JP2001045020A - Reply distribution type communication method and communication system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To avoid congestion in a communication system on the occurrence of a traffic flow in excess of a communication capacity of a transmission medium or a data reception capability or the like caused by transmission of a request by a plurality of nodes. SOLUTION: The communication system adopts the reply distribution type communication method where each of a plurality of nodes 3 stores an assigned specific value and a transmission timing of a reply for transmission of the reply is delayed by a transmission time obtained by multiplying a specific value such as a node number with 10 ms. Furthermore, the communication system is configured by connecting a plurality of nodes that delay a reply transmission timing by the transmission time obtained by multiplying the specific value such as the node number with 10 ms for the transmission of the reply in a way of enabling data transmission reception.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a distributed communication method for a communication system in which a plurality of nodes mutually transmit and receive data, and which is applicable to a disaster prevention system and the like, and a communication system thereof.

[0002]

2. Description of the Related Art Conventionally, various communication methods have been proposed and adopted. For example, Japanese Patent Laid-Open Publication No. Hei 5-212503 has a plurality of stations connected to a network and transmits a call from a calling station. In a broadcast communication system in which a plurality of called stations receiving the received broadcast request information transmit broadcast response information to the calling station, each called station has a different delay time from the other called stations. Thereafter, a broadcast communication method is described in which broadcast response information is transmitted, thereby suppressing an increase in local traffic on a network.

Japanese Patent Laid-Open Publication No. Hei 10-79753 discloses that a broadcast message transmitting station broadcasts delay type information for designating a waiting time until a broadcast message receiving station transmits a response message. The broadcast message receiving station inserts or adds the message in the message and sends it. The broadcast message receiving station receives the broadcast message, reads the delay type information in the message, sets the delay time of its own station, and sets it. A broadcast communication control method is disclosed in which a response message is transmitted after the elapse of the delay time, whereby the timing at which the response message is returned in a short broadcast message can be controlled.

[0004]

However, according to the methods described in the above publications, the operation in the case where a single node broadcasts is limited, and a plurality of nodes are required to perform the method in a case where a single node broadcasts. It cannot be applied to a communication system for transmitting and receiving data.

By the way, such a communication system includes:
There is a problem that data is lost when a plurality of nodes transmit.

FIG. 10 is an explanatory diagram of this problem. In this example, when one node transmits a request at time t1 and another node transmits a request at time t2, the time t2
In the subsequent section TM1, twice as much traffic as when a single node transmits a request will occur, so that data may be lost due to data collision, reception buffer leakage, or the like. As the number of nodes sending requests further increases, the traffic volume further increases,
The risk of data loss is further increased. In particular, in a distributed system in which each node operates autonomously and transmits data asynchronously, it is difficult to predict the transmission timing of other nodes. For this reason, when all nodes transmit requests at the same time, It is necessary to design a communication system in consideration of the following. However, in FIG. 10, the transmission signal other than the time points t1 and t2 is a response transmission signal.

[0007] In a communication system in which a plurality of nodes transmit and receive data to and from each other, a communication method for minimizing the delay of a request is also desired. Also, a communication method that minimizes the response delay is desired. In the latter case, if each node is provided with a parallel processing function, transmission of the next request becomes possible before collection of a response to a certain request is completed. To do so, more complicated transaction management functions are required, and the design of each node, and thus the communication system, becomes more complex. For this reason, designing without a parallel processing function becomes much easier, and in this case, it is required to minimize the response delay.

The present invention has been made in view of the above circumstances, and has been made in consideration of the above circumstances, and has been made in consideration of the case where traffic exceeding the communication capacity or data receiving capacity of a transmission medium occurs due to transmission of a request by a plurality of nodes. It is an object of the present invention to provide a response-distributed communication method capable of avoiding congestion.

[0009]

According to a first aspect of the present invention, there is provided a distributed response communication method for distributing a response for a communication system in which a plurality of nodes mutually transmit and receive data. In the method, each of the plurality of nodes holds an assigned unique value, and when transmitting a response, delays the transmission timing of the response according to the unique value.

