually access each of the DS1 terminating unit 802 and the DS3 terminating unit 804. The processing can be performed efficiently.

Here, in a broadband connectionless data service such as SMDS, a function of performing charging processing for each subscriber is required. In the billing process, it is necessary to collect, for each subscriber, subscriber information generated as a result of communication performed by each subscriber.

In this case, it is necessary to charge for the communication data provided by the network. Therefore, the charging process is performed on the destination station side. Specifically, LP8 shown in FIG. 8 in the destination station
At 13.

As a method that can be easily considered as a billing process, in the LP 813, the L which is addressed to the station received by the LP 813 is received.
There is a method of collecting and accumulating subscriber information (billing data) necessary for billing processing for each of the source addresses SA stored in 3-PDU. But L
The source address SA stored in the 3-PDU has an information amount of 64 bits, for example, and the number of source addresses SA is 2 ⁶⁴ . Therefore, it is actually impossible for the LP 813 to store the charging data for such a number of source addresses SA.

Therefore, the present applicant has filed Japanese Patent Application No. 5-2334.
In the patent application of No. 61, the source address SA extracted from the received L3-PDU is degenerated to a few thousand degenerate source addresses represented by data having a bit number of, for example, ten and several bits, and the degenerate transmission is performed. We have proposed a method to collect and store accounting data for each original address.

As a result, L can be formed by using an ordinary memory element.
The billing process in P813 can be realized. Here, the following sequence can be considered as a specific processing sequence of the charging process in the LP 813.
That is, first, the process of collecting the accounting data for each degenerate transmission source address is executed, for example, every one minute. Next, 15 minutes of the charge data for each 1 minute is totaled. Then, the control device 809 (see FIG. 8) is notified of the aggregated billing data for the 15 minutes.

FIG. 9 is an explanatory diagram of a conventional charging process that can be considered to realize the above-described charging process sequence. The configuration shown in FIG. 9 is provided in LP813 of FIG.

The billing data collection unit 901 includes a switch unit 8
The source address SA extracted from the L3-PDU received from 08 is degenerated to, for example, several thousand degenerate source addresses RDSCA represented by data having a bit number of about ten and several bits, and the degenerate source address RDS.
The accounting data DATA is collected for each CA. Then, the billing data collection unit 901 alternately displays the degenerate source address RDS on the collection surface 902 (0 surface) and the collection surface 902 (1 surface) which are memory elements, alternately every minute.
Collected billing data DATA with CA as address
Write.

In parallel with this, the MPU 905 is connected to the bus 9
ROM connected via bus 908 while using RAM 907 connected via 08 as work memory
0 plane and 1 according to the control program stored in 906.
A memory element for storing the charging data DATA for one minute collected on the collecting surface 902, which is not written by the charging data collecting unit 901, of the charging data DATA for 15 minutes via the bus 908. Read surface 903
(The thick line A in FIG. 9).

When the charge data DATA for the past 15 minutes is accumulated on the reading surface 903, the MPU 905 causes the reading surface 9 to read.
The entire contents of 03 are collectively transferred to the notification surface 904, which is a memory element having the same storage capacity as the reading surface 903, via the bus 908 (thick line B in FIG. 9).

Finally, the controller 809 of FIG.
The contents of the notification surface 904 shown in FIG.

[0024]

In the conventional charging processing sequence shown in FIG. 9 described above, the MPU 905 is used.
1 minute of billing data DA collected on the collection surface 902
When the TA is transferred to the reading surface 903 that accumulates the billing data DATA for 15 minutes, the MPU 905 calculates each billing data DATA on the collecting surface 902 in order to calculate the tally for 15 minutes. It is necessary to execute a process of searching the reading surface 903 for the billing data DATA corresponding to DATA. And MPU905
When the billing data DATA corresponding to one billing data DATA on the collecting surface 902 is found on the reading surface 903, the billing data DATA on the collecting surface 902 is found on the reading surface 903. If it is not found, the charging data DATA on the collecting surface 902 is added to the new charging data area on the reading surface 903.

Thus, the collecting surface 902 to the reading surface 90
The transfer process of the charging data DATA to the data No. 3 is not a simple transfer process, but a process that involves searching for the charging data DATA on the reading surface 903. In this case, the maximum number of messages that generate the billing data DATA collected on the collection surface 902 matches the maximum number of the degenerate source address RDSCA as described above, and is about several thousand messages. Therefore,
During the one minute when the processing surface of the collection surface 902 is switched, the reading surface 9 of the billing data DATA for several thousand messages is read.
The above search process must be executed on 03.

