JPH01108646A - Template matching circuit - Google Patents

Template matching circuit

Info

Publication number
JPH01108646A
JPH01108646A JP62265736A JP26573687A JPH01108646A JP H01108646 A JPH01108646 A JP H01108646A JP 62265736 A JP62265736 A JP 62265736A JP 26573687 A JP26573687 A JP 26573687A JP H01108646 A JPH01108646 A JP H01108646A
Authority
JP
Japan
Prior art keywords
data
memory
memory access
input
matching
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
JP62265736A
Other languages
Japanese (ja)
Other versions
JPH0586572B2 (en
Inventor
Koji Komatsu
宏二 小松
Shinichi Yoshida
芳田 真一
Soichi Miyata
宗一 宮田
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sharp Corp
Original Assignee
Sharp Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sharp Corp filed Critical Sharp Corp
Priority to JP62265736A priority Critical patent/JPH01108646A/en
Priority to US07/260,061 priority patent/US5113339A/en
Publication of JPH01108646A publication Critical patent/JPH01108646A/en
Priority to US07/779,805 priority patent/US5317756A/en
Publication of JPH0586572B2 publication Critical patent/JPH0586572B2/ja
Priority to US08/208,032 priority patent/US5392405A/en
Granted legal-status Critical Current

Links

Landscapes

  • Memory System (AREA)
  • Computer And Data Communications (AREA)

Abstract

PURPOSE:To accelerate template matching by setting the execution time of a pipeline processing at each stage less than a transmission delay time per stage of data transmission path. CONSTITUTION:In a memory access control circuit 7, only one of memory access parts in memory readout parts 2A and 2B and memory write parts 5A and 5B is permitted to make access to a memory by a memory access permission signal. The permission of memory access is issued preferentially to the memory access part 2A, 2B, 5A, or 5B generating the memory access, and when plural number of memory access are superposed, priority is attached on the memory access part 2A, 2B, 5A, or 5B from which the request of the memory access is issued earlier. Thus, by controlling the data transmission path by a memory access permission signal for the request of the memory access in the period of memory access, which makes the memory access by another data wait. In such a way, it is possible to evade the simultaneous memory access by plural data.

Description

【発明の詳細な説明】 〈産業上の利用分野〉 この発明はメモリを用いてテンプレートマツチングを行
う回路に関する。
DETAILED DESCRIPTION OF THE INVENTION <Industrial Application Field> The present invention relates to a circuit that performs template matching using a memory.

〈従来の技術〉 従来、単一人力ボートから連続的に入力されるデータに
よりマツチングメモリをアクセスした場合、パイプライ
ン処理を行ってメモリのアクセスを密にしシステムの処
理能力を上げている。
<Conventional Technology> Conventionally, when matching memory is accessed using data continuously input from a single human-powered boat, pipeline processing is performed to access the memory densely and increase the processing capacity of the system.

〈発明が解決しようとする問題点〉 パイプライン処理でメモリのアクセスを行ってもアクセ
スを密にするには限界があり、マツチングメモリ部での
処理能力がシステム全体の処理能力を規定する。本発明
では前記の問題点を解決し、高速なテンプレートマツチ
ングを実現することを目的としている。
<Problems to be Solved by the Invention> Even when memory is accessed by pipeline processing, there is a limit to how dense the access can be, and the processing capacity of the matching memory section determines the processing capacity of the entire system. The present invention aims to solve the above-mentioned problems and realize high-speed template matching.