According to this method, the transmission timing of the response from each node is delayed according to a unique value, so that the communication system in a case where traffic exceeding the communication capacity of the transmission medium or the data receiving capacity occurs. Can be avoided.

[0011] In the distributed response communication method according to claim 1, each of the plurality of nodes, when transmitting a request, includes, in header information, designation information of at least one destination node to which the request is to be transmitted. The request is transmitted as data together with this header information, and when the data is received, based on the header information obtained from this data,
Determine whether the received data is addressed to itself,
If the data is addressed to itself, a method of transmitting a response may be used (claim 2). According to this method, since a response is transmitted if the data is addressed to itself, redundant traffic can be suppressed, and the probability of data loss of a request due to signal collision or reception buffer overflow can be suppressed.

Further, in the response-distributed communication method according to claim 1, each of the plurality of nodes individually has a response transmission function and a request transmission function as a data transmission function,
When transmitting a response by the response transmission function, the transmission timing of the response is delayed according to the unique value, and when transmitting a request by the request transmission function, the request is transmitted without delaying the transmission timing. It may be a method (claim 3). According to this method, since the transmission timing of the request is not delayed, the request can be transmitted quickly.

In the response-distributed communication method according to claim 1, each of the plurality of nodes, when transmitting the response, determines the response of the response from a transmission speed of the transmission medium and a data length of the response. Calculate the stay time in the transmission medium, calculate the transmission time interval of the response required to avoid congestion of the communication system from the stay time, and determine the transmission order of the response from the unique value held by the node. Then, the response may be transmitted according to the transmission order and the transmission time interval (claim 4). According to this method, the transmission time interval of the response is calculated based on the transmission speed of the transmission medium and the data length of the response, so that the response can be returned with a minimum delay.

Further, in the distributed response communication method according to claim 1 or 3, a load on the communication system is calculated from a data reception time, and it is determined whether the load on the communication system is increased based on the load. If the load on the communication system is increasing, a method of extending the transmission time interval of the response may be used (claim 5). According to this method, if the load on the communication system increases, the transmission time interval of the response becomes longer, so that the probability of data loss of the request due to signal collision, reception buffer overflow, etc. can be suppressed.

Further, in the distributed response communication method according to claim 2, each of the plurality of nodes has a timer and stores in advance a list of nodes which are expected to return a response, and when transmitting the request, Start the timer,
The header information includes at least one destination node designation information to which the request is to be transmitted using the list, and the request is transmitted together with the header information. After that, a timeout occurs due to timing of the timer. If the response is received from all of the nodes expecting to return the response included in the header information before, the timer is stopped, and the response is received from all of the nodes expecting to return the response. In the case where a time-out occurs due to the counting of the timer, a method may be employed in which the request is retransmitted only to a node which has not returned a response among the nodes which are expected to return the response (claim 6). ).
According to this method, a request is retransmitted only to a node that does not return a response among nodes that expect a response, so redundant traffic is suppressed, signal collision, reception buffer overflow, and the like. , The probability of data loss of the request due to the request can be reduced.

In the response-distributed communication method according to claim 6, each of the plurality of nodes is configured such that, even if the request is retransmitted a predetermined number of times, no response is returned from the retransmission partner. Alternatively, a method may be adopted in which the partner node is treated as a failed node, and thereafter, the number of retransmissions for the failed node is made smaller than the number of retransmissions for a node not treated as the failed node (claim 7). According to this method, the number of retransmissions to the failed node is reduced, so that the next request can be transmitted immediately. That is, by shortening redundant waiting time due to repetition of retransmission to a node that cannot expect a response, prompt transmission of the next request becomes possible.

Further, in the response-distributed communication method according to claim 3, at least one of the plurality of nodes, which has a higher request for data transmission than other nodes, is set as a high-priority node. The communication system is set so that a response from the priority node is transmitted before a response from the other node, and no response is returned from the high-priority node to the other node. In this case, the method may be such that the collection of the response is interrupted and the request is retransmitted to the high priority node (claim 8). According to this method, the request to the high-priority node is given priority, so that the delay of the request to the high-priority node can be minimized.