However, in the software processing executed by the MPU 905, it may not be possible to execute such processing in one minute, and in the worst case,
All billing data DA from collection surface 902 to read surface 903
There is a problem in that the TA cannot be transferred, and communication may occur in which charging is impossible.

More specifically, the maximum number of sets of billing data DATA collected on the collection surface 902 is n sets. Then 32
It is assumed that a data area for four addresses is searched for for one set of billing data DATA with one bit as one address. The data amount of one set of the billing data DATA may be four addresses or more. Further, the processing clock width of the MPU 905 is set to 40 × 10 ⁻⁹ seconds, and the MPU 905 sets the reading surface 9
It takes 20 clocks on average to retrieve the data of one address on 03, that is, 40 × 10 ^-9 × on average
Suppose it takes 20 seconds.

In this case, a maximum of (n ×
14) Different charging data DATA for the set is read out on the reading surface 903.
Accumulated in the last 15 minutes, and in the first minute, collecting surface 9
Assuming that n sets of billing data DATA equal to the maximum number of sets have been collected in 02, one set of billing data DATA on the collecting surface 902 has four addresses on the reading surface 903 for the past 14 minutes. In order to retrieve the charging data DATA of, the time required by the following formula 1 is required.

[0029]

[Equation 1] (40 × 10 ⁻⁹ × 20) seconds × 4 addresses × (n
Therefore, in order to retrieve the charging data DATA for the past 14 minutes on the reading surface 903 for each of the n sets of charging data DATA on the collecting surface 902, the following formula 2 is used. It will take time to get there.

[0030]

[Equation 2] {(40 × 10 ⁻⁹ × 20) seconds × 4 addresses ×
(N × 14) pairs} × n pairs Since the upper limit of this time is 1 minute as the processing surface of the collecting surface 902 is switched as described above, the formula 2 is set to 60 seconds, and thus the collecting surface 902 is set. The maximum number n of billing data DATA that can be collected is as shown by the following equation (3).

[0031]

[Equation 3] n≈1157 sets That is, the processing clock width is 40 × 10 ⁻⁹ seconds (processing clock frequency is 25 MHz) and the processing bit width is 32 bits M.
Even if the PU 905 is used, there is a restriction that it is impossible to collect more than n = 1157 sets of billing data DATA.

Such a problem is not limited to the billing data processing device, but is generally applied to a device for transferring the transfer source data to the transfer destination while comparing the transfer source data and the transfer destination data. It is a problem that can occur.

It is an object of the present invention to improve the processing capability of a device for transferring the transfer source data to the transfer destination while comparing the transfer source data and the transfer destination data.

[0034]

FIG. 1 is a block diagram of the present invention. The present invention compares the transfer source data 103 of the transfer source device 101 with the transfer source data 103 stored in the transfer source device 101 and the transfer destination data 104 stored in the transfer destination device 102. Assuming a device to transfer to.

First, the transfer source data storage means 105 stores the transfer source data 103 extracted from the transfer source device 101. The transfer destination data access unit 106 accesses the transfer destination device 102.

The transfer destination data storage means 107 stores the transfer destination data 104 extracted from the transfer destination device 102 by the access by the transfer destination data access means 106.

The comparison means 108 compares the transfer source data 103 stored in the transfer source data storage means 105 with the transfer destination data 104 stored in the transfer destination data storage means 107.

The control means 109 changes from the transfer source device 101 to the transfer destination device 1 according to the comparison result by the comparison means 108.
Control the transfer of the transfer source data 103 to 02. While the transfer destination data access unit 106 accesses the transfer destination device 102 at high speed, the transfer destination data 104 is extracted to the transfer destination data storage unit 107, and the transfer source data 103 extracted to the transfer source data storage unit 105 is compared. By the means 108 performing a high-speed comparison, the transfer source device 1
The transfer of the transfer source data 103 from 01 to the transfer destination device 102 is controlled.

[0039]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will now be described in detail with reference to the drawings. <Type of Billing Data DATA> First, before describing the specific configuration of the present embodiment, the type of billing data DATA processed in the present embodiment will be described.