〈問題点を解決するだめの手段〉 本発明では、データ伝送路をパイプラインステージに分
割し、データ伝送路を伝送するデータに対して、マツチ
ングメモリからのデータの読み出し、読み出しデータと
当該入力データの比較、比較結果による前記各データの
処理、マツチングメモリへのデータの書き込みをそれぞ
れのノ々イブラインステージで順に実行するように構成
し、更に各パイプラインステージで実行される処理に要
する時間をデータ伝送路の1段当たりの伝送遅延時間以
下にし、当該データ伝送路の入力データに対して前記処
理に必要な段数のデータ伝送路でノくイブライン処理す
る。少なくとも一つの前記ノくイブライン処理を行うデ
ータ伝送路に対して並列に入力されるデータから発生す
る単一のマツチングメモリに対するメモリアクセス要求
をメモリアクセス制御回路で順序付けし同時に複数のメ
モリアクセスが起こらないように制御する。
<Means to Solve the Problem> In the present invention, the data transmission path is divided into pipeline stages, and the data transmitted through the data transmission path is read from the matching memory, and the read data and the input data are matched. Comparison of data, processing of each data based on the comparison results, and writing of data to the matching memory are configured to be executed in order at each pipeline stage, and furthermore, the processing required for the processing executed at each pipeline stage is The time is set to be equal to or less than the transmission delay time per stage of the data transmission line, and the input data of the data transmission line is processed through the data transmission line of the number of stages necessary for the processing. A memory access control circuit orders memory access requests for a single matching memory generated from data input in parallel to at least one data transmission line that performs the above-mentioned no-blind processing, and prevents multiple memory accesses from occurring at the same time. control so that it does not occur.

〈作 用〉 前記各段のパイプライン処理の実行時間をデータ伝送路
の1段当たりの伝送遅延時間以下に収めることにより、
メモリアクセスが重ならない限りデータの伝送は乱され
ず、データ伝送路の能力で決まる時間間隔でデータを入
力ポートから入力でき、メモリアクセスが重なった場合
でもメモリアクセス制御回路によりデータ伝送路が制御
され、実行されているメモリアクセスが終了するまで該
メモリアクセスが待たされ処理に矛盾を生じない。
<Function> By keeping the execution time of the pipeline processing at each stage within the transmission delay time per stage of the data transmission path,
As long as memory accesses do not overlap, data transmission will not be disrupted, and data can be input from the input port at time intervals determined by the data transmission path's capacity.Even if memory accesses overlap, the data transmission path is controlled by the memory access control circuit. , the memory access being executed is delayed until the memory access is completed, so that no inconsistency occurs in the processing.

メモリのアクセスを要求するデータが並列に入力される
ことにより、従来の単一人力からのデータによりメモリ
アクセスを行う方式よりメモリアクセスを密に出来る。
By inputting data requiring memory access in parallel, memory access can be made more dense than in the conventional method of accessing memory using data from a single person.

〈実施例〉 第1図は本発明の2並列の場合の一実施例の構成図であ
り、データ伝送路IA、IB、メモリ読み出し部2A、
2B、比較判定部8A、、3B、データ処理部4A、4
B1メモリ書き込み部5A。
<Embodiment> FIG. 1 is a block diagram of an embodiment of the present invention in the case of two parallel systems, in which data transmission paths IA and IB, memory reading section 2A,
2B, comparison/judgment section 8A, 3B, data processing section 4A, 4
B1 memory writing section 5A.

5B、マツチングメモリ6、メモリアクセス制御回路7
で構成されている。  − データ伝送路IA又はIBに入力されたデータに対して
、メモリ読み出し部2A又は2Bではデータの一部或い
は全部からなる識別子に対応するアドレスのデータをマ
ツチングメモリ6から読み出し、比較判定部3A又は3
Bでは読み出されたデータと、入力された当該データの
各識別子を比較し各データの関係を判定する。データ処
理部4A又は4Bでは前記比較判定部3A又は3Bで得
られた判定結果に基づき各データを処理し、更にメモリ
書き込み部5A又は5Bでは前記判定結果に基づき前記
の何れかのデータをマツチングメモリ6の当該データの
識別子に対応するアドレスに書き込む、或いはマツチン
グメモリ6のデータを消去する。
5B, matching memory 6, memory access control circuit 7
It consists of - With respect to the data input to the data transmission path IA or IB, the memory reading section 2A or 2B reads data at the address corresponding to the identifier consisting of part or all of the data from the matching memory 6, and the comparison judgment section 3A or 3
In B, the read data is compared with each identifier of the input data to determine the relationship between each data. The data processing section 4A or 4B processes each data based on the judgment result obtained by the comparison judgment section 3A or 3B, and the memory writing section 5A or 5B matches any of the above data based on the judgment result. Write to the address corresponding to the identifier of the data in the memory 6, or erase the data in the matching memory 6.