Further, in the distributed response communication method according to claim 4, each of the plurality of nodes has a timer,
A method may be used in which the length of time measured by the timer is set to a value that can avoid congestion in the communication system according to the number of nodes connected to the communication system (claim 9).
According to this method, even if the number of nodes connected to the communication system changes, congestion of the communication system can be avoided. Further, it is possible to minimize the response delay according to the number of nodes connected to the communication system.

Further, in the distributed response communication method according to the ninth aspect, each of the plurality of nodes may set a timer setting time for response collection according to the number of the failed nodes. ). According to this method, redundant delay due to waiting for a response from the failed node can be suppressed, and the delay of the response can be minimized.

According to an eleventh aspect of the present invention, there is provided a communication system in which a plurality of nodes mutually transmit and receive data, wherein each of the plurality of nodes holds an assigned unique value and transmits a response. At this time, the transmission timing of this response is delayed according to the unique value.

In this configuration, since the transmission timing of the response of each node is delayed according to a unique value, congestion of the communication system occurs when traffic exceeding the communication capacity of the transmission medium or the data receiving capacity occurs. Can be avoided.

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is an explanatory diagram of a distributed response communication method according to a first embodiment of the present invention, and the first embodiment will be described below with reference to this drawing.

First, a communication system to which the response-based communication method according to the first embodiment is applied will be described together with an embodiment according to the communication system of the present invention.

FIG. 2 is a diagram showing an example of the configuration of the communication system, and FIG. 3 is a diagram showing an example of a frame format of data transmitted and received by the communication system. The communication system 1 shown in FIG. Network 2 and a plurality of devices connected to the network 2 (8 in the example of FIG. 2).
) Nodes 3. However, FIG.
Indicate the node number (the same applies to FIG. 1).

A frame F of data transmitted and received by the communication system 1 includes a header section F1 and a data section F2. The header section F1 includes a source node number F11,
Destination node / group number F12, destination destination designation F
13, a message type F14 of the request / response, and a transaction number F15.

FIG. 4 is a hardware configuration diagram of each node shown in FIG. 2, and FIG. 5 is a functional block diagram provided on the hardware, particularly on the control unit, and each node 3 is shown in FIG. Thus, it has a transmission unit 30 for transmitting data to the network 2 and a reception buffer memory (reception buffer) 31a, and is connected to the network 2 to receive data from the network 2. , A timer 32 for measuring time, a memory 33 for storing various data,
The receiving unit 31, the timer 32, the memory 33, and the like are connected to each other, and are configured by a control unit (CPU) 34 that performs various control processes for the node 3.

A control unit 3 operating according to a predetermined software program is provided on the hardware configured as described above.
5, a driver 300 for the transmission unit 30 and the reception unit 31 and an application task unit 301 for performing information processing for the data unit F2 as shown in FIG.
And a communication protocol layer unit 302 for processing a frame F to be transmitted / received based on the information in the header F1.

The communication protocol layer unit 302 includes a header analysis unit 316 described later as a function for performing various managements.
, A frame reception history management unit 303 that manages the reception history of frames in response to the notification from the application task unit 301 and notifies the application task unit 301, a system state management unit 304, and a response management unit 305 that manages responses to frame transmission.
And a transaction management unit 306 for designating the transaction number of the header F1. Then, the response management unit 305 and the transaction management unit 3
In addition to a queue 307 for notifying the application task unit 301 of information from 06, an event flag queue 308 is provided.

The communication protocol layer section 302 has a request transmission function as a request input queue 309 to be transmitted from the application task section 301 and a transaction management section 30 according to the format shown in FIG.
In addition to assembling the frame F by including the transaction number and the like designated by No. 6 in the header section F1 and including the request information obtained from the queue 309 in the data section F2, it also responds with the node number of at least one node designated as the transmission destination. A request frame creation unit 310 for notifying the management unit 305 and registering a request event in the queue 308 is provided with a queue 311 for taking in a frame obtained from the request frame creation unit 310 for transmission.
In addition, the communication protocol layer unit 302 includes, as a response transmission function, a response frame creation unit 312 that assembles a response frame F, and a queue 313 that captures a frame obtained from the response frame creation unit 312 for transmission. ing. A transmission delay unit 314a is provided at a subsequent stage of the queues 311, 313, and the frame from the queue 311 is immediately passed to the driver 300, while the queue 311, 313
A frame transmission unit 314 that passes the frame from the third through the transmission delay unit 314a, delays the transmission timing, and passes the frame to the driver 300 is created.