As the charging data DATA, destination address DA, source address SA, group address, carrier information, condition code, L2-PDU number (segment count), L3-PDU number (packet count), SNI address, etc. are considered. However, in this embodiment, for the sake of simplicity of explanation, the LP installed in each destination station
813 (FIG. 8), the degenerate source address RD in which the source address SA in which communication is actually performed is degenerated.
For each SCA, the degenerate destination address RDA obtained by degenerating the destination address DA, the carrier information Carrier, the source address SA, the number of L2-PDUs, and the number of L3-PDUs are collected.

First, in this embodiment, basically, the charging data DATA is the source address SA (L3-PDU) indicating the source that transferred the message (L3-PDU) to the destination address DA in the station processed by the LP 813. (Stored in the header part of the). However, since the source address SA has a large bit number of 64 bits, the source address SA extracted from the received L3-PDU has about ten and several bits in accordance with the technology disclosed in the patent application of Japanese Patent Application No. Hei 5-233461. The degenerate source address RDSCA having the number of bits is degenerated, and the charging data DATA is collected for each of the degenerate source address RDSCA. This degenerate source address RDS
CA is determined as follows. That is, LP813
(FIG. 8) can handle up to, for example, 64 SNIs accommodated by the station in which it is included. Therefore, 1
The number of source addresses SA stored in one LP 813 is
This is the number of sender subscribers who have transferred the message to any of the destination addresses DA corresponding to a maximum of 64 SNIs. This number is much smaller than the number of addresses represented by the number of bits (for example, 64 bits) of the source address SA, and it can be estimated that about several thousand is sufficient as described later. Therefore, the input source address SA is sequentially held in thousands of source address registers provided in the LP 813, and the register number is set as the degenerate source address RDSCA. As a result, the number of sets of billing data DATA to be collected becomes a few thousand sets indicated by this degenerate source address RDSCA. As a result, the 64-bit source address SA can be degenerated to the degenerate source address RDSCA of about a dozen or more bits. However, in order to determine the actual SA, the original 64-bit source address SA itself is also collected.

Next, the degenerate destination address RDA, which is a part of the billing data DATA, is the degenerate one of the destination address DA which is the address of the destination subscriber within the station processed by the LP 813. As the destination address DA,
There are individual addresses, which are subscriber addresses of normal destinations, and group addresses. The group address is an address defined to realize the broadcast communication. Now 1
A plurality of individual addresses are defined for one group address, and the message sent by storing the group address as the destination address DA is a plurality of individual addresses defined for the group address by the function of the network. Are sent simultaneously to subscribers with. The destination address DA is stored in the header part of the L3-PDU. Here, the destination address DA has a data width of 64 bits, but the destination address DA accommodated in one station can be represented by a much smaller data width. Therefore, similarly to the case of the source address SA, only the destination address DA processed by the LP 813 in one station is given a serial number in advance, and the serial number is set as the degenerate destination address RDA. This gives 64
This means that the bit destination address DA can be degenerated to a degenerate destination address RDA of about several bits to a dozen or more bits.
In this case, the destination address DA accommodated in one station is 1.
Once set, it is basically fixed, so the relationship between the degenerate destination address RDA and the destination address DA is also fixed.
Therefore, the original destination address D is used as the charging data DATA.
There is no need to collect A.

Next, the carrier information, which is a part of the billing data DATA, is information indicating the type of transmission path. Refer to this information, for example, when there are multiple types of transmission lines by multiple transmission line providers and the charging system is different for each transmission line, for example, in a network that provides long-distance calls called inter-LATA communication. To be done. In this case, the LP 813 (FIG. 8) of each station receives the message,
If it is not a message addressed to the inside of the station, after adding a header for the station to the message, the message is transferred to the gateway device using a plurality of cells. The gateway device determines the station having the area code by discriminating the higher area code of the destination address DA in the message, and transfers the message to the station. In this case, the above-mentioned message is transferred to the transmission path of the transmission path provider if there are a plurality of transmission paths by the transmission path providers in the middle of the route to which the message is transferred. On the other hand, the LP 813 that has received the message from the transmission path of another transmission path carrier (the interoffice relay line 812 of FIG. 8)
Carrier information indicating from which transmission line carrier the message was received is added to the inter-office header area of the received message. The carrier information added in this way is collected in the charging process.