入力データ一つに対して前記の一連のパイプライン処理
が順に起こり、メモリアクセスの要求が重なることは無
いが、同一データ伝送路に連続してデータが入力された
場合、或いは異なるデータ伝送路に並列にデータが入力
された場合、メモリアクセス要求が重なる場合がある。
The above series of pipeline processing occurs sequentially for one piece of input data, and memory access requests do not overlap, but if data is input consecutively to the same data transmission path or to different data transmission paths. When data is input in parallel, memory access requests may overlap.

メモリアクセス制御回路7ではメモリ読み出し部2A、
2B。
In the memory access control circuit 7, a memory reading section 2A,
2B.

メモリ書き込み部5A、5Bのメモリアクセス部の内1
ケ所だけにメモリアクセス許可信号(読みだし許可信号
A / B 、又は書き込み許可信号A / B )に
よりメモリのアクセスを許可する。メモリアクセスの許
可はメモリアクセス要求が発生したメモリアクセス部2
A、又は2B、又は5A、又は5Bが優先され、複数の
メモリアクセスが重なった場合はメモリアクセス要求が
早く起こったメモリアクセス部2A、又は2B、又は5
A、又は5Bから順に優先される。メモリアクセスを許
可されたメモリアクセス部はメモリをアクセスしている
期間を示すメモリアクセス実行信号(読み出し実行信号
A / B 、書き込み実行信号A/B)を発生し、メ
モリアクセス期間中は他のメモリアクセス部のメモリア
クセスを許可せず、メモリアクセスが終了次第、他のメ
モリアクセス要求が起こったメモリアクセス部2A、又
は2B、又は5A、又は5Bのメモリアクセスを許可す
る。このようにメモリアクセス要求に対してメモリアク
セス期間中、データ伝送路をメモリアクセス許可信号に
より制御し、他のデータによるメモリアクセスを待たせ
ることにより、同時に複数のメモリアクセスが起こるこ
とを回避する。
One of the memory access parts of the memory write parts 5A and 5B
Access to the memory is permitted only to these locations using a memory access permission signal (read permission signal A/B or write permission signal A/B). Permission for memory access is granted by the memory access unit 2 where the memory access request occurred.
Priority is given to A, 2B, 5A, or 5B, and if multiple memory accesses overlap, the memory access unit 2A, 2B, or 5 where the memory access request occurred earlier
Priority is given in order from A or 5B. A memory access unit that is permitted to access the memory generates a memory access execution signal (read execution signal A/B, write execution signal A/B) indicating the period during which the memory is being accessed, and during the memory access period, other memory The access unit is not permitted to access the memory, and as soon as the memory access is completed, the memory access unit 2A, 2B, 5A, or 5B that has received another memory access request is permitted to access the memory. In this manner, the data transmission path is controlled by the memory access permission signal during the memory access period in response to a memory access request, and by making memory access by other data wait, it is possible to avoid multiple memory accesses occurring at the same time.