Further, the communication protocol layer unit 302
A queue 315 for inputting a frame received from the driver 300 and a header analyzing unit 316 for analyzing a header F1 in a frame obtained from the queue 315 are provided.

The processing function of the header analyzer 316 will be described in more detail. For example, referring to the destination node / group number F12 in the header F1, if it is not addressed to itself, the frame having the header F1 Is discarded.

If the request is made by referring to the message type F14 in the header part F1, if the request is not addressed to itself, a process of discarding the frame is performed by further referring to the destination designation F13. Done. On the other hand, when the frame is addressed to itself, a process of passing the frame or its header F1 to the response frame creating unit 312 is performed. As a result, a response frame is created by the response frame creation unit 312 and queued in the queue 313, and thereafter, the driver 3 is sent through the frame transmission unit 314.
00 will be passed. In addition, a process of registering a response event in the queue 308 and notifying the transaction management unit 306 of the received frame is performed. At this time, if the transaction management unit 306 determines that there is no double reception, a request message is notified to the application task unit 301 via the queue 307.

On the other hand, referring to the message type F14, if the response is a response, the response
A process of notifying the first information is performed. In this case, the response management unit 305 stops the response reception completion wait timer when all the responses have been returned from all the nodes that should receive the response, and performs a process of notifying the application task 301 of the normal termination. .

Here, the timer 32 shown in FIG. 4 is used as a communication event period timer, a response reception completion waiting timer, and a response transmission waiting timer.

The communication event period timer transmits a request when an event does not exist in the event flag queue 308, or receives a request irrespective of whether it is addressed to itself, or counts the time of the communication event period timer. If at least one event is queued in the queue 308 at the time of termination, it is activated and has no stop condition. At the time of timeout, a process for referring to the queue 308 is executed.

The response reception completion wait timer is set in the queue 308
If there is no event in the queue, or if a request is sent, or
If the event obtained from 08 is a request, it is activated, and thereafter, when all responses are received from all nodes to be returned, the response management unit 305 stops. In addition, at the time of timeout, if the number of retransmissions is less than the number of times, a node that has not returned a response is set in the destination destination designation F13 of the header F1 and retransmitted. When retransmission has been performed for the number of retransmissions, the node number of the node that has not returned a response is notified to the system state management unit 304, and the application task unit 30
1 is performed by the response management unit 305.

The response transmission waiting timer is used to transmit a request when there is no event in the queue 308 or when the communication event period timer finishes counting.
If the event obtained from is a response, it is activated and has no stop condition. At the time of timeout, a process of transmitting a response is performed.

Next, a description will be given of a response distributed communication method according to the first embodiment applied to the communication system 1 having the above configuration. In the first embodiment, a unique value is assigned to each of a plurality of nodes 3. When a response is transmitted from the frame transmission unit 314, for example, a node number is allocated and held, and the transmission timing of this response is set to the transmission delay unit 314a.
A method of delaying according to the unique value is adopted. The delay time is set to a value obtained by a node number × 10 ms. This delay processing is performed by the transmission delay unit 314a shown in FIG.

An operation example according to the above method will be described with reference to FIG. However, the transmission delay time in the transmission medium and the data processing time in the driver 300 are sufficiently short and can be ignored.