Finally, the number of L2-PDUs and the number of L3-PDUs that are part of the billing data DATA are the L2-PD transferred in the communication corresponding to the billing data DATA.
The numbers of U and L3-PDUs are shown, and in this embodiment, they are collected in units of 1 minute and are aggregated in units of 15 minutes as described later. <Collection Number of Billing Data DATA> Next, in the present embodiment, the billing data DATA is collected in units of 1 minute. Here, the maximum number of billing data DATA groups collected in 1 minute, that is, degeneration. Estimate the number of destination addresses RDA. This number is equal to the number of communications made to the destination address DA, which is the address of the destination subscriber in the station processed by the LP 813.

As communication using the SMDS service, data communication of general users and RIP (dynamic route information which is communicated between networks when computer networks are interconnected using the SMDS service). Routing Information Protocol) There is data communication. In RIP, routing information for route control is transferred from a route information server in one network to a route information server in another network, for example, 30
Broadcast every second.

From this point of view, first, one LP8
Estimate the number of data communication of general users that 13 processes.
Here, one LP in the range where the loss rate of communication data is 10 ⁻⁴ or less per unit time (for example, per minute)
The number of communications processed by 813 is based on the following prerequisites.
It is estimated using a stochastic process (Poisson arrival process). Number of SNIs per station: 1024 Number of SNIs per 1LP813: 64 Bandwidth per 1LP813: 622Mbps Number of individual addresses per SNI: 16 When estimated under these assumptions, one LP81
The number of data communication of general users processed by 3 was 2048.

Next, the RIP processed by one LP 813
Estimate the number of data communication. This communication is performed for each subscriber device. Therefore, the following preconditions are assumed. Number of SNIs per 1LP813: 64 Band per 1LP813: 622 Mbps Subscriber bandwidth: 40.7 Mbps (DS3): 1.416 Mbps (DS1) Number of subscriber devices: 16 (DS3): 1 (DS1) Also, 1 When the LP813 uses the entire bandwidth of 622 Mbps, the maximum number of DS3 lines processed by one LP813 is 14 and the maximum number of DS1 lines is 36.
It will be a book. Therefore, based on the above prerequisites, one LP8
The maximum number of subscriber units with which 13 communicates is:

[0048]

[Equation 4] 16 units x 14 (DS3) + 1 unit x 36 (D
S1) = 260 Further, the probability that a certain subscriber unit simultaneously sends a message to all the other twelve subscriber units is 10 ⁻¹⁰ or less. Assuming that this number is the minimum unit to which the subscriber unit must connect, the minimum number of members in RIP data communication is 12. Therefore, 260 shown by the equation 4
The number of communication of RIP data by all the subscriber devices of the unit is
It becomes 260 × 12 = 2860.

From the above estimation, the number of communications processed by one LP 813 is equal to the number of data communications of general users 2048 and R.
Total number of IP data communications 2860 2048 + 2860
= 4908. From this, the maximum number of sets of billing data DATA collected in one minute is set to 6144 (2048 × 3) sets by adopting values with good delimitation.

As described above, this charging data DATA
The number of sets is a number that cannot be processed by the conventional technique shown in FIG. In this embodiment, as described above, the maximum is 6144.
The transfer processing of the billing data DATA from the collecting surface where the billing data DATA forming a set is collected to the reading surface is made possible by executing the retrieval of the billing data DATA on the reading surface with a dedicated search LSI. is there.

The performance required for this search LSI will be roughly calculated below. First, the maximum number of sets of billing data DATA collected on the collecting surface is n sets. Next, assuming that 64 bits are one address, it is assumed that a data area for two addresses is searched for for one set of accounting data DATA. The data amount of one set of the charging data DATA is three addresses in this embodiment, as will be described later. It is also assumed that it takes an average of 40 × 10 ⁻⁹ seconds for the search LSI to search the data of one address on the reading surface.

In this case, at most (n ×
14) Assuming that different sets of billing data DATA are accumulated in the reading surface, and n sets of billing data DATA equal to the maximum number of sets are collected in the collecting surface in one minute of the last 15 minutes. In order to search the charge data DATA for the past 14 minutes by using two addresses for each reading surface for the above-mentioned set of charge data DATA, time required by the following formula 5 is required. Becomes

[0053]

## ^EQU00005 ## ^{40.times.10.sup.-9} seconds.times.2 addresses.times. (N.times.14) sets Therefore, for each of the n sets of charging data DATA on the collecting surface, the charging data DA for the past 14 minutes on the reading surface
In order to retrieve each TA, the time required by the following formula 6 is required.