第2図は、第1図の比較判定部3Aとデータ処理部4A
、比較判定部3Bとデータ処理部4Bを同じパイプライ
ンステージで実行させた場合の構成図である。図示はし
々いが、パイプライン処理の実行時間によりデータ伝送
路の1段当たりの伝送遅延時間に収まる範囲で複数のパ
イプライン処理を一つのパイプラインステージで実行さ
せたり、前記遅延時間に収まらない処理を複数のパイプ
ラインステージに分割する構成が実現できることは明ら
かである。
FIG. 2 shows the comparison/judgment section 3A and data processing section 4A in FIG.
, is a configuration diagram when the comparison/determination unit 3B and the data processing unit 4B are executed in the same pipeline stage. Although the illustration is difficult, it is possible to execute multiple pipeline processes in one pipeline stage within the range of the transmission delay time per stage of the data transmission path, depending on the execution time of the pipeline processing, or to execute multiple pipeline processes in one pipeline stage within the range of the transmission delay time per stage of the data transmission path. It is clear that it is possible to implement a configuration in which processing that does not exist is divided into multiple pipeline stages.

〈発明の効果〉 以上のように本発明を用いることにより、テンプレート
マツチングを高速に実現できる。
<Effects of the Invention> By using the present invention as described above, template matching can be realized at high speed.

【図面の簡単な説明】[Brief explanation of the drawing]

第1図及び第2図は、本発明の一実施例を示す構成図で
ある。第1図、第2図において、IA。 IBはデータ伝送路、2A、2Bはメモリ読み出し部、
3A、3Bは比較判定部、4A、4Bはデータ処理部、
5A、5Bはメモリ書き込み部、6はマツチングメモリ
、7はメモリアクセス制御回路であり、読み出し実行信
号A / B及び書き込み実行信号A/Bをメモリアク
セス実行信号、読み出し許可信号A/B及び書き込み許
可信号A / Bをメモリアクセス許可信号、読み出し
要求信号A / B及び書き込み要求信号A / Bを
メモリアクセス要求信号とする。
FIGS. 1 and 2 are configuration diagrams showing one embodiment of the present invention. In FIGS. 1 and 2, IA. IB is a data transmission path, 2A and 2B are memory reading units,
3A and 3B are comparison and determination sections, 4A and 4B are data processing sections,
5A and 5B are memory write units, 6 is a matching memory, and 7 is a memory access control circuit, which converts read execution signals A/B and write execution signals A/B into memory access execution signals, read permission signals A/B, and write. The permission signals A/B are used as memory access permission signals, and the read request signals A/B and write request signals A/B are used as memory access request signals.

Claims (1)

【特許請求の範囲】[Claims] 1、少なくとも一つの入力ポートを持ち、これらの入力
ポートから入力されるデータに対してデータの一部或い
は全部からなる識別子に対応するアドレスを持つマッチ
ングメモリと、前記マッチングメモリから当該データの
前記識別子に対応するアドレスの読み出しを行うメモリ
読み出し部と、読み出されたデータと当該データとの比
較判定を行う比較判定部と、比較結果に基づき各データ
の処理を行うデータ処理部と、前記マッチングメモリの
当該データの前記識別子に対応するアドレスへ前記デー
タ処理部の処理結果の書き込みを行うメモリ書き込み部
と、これらの処理をパイプライン処理しつつデータ伝送
が可能な前記入力ポートに対応する一つ以上のデータ伝
送路と、少なくとも一つのデータ伝送路上を伝送される
二つ以上のデータに対して前記単一のマッチングメモリ
のアクセスが同時に発生しないように前記データ伝送路
上のデータの伝送を調停するメモリアクセス制御回路と
を備え、少なくとも一つのデータの入力に対してそれぞ
れのデータのパイプライン処理を並列に実行し、高速に
テンプレートマッチングを行うことを特徴とするテンプ
レートマッチング回路。
1. A matching memory having at least one input port and having an address corresponding to an identifier consisting of part or all of the data for data input from these input ports, and a matching memory that has an address corresponding to an identifier consisting of a part or all of the data, and a matching memory that has an address corresponding to an identifier consisting of a part or all of the data input from the input port, and a matching memory that has an address corresponding to an identifier consisting of a part or all of the data input from the input port, a memory reading unit that reads an address corresponding to the data, a comparison judgment unit that makes a comparison judgment between the read data and the data, a data processing unit that processes each data based on the comparison result, and the matching memory. a memory writing unit that writes a processing result of the data processing unit to an address corresponding to the identifier of the data, and one or more corresponding to the input ports capable of data transmission while pipeline processing these processes. a data transmission path, and a memory that arbitrates the transmission of data on the data transmission path so that accesses of the single matching memory do not occur simultaneously for two or more data transmitted on at least one data transmission path. What is claimed is: 1. A template matching circuit, comprising: an access control circuit; the circuit executes pipeline processing of each data in parallel for at least one data input, and performs template matching at high speed.
JP62265736A 1987-10-20 1987-10-20 Template matching circuit Granted JPH01108646A (en)