The node 3 [2] sends a request to the other nodes 3 [1], 3 [3] to 3 by, for example, group communication or multicast.
When transmitted to [8] (t10), from each of the other nodes,
Responses are sequentially returned after a delay time of a value obtained by node number × 10 ms. That is, the response from the node 3 [1] is received at a time t11 after a delay of about 10 ms, the response from the node 3 [3] is received at a time t13 after a delay of about 30 ms, and finally, the node 3 [8] ] Is received at time t18, approximately 80 ms later.
As described above, since the response transmission time of each node is uniquely determined in the communication event period, a case where traffic exceeding the communication capacity of the transmission medium or the data receiving capacity occurs due to transmission of the request by a plurality of nodes. It is possible to avoid congestion in a communication system such as However, it is assumed that the nodes 3 [1] to 3 [8] belong to the same group.

As shown in FIG. 1, if node 3 [3] sends a request to all other nodes at time t123 between time t12 and time t13, for example,
Since the response is sequentially returned from each of the other nodes based on 123, it is possible to prevent signal loss due to collision and also to avoid congestion in the communication system as described above.

FIG. 6 is an explanatory diagram of a distributed response communication method according to a second embodiment of the present invention, and the second embodiment will be described below with reference to FIG. This response-distributed communication method is also applied to the communication system 1 as in the first embodiment. In the second embodiment, when transmitting a request to each of the plurality of nodes 3, the specification information of at least one transmission destination node to which the request is to be transmitted is transmitted to the transmission destination partner specification F1 of the header F1.
3 together with the header part F1 and the request (data part F
When 2) is transmitted as data and data is received,
Based on the information in the header section F1 obtained from this data,
A method is adopted in which it is determined whether the received data is data addressed to itself, and if the data is addressed to itself, a response is transmitted.

An operation example according to the above method will be described with reference to FIG. Node 3 [2] sends a request to nodes 3 [1], 3
If it is transmitted to [8] (t20), nodes 3 [1] and 3
In [8], since the data received from the node 3 [2] is the data addressed to itself, a process of transmitting a response is performed at time t21 after a delay of approximately 10 ms and at time t28 after a delay of approximately 80 ms, respectively. In each of the nodes 3 [3] to 3 [7], since the data received from the node 3 [2] is not the data addressed to itself, the process of transmitting the response is not executed. Accordingly, the node 3 [2] receives the response from the node 3 [1] at the time point t21, and receives the response from the node 3 [8] at the time point t28. In this way, by eliminating a response from a node that does not need to transmit a request, redundant traffic can be suppressed, and the probability of losing requested data due to signal collision or reception buffer overflow can be reduced. Can be.

As shown in FIG. 6, if node 3 [3] transmits a request to nodes 3 [1] and 3 [2] at time t223 between time t22 and time t23, for example, Since the responses are sequentially returned only from the nodes 3 [1] and 3 [2] based on the time point t223, the loss of the signal due to the collision can be prevented.

In the first and second embodiments, as an operation example, when the node 3 [3] transmits a request during the collection of the response of the node 3 [2], the collection of the response to the node 3 [3] is performed. This is done in parallel with the collection of responses from node 3 [2],
However, the present invention is not limited thereto, and the collection of the response to the node 3 [3] may be started after the collection of the response of the node 3 [2] is completed. In this case, when the communication event period timer started at the time of transmitting the request from the node 3 [2] expires, a response to the node 3 [3] starts, and thereafter, the node number ×
Responses may be sequentially returned in accordance with 10 ms. FIG. 7 shows an operation example in this case. In FIG. 7, when the node 3 [2] transmits a request to, for example, all other nodes (t300), responses are sequentially returned from the other nodes 3 [1], 3 [3] to 3 [8]. Come (t301, t30
3,..., T308). Thereafter, when the communication event period timer started at the time of transmission of the request from the node 3 [2] expires (t309), the node 3 [3] receives the other node 3 [3].
Responses are sequentially returned from each of [1], 3 [2], 3 [4] to 3 [8] (t310, t311,..., T317).
This makes it possible to avoid congestion in the communication system.