[0054]

[Equation 6] {40 × 10 ⁻⁹ seconds × 2 addresses × (n × 14)
Set} × n set The upper limit of this time is 1 minute for the processing surface of the collecting surface to switch as described above, so by setting Equation 6 = 60 seconds, the billing data that can be collected on the collecting surface. The maximum number n of groups of DATA is as shown by the following Expression 7.

[0055]

(7) n≈7319 sets As a result, by using the search LSI, it becomes possible to collect the 6144 sets of accounting data DATA described above. <Specific Configuration of Embodiment> FIG. 2 is a specific configuration diagram of an embodiment of the present invention. This configuration is provided within LP813 of FIG.

The billing data collection unit 201 includes a switch unit 8
The source address SA extracted from the L3-PDU received from 08 is degenerated to, for example, a maximum of 6144 degenerate source addresses RDSCA, and the degenerate source address RD.
The billing data DATA is collected for each SCA. As described above, the collected charging data DATA includes the degenerate destination address RDA, carrier information Carrier,
Source address SA, number of L2-PDU and L3-PDU
Is a number. Then, the billing data collection unit 201 alternately collects the collection planes 202 (0
Collected data D and the collection surface 202 (one surface) with the degenerate sender address RDSCA as an address.
Write ATA.

In parallel with this, the MPU 205 is
ROM connected via bus 208 while using RAM 207 connected via 08 as work memory
According to the control program stored in 206, 0 side and 1 side
A memory element for storing the charging data DATA for one minute collected on the collecting surface 202, which is not written by the charging data collecting unit 201, of the charging data DATA for 15 minutes via the bus 208. Read surface 203
Transfer to. In this case, the search LSI 209 searches for the charging data DATA on the reading surface 203 (see FIG. 2).
The thick line A), and the MPU 205 uses the search result to execute the transfer process. This is a major feature related to the present invention.

When the charging data DATA for the past 15 minutes is accumulated on the reading surface 203, the MPU 205 determines that the reading surface 2
The entire contents of 03 are collectively transferred via the bus 208 to the notification surface 204 which is a memory element having the same storage capacity as the reading surface 203 (thick line B in FIG. 2).

Finally, the controller 809 of FIG.
The contents of the notification surface 204 shown in FIG. Next, FIG. 3 is a configuration diagram of the search LSI 209 of FIG. A collection plane data storage unit 303 that stores data from the collection plane 202, a read plane address output unit 301 that sequentially specifies addresses for the read plane 203, and a read plane 2
03 search end address (read side end address WTEN
DP) to stop the read surface address output unit 301, a read surface control unit 302, a read surface data storage unit 304 for storing data from the read surface 203, and a collection surface data storage unit 303 for collecting data. A comparator 305 that compares the data from the plane 202 with the data from the read plane 203 stored in the read plane data storage unit 304, and a comparator 305.
And a comparison result determination control unit 306 that determines the comparison result in 1.

In the embodiment having the above structure, MP
The operation of the U205 and the search LSI 209 transferring the billing data DATA on the collection surface 202 to the reading surface 203 will be described below with reference to FIGS.

First, FIG. 7 is a processing flow corresponding to a program executed as interrupt processing every time the charging data collecting unit 201 switches the collecting surface 202 between the 0th surface and the 1st surface. This processing flow is such that the MPU 205 uses the RAM 207 connected via the bus 208 as a work memory and the ROM 206 connected via the bus 208.
It is realized as an operation for executing the interrupt program stored in.

When the charging data collecting unit 201 of FIG. 2 comes to the timing of switching the collecting surface 202 between the 0th surface and the 1st surface every one minute, the MPU 205 executes step 701 of FIG.
It is determined whether or not the pointer PADDR indicating the number of sets of billing data DATA collected on the collection surface 202 of the transfer target is 0. The pointer PADDR indicates the final pointer position of the set of billing data DATA collected by the billing data collection unit 201, and when the value of the pointer PADDR is 0, the billing data DATA collected in 1 minute.
Indicates that there is none.

When the determination in step 701 is NO,
The MPU 205 ends the interrupt processing of FIG. 7 as it is.
When the determination in step 701 is YES, the MPU 20
5 is step 702, and the current process is continuous 15
It is determined whether the process is the first minute of the minutes.