Priority Applications (4)

Application Number Priority Date Filing Date Title
JP62265736A JPH01108646A (en) 1987-10-20 1987-10-20 Template matching circuit
US07/260,061 US5113339A (en) 1987-10-20 1988-10-20 Data processor for detecting identical data simultaneously coexisting in a plurality of data sections of data transmission paths
US07/779,805 US5317756A (en) 1987-10-20 1991-10-21 Data processor for detecting identical data coexisting in a plurality of data section of data transmission paths
US08/208,032 US5392405A (en) 1987-10-20 1994-03-09 Data processor for detecting identical data coexisting in a plurality of data sections of data transmission paths

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP62265736A JPH01108646A (en) 1987-10-20 1987-10-20 Template matching circuit

Publications (2)

Publication Number Publication Date
JPH01108646A true JPH01108646A (en) 1989-04-25
JPH0586572B2 JPH0586572B2 (en) 1993-12-13

Family

ID=17421279

Family Applications (1)

Application Number Title Priority Date Filing Date
JP62265736A Granted JPH01108646A (en) 1987-10-20 1987-10-20 Template matching circuit

Country Status (1)

Country Link
JP (1) JPH01108646A (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS6079469A (en) * 1983-10-07 1985-05-07 Fujitsu Ltd Detecting device of ambiguity in vocabulary of japanese language sentence
JPS60101683A (en) * 1983-11-08 1985-06-05 Fujitsu Ltd Matching circuit

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS6079469A (en) * 1983-10-07 1985-05-07 Fujitsu Ltd Detecting device of ambiguity in vocabulary of japanese language sentence
JPS60101683A (en) * 1983-11-08 1985-06-05 Fujitsu Ltd Matching circuit

Also Published As

Publication number Publication date
JPH0586572B2 (en) 1993-12-13

Similar Documents

Publication Publication Date Title
US5201040A (en) Multiprocessor system having subsystems which are loosely coupled through a random access storage and which each include a tightly coupled multiprocessor
JPH01108646A (en) Template matching circuit
JPH04120652A (en) Parallel processors
JPS6381557A (en) Dual port memory
JPH01293456A (en) Shared memory
JPH0368045A (en) Main memory control system
JPS6130300B2 (en)
JPS61165144A (en) Memory access controlling system
JPS62272352A (en) Memory control circuit
JPH06314251A (en) Scsi data transfer device
JPH04104355A (en) Multi-processing system
JPH01191964A (en) Transfer method for memory bus data
JPH01108647A (en) Template matching circuit
JPH0253161A (en) Memory access system
JPS61206065A (en) Multi-processor system
JPS6260329A (en) Digital signal detection system
JPH0236443A (en) System for controlling expansion storage
JPH01108642A (en) Data processor
JPS62274346A (en) Memory access circuit
JPH01121955A (en) Storage access controller
JPS5931737B2 (en) Multiple control data processing system
JPH06282456A (en) Artificial oiperation device
JPS642985B2 (en)
JPS63156268A (en) Microprocessor system
JPS635460A (en) Parallel computer

Legal Events

Date Code Title Description
EXPY Cancellation because of completion of term