Next, a description will be given of a distributed response communication method according to a third embodiment of the present invention. This response-distributed communication method is also applied to the communication system 1 as in the first embodiment. First
In the second embodiment, the response transmission time interval is a fixed value of 10 ms. However, in the third embodiment, a method of setting the response transmission time interval to a value suitable for the communication system 1 is adopted. That is, when transmitting a response to each of the plurality of nodes 3, a response time of the response in the transmission medium is calculated from the transmission speed of the transmission medium and the data length of the response frame. 1 to calculate the response transmission time interval necessary to avoid congestion, determine the transmission order of the response from the unique value held by the node 3, and determine the response according to the transmission order and the transmission time interval. The transmission method is adopted.

An example of calculating the response transmission time interval by this method will be described. For example, the transmission speed of the transmission medium is set to 80 k.
If the average data length of the request and response frames transmitted by each node 3 is 200 bytes, the response time in the transmission medium is 2.5 ms (= 200 by).
te / 80 kbps) is calculated. Subsequently, a transmission time interval is calculated by adding a predetermined value (≧ 0) to the stay time or multiplying the predetermined value (≧ 1). For example, if the predetermined value is zero, the response transmission time interval is 2.5 ms. As a result, congestion in the communication system can be avoided, and a response can be returned with a minimum delay.

Next, a distributed response communication method according to a fourth embodiment of the present invention will be described. This response-distributed communication method is also applied to the communication system 1 as in the first embodiment. 4th
In the embodiment, the load on the communication system 1 is calculated from the data reception time, and it is determined whether or not the load on the communication system 1 is increasing based on the load. A method of increasing the transmission time is adopted.

As a specific example, of the plurality of nodes 3 in the communication system 1, the node 3 serving as a master for determining the response transmission time is selected. In FIG. 2, for example, when node 3 [1] is a master, when data is received from driver 300 for node 3 [1], frame reception history is transmitted from header analysis unit 316 regardless of whether the data is addressed to itself. The management unit 303 is notified of the reception event (see FIG. 5), and calculates the response transmission time interval (for example, 2.5 ms) in the same manner as in the third embodiment.

Here, for example, if the monitoring result of the load state of the communication system 1 is to be verified every second, 400 frames (= 1 s / 2.5 ms) are sent to determine whether the load of the communication system 1 is increasing. Can be obtained as a reference value for determining. A value obtained by adding a predetermined value to 400 may be used, but here, for simplicity, 400 is used as it is as a reference value.

In this case, the node 3 [1] is made to calculate the number of frames received every second from the data reception time during the second as the load of the communication system 1, and the calculated number of frames is calculated based on the above-mentioned reference value. When the value is 400 or more, it is determined that the load on the communication system 1 is increasing.
If it is less than 400, it is determined that the load on the communication system 1 has not increased. Further, when the node 3 [1] determines that the load of the communication system 1 is increasing, the node 3 [1] is notified to all the other nodes so as to temporarily increase the response transmission time interval to 5 ms, Thereafter, if it is determined that the load on the communication system 1 has not increased, all the other nodes are notified to return the response transmission time interval to 2.5 ms. On the other hand, the transmission time interval is changed to 5 ms or 2.5 ms for all of the other nodes receiving these notifications in accordance with the notification from node 3 [1].

According to this specific example, if the load on the communication system 1 increases, the transmission time of the response becomes longer, so that the probability of losing the required data due to collision of signals or overflow of the receiving buffer can be suppressed. .

FIG. 8 is an explanatory diagram of a distributed response communication method according to a fifth embodiment of the present invention. The fifth embodiment will be described below with reference to FIG. This response-distributed communication method is also applied to the communication system 1 as in the first embodiment. In the fifth embodiment, for each of the plurality of nodes 3, a list of the nodes 3 that are expected to return a response is stored in advance, and when transmitting a request, a response reception completion wait timer is started and the list is used. In the header part F1, the request information is transmitted and the request is transmitted together with the header part F1. After that, before the time-out occurs due to the time measurement of the response reception completion wait timer. When the response is received from all of the nodes expecting the return of the response included in the header F1, the response reception completion wait timer is stopped, and the response is received from all of the nodes expecting the return of the response. Previously, if a timeout occurred due to the time measurement of the response reception completion wait timer, a node that did not return a response among nodes that are expected to return a response. How to perform a retransmission of the request is taken only for.