When the MPU 205 determines in step 702 that the current process is the first one-minute process, it does not need to retrieve the data on the reading surface 203, and therefore, in step 703, the process shown in FIG. As described above, all the accounting data DATA up to the address indicated by the pointer PADDR stored in the collection surface 202 which is the transfer target is DMA (direct memory access) transferred to a continuous area from the start address of the reading surface 203. . After that, MPU2
Reference numeral 05 is the charging data DATA transferred to the reading surface 203.
Is set as the read surface end address WTENDP, and the next address is set as the new write address NEWWWTP. Then, the interrupt processing of FIG. 7 is ended. The billing data DATA is composed of three addresses (one address is 64-bit data) as one set, and the degenerate destination address RDA that constitutes the upper 32 bits of the first address and the carrier information that constitutes the lower 32 bits of the first address. From the carrier, the source address SA forming 64 bits of the second address, the L2-PDU number forming the upper 32 bits of the third address, and the L3-PDU number forming the lower 32 bits of the third address Composed.

On the other hand, when the MPU 205 determines in step 702 that the current process is not the first one-minute process (the process from the second minute to the fifteenth minute), first in step 704, FIG. , The degenerate destination address RDA, the carrier information Carrier, and the source address SA that form the charge data DATA at the beginning of the collection surface 202 that is the transfer target are searched for via the bus 208.
9 in the collection plane data storage unit 303 (see FIG. 3). These data are data for two addresses.

The MPU 205 then proceeds to step 705.
Then, the write end signal WTEND is sent to the search LSI 209.
Output to Step 706 is actually a search LS.
This is a search process according to I209. After that, MPU20
No. 5 waits until SERCH in the search signal changes from 1 to 0.

Upon receiving the write end signal WTEND from the MPU 205, the search LSI 209 receives the MPU 20.
5, the degenerate destination address RDA written in the collection side data storage unit 303, the carrier information Carrier, and the source address SA, and the degenerate destinations included in each charging data DATA set already transferred to the reading side 203. The data set of the address RDA, the carrier information Carrier, and the source address SA are sequentially compared from the start address of the read surface 203 to the address indicated by the read surface end address WTENDP.

Specifically, the read surface address output unit 301 in the search LSI 209 outputs a read surface address having the initial address of the read surface 203 as an initial value to the read surface 203 and, as a result, outputs from the read surface 203. The read data set is stored in the read surface data storage unit 304. Read surface 203
In the above, the charging data DATA is 3 as shown in FIG.
The addresses are stored as one set, but the degenerate destination address RDA and carrier information Carrier stored in the first address and the source address SA stored in the second address are data to be compared. It is a group. Therefore, the read side address output unit 301 first specifies the first address, and as a result, the degenerate destination address RDA and carrier information Ca output from the read side 203.
The rrier is stored in the read surface data storage unit 304, the second address is then designated by the read surface address output unit 301, and as a result, the source address SA output from the read surface 203 is also stored in the read surface data storage unit 304. To be done. Then, the read surface address output unit 301 skips the third address.

Next, the comparator 305 determines the data set stored in the acquisition surface data storage unit 303 and the readout surface data storage unit 3
The data set stored in 04 is compared. When the comparator 305 detects a match, the comparison result determination control unit 306 outputs the start address of the charging data DATA which is the comparison target output by the reading surface address output unit 301 as the search address MADDR, and the same charging is performed. The data presence / absence signal FOUND is set to 1, and the searching signal SERCH is reset to 0.

On the other hand, when the comparator 305 detects a mismatch, the read surface address output unit 301 outputs the read surface address corresponding to the next accounting data DATA set to the read surface 203 as described above. As a result, the reading surface 2
The data set output from 03 is used as the reading surface data storage unit 30.
Store in 4.

Here, the reading surface control unit 302 has an MPU.
The reading surface end address WTENDP is set by 205. This read surface end address WTENDP indicates the final address of the accounting data DATA written on the read surface 203 in the previous interrupt processing of FIG.
Then, as a result of the reading surface address output unit 301 sequentially outputting the reading surface addresses, the last billing data DATA written in the reading surface 203 in the interrupt processing of the previous FIG.
When the read surface addresses exceed the read surface end address WTENDP, the read surface control unit 302 causes the read surface address output unit 301 to stop. In this case, the comparison result determination control unit 306 causes the same billing data presence / absence signal FOUND and the in-search signal SER.
Reset both CHs to 0.