Next, an example of the operation according to the above method will be described.
When the node 3 [2] transmits a request to, for example, all the other nodes (t400), a response is sequentially returned from each of the other nodes. In the example of FIG. 8, the node 3 [3] is transmitted for some reason. ] Does not return a response, and node 3 [3]
Responses are sequentially returned from each of the other nodes 3 [1], 3 [4] to 3 [8] except for (t401,..., T408).

In this case, before receiving a response from all of the nodes that are expected to return the response, a timeout occurs due to the timing of the response reception completion wait timer (t409). At this time, the node 3 [2] transmits. The request is retransmitted with the destination designation F13 set to only the node 3 [3]. At this time, the destination node / group number F12 and the transaction number F15 are not changed. Thereafter, a response is returned from the node 3 [3] to the node 3 [2] at a time t412 after a lapse of 30 ms.

As described above, by performing retransmission only to a node that has not transmitted a response, a response from a node that does not need to be retransmitted is eliminated, so that redundant traffic can be suppressed. In addition, it is possible to suppress the probability of the loss of the requested data due to the collision of the signal or the overflow of the receiving buffer. In addition, when a request is received normally and a response to the request is lost due to a signal collision, if the request is retransmitted with a different destination node / group number F12 and transaction number F15, double reception occurs and the application runs away. Although there is a danger that the destination node / group number F12 and the transaction number F
By performing retransmission without changing No. 15, communication without such danger becomes possible.

In the fifth embodiment, if each of the plurality of nodes 3 does not return a response from the retransmission destination after performing the retransmission of the request a predetermined number of times, the destination node 3 May be treated as a failed node, and thereafter, the number of retransmissions for the failed node may be made smaller than the number of retransmissions for the node not treated as the failed node. For example, if the predetermined number is set to three times and the number of retransmissions to the failed node is set to one, the node 3 [2] becomes the node 3 [3] as in the operation of FIG.
If the response is not returned from the node 3 [3] even if retransmission is performed three times, the response management unit 305 in the node 3 [2]
A process of notifying that [3] is the failed node and reducing the number of retransmissions to node 3 [3] to one after this is performed.
Subsequently, a process of notifying the application task unit 301 of the abnormal termination is performed, whereby the communication process is completed. Thereafter, if the node 3 [2] transmits a request to the node 3 [3] and a response to the request is not returned, retransmission to the node 3 [3] is performed only once. Thereafter, if there is no response to the retransmission, the communication process is completed. As a result, redundant waiting time due to repetition of retransmission to a node for which no response can be expected can be reduced, and the next transmission can be quickly performed by the reduced time.

FIG. 9 is an explanatory diagram of a distributed response communication method according to a sixth embodiment of the present invention. The sixth embodiment will be described below with reference to FIG. This response-distributed communication method is also applied to the communication system 1 as in the first embodiment. In the sixth embodiment, at least one of the plurality of nodes 3 that has a higher request for data transmission than other nodes is set as a high-priority node, and a response from this high-priority node is the other node. The communication system 1 is set to be transmitted before the response from
If no response is returned from the high-priority node, a method is adopted in which the collection of the response is interrupted and the request is retransmitted to the high-priority node.

Next, an operation example according to the above method will be described.
However, in FIG. 9, it is assumed that the node 3 [1] has more requests for data transmission than the other nodes and is set as the high priority node.

When the node 3 [2] transmits the request to, for example, all the other nodes (t50), the response from the node 3 [1] set as the high-priority node is transmitted to the other nodes 3 [3] to 3 [ 8]
Responses are sequentially returned so as to be returned earlier than the response from. However, in the example of FIG. 9, the response is sent from the node 3 [1] set as the high priority node for some reason. It is not returned at time t51 when it should be received. In this case, the node 3 [2] suspends the collection of the response and, for example, by the time when the response is to be returned from the next node 3 [3],
The request is retransmitted to the node 3 [1] and the same request is retransmitted to the other nodes 3 [3] to 3 [8] (time t5).
2). At this time, the destination node / group number F12 and the transaction number F15 are not changed. As a result, runaway of the application due to double reception can be prevented.

Thereafter, in the example of FIG. 9, the response from the node 3 [1] set as the high-priority node is transmitted to another node 3 [3].
Responses are sequentially returned so as to be returned before the response from 3 [8] (t53, t55 to t60). As a result, it is possible to minimize the delay of the response from the high priority node.

Next, a distributed response communication method according to the seventh embodiment of the present invention will be described. This response-distributed communication method is also applied to the communication system 1 as in the first embodiment. Seventh
In the embodiment, the length of each time measured by the communication event period timer and the response reception completion waiting timer is determined for each of the plurality of nodes 3 in accordance with the number of nodes connected to the communication system 1. Is set to a value that can avoid the above.

Specifically, at the time of initial setting, the application task unit 3 stores the node number of its own node and the list (connection list) of the number of nodes and the node number of the communication system 1.
01 sets each node 3 so as to notify the communication protocol layer unit 302. Further, each node 3 is set so as to compare the node number of the own node with the node number in the connection list and calculate the response transmission order of the own node.

For example, the node number of the own node is 7, and the node numbers in the connection list are 1, 3, 4, 6, 9,
If the numbers are 10, 13, 16, and 19, the response transmission order of the own node is next to 1, 3, 4, and 6, which is the fifth. As described above, in view of the fact that the node number and the response transmission order do not always match, in the seventh embodiment, the communication event period timer and the response reception completion wait timer are used in accordance with the number of nodes connected to the communication system 1. The length of each time period is set adaptively. In this case, the number of nodes is 10, and the number of nodes is increased by two compared to the number of nodes 8 in FIG. 2. Therefore, in order to avoid congestion in the communication system, each time counted by the communication event period timer and the response reception completion timer is measured. The length is changed to 1.25 (= 10/8) times when the number of nodes is eight. This allows
The response delay time can be set to a minimum according to the number of nodes connected to the communication system.

In the seventh embodiment, when the above-described failure mode is detected, the number of detected failure nodes may be added to the setting of the length of each of the clock times.
In other words, according to the number of nodes connected to the communication system 1 and the number of failed nodes, the length of each time measured by the communication event period timer and the response reception completion wait timer is set to a value that can avoid congestion in the communication system 1. You may make it. For example, when the number of nodes connected to the communication system 1 is 10 and the number of failed nodes is 1, the length of each time counted by the communication event period timer and the response reception completion wait timer is set to the case where the number of nodes is 10. It may be changed to 0.9 times. Note that the node number of the own node is 7,
Node numbers in the connection list are 1, 3, 4, 6, 9, 1
When the node number of the failed node is 13, when the node number of the failed node is 13, the response transmission order of the own node remains at 5, and when the node number of the failed node is 4, the response transmission order of the own node. Changes to the fourth, so node number x 1
Needless to say, the delay time obtained by 0 ms changes.

[0066]

As is apparent from the above, according to the first aspect of the present invention, there is provided a communication method for distributing a response for a communication system in which a plurality of nodes mutually transmit and receive data. Each of the plurality of nodes holds the assigned unique value and, when transmitting a response, delays the transmission timing of the response according to the unique value, so that the communication capacity or the data reception capability of the transmission medium is reduced. It is possible to avoid congestion in the communication system such as when excessive traffic occurs.

According to a second aspect of the present invention, in the distributed response communication method according to the first aspect, when each of the plurality of nodes transmits a request, at least one destination to which the request is to be transmitted is provided. The node specification information is included in the header information, the request is transmitted together with the header information as data, and when the data is received, the received data is data addressed to itself based on the header information obtained from the data. It is also possible to determine whether the data is addressed to itself or to send a response (claim 2). According to this method,
If the data is addressed to itself, a response is transmitted, so that redundant traffic can be suppressed, and the probability of data loss of the request due to signal collision or reception buffer overflow can be suppressed.

According to the third aspect of the present invention, in the response-distributed communication method according to the first aspect, each of the plurality of nodes individually has a response transmission function and a request transmission function