When the in-search signal SERCH changes from 1 to 0, the MPU 205 makes an N determination in step 707.
It becomes O. As a result, the MPU 205 makes a step 708.
Then, it is determined whether or not the value of the same billing data presence / absence signal FOUND is 1.

When the value of the same billing data presence / absence signal FOUND is 1, that is, when the search LSI 209 detects a match and the determination in step 708 is YES, the MPU 205
5, in step 709, as shown in FIG. 5, the number of L2-PDU and the number of L3-PDU stored in the third address of the set of accounting data DATA starting from the search address MADDR output from the search LSI 209. And the L2-PD included in the charging data DATA on the collecting surface 202
The number of U and the number of L3-PDU are added respectively.

On the other hand, the same billing data presence / absence signal FOUND
Is 1, that is, if the search LSI 209 does not detect a match and the determination in step 708 is NO, M
In step 710, the PU 205 sets the set of billing data DATA on the collecting surface 202 into a region for three consecutive addresses from the address indicated by the new write address NEWWWTP on the reading surface 203, as shown in FIG. Forward. After that, the MPU 205, as shown in FIG. 6, charges data DA newly written on the reading surface 203.
The address next to the TA pair is newly written as a new write address N.
Reset as EWWTP.

The MPU 205 then proceeds to step 711.
Then, it is determined whether or not the set of billing data DATA on the collecting surface 202 is the final data. Specifically, MPU20
5 is a collection PAD output by the billing data collection unit 201
The final address is determined by DR.

If the determination in step 711 is NO, MP
In step 712, the U 205 collects, via the bus 208, the degenerate destination address RDA, the carrier information Carrier, and the source address SA that form the next billing data DATA of the collection surface 202 that is the transfer target via the bus 208 in the collection surface of the search LSI 209. Write to the data storage unit 303.

The MPU 205 then proceeds to step 705.
The following processing is repeated. <Other Embodiments> Although the embodiments described above relate to the billing data processing device in the SMDS service, the present invention is not limited to this, and the transfer source data and the transfer destination data are It can be similarly applied to an apparatus that transfers the data of the transfer source to the transfer destination while making a comparison.

In the above embodiment, the search LSI 2
09 is configured to set search data one data set at a time,
All the search data on the collecting surface 202 may be set in the search LSI 209 by having a memory having the same capacity as the above.

Further, the search LSI 209 is configured to execute only the search operation, but it may be configured to execute the transfer operation as well.

[0080]

According to the present invention, the transfer destination data access means fetches the transfer destination data into the transfer destination data storage means while accessing the transfer destination device at high speed, and the transfer fetched into the transfer source data storage means. Since the original data can be compared at high speed by the comparison means, for example, it becomes possible to improve the transfer function of the billing data in the SMDS service requiring real-time property.

[Brief description of drawings]

FIG. 1 is a block diagram of the present invention.

FIG. 2 is a configuration diagram of an embodiment of the present invention.

FIG. 3 is a configuration diagram of a search LSI.

FIG. 4 is an explanatory diagram of a first one-minute transfer process.

FIG. 5 is an explanatory diagram of a transfer process when the same data exists.

FIG. 6 is an explanatory diagram of a transfer process when the same data does not exist.

FIG. 7 is a diagram showing a processing flow of an MPU.

FIG. 8 is a general configuration diagram of an SMDS system.

FIG. 9 is an explanatory diagram of a conventional method.

[Explanation of symbols]

 101 Transfer Source Device 102 Transfer Destination Device 103 Transfer Source Data 104 Transfer Destination Data 105 Transfer Source Data Storage Means 106 Transfer Destination Data Access Means 107 Transfer Destination Data Storage Means 108 Comparison Means 109 Control Means

Claims (1)

[Claims] 1. A device for transferring transfer source data of a transfer source device to a transfer destination device while comparing transfer source data stored in the transfer source device and transfer destination data stored in the transfer destination device. A transfer source data storage unit that stores the transfer source data extracted from the transfer source device, a transfer destination data access unit that accesses the transfer destination device, and a transfer destination device by access by the transfer destination data access unit. Transfer destination data storage means for storing the transfer destination data retrieved from the transfer source data, and comparison means for comparing the transfer source data stored in the transfer source data storage means with the transfer destination data stored in the transfer destination data storage means. And control means for controlling the transfer of the transfer source data from the transfer source device to the transfer destination device according to the comparison result by the comparison means. Comparison / forwarding device data, characterized in that it comprises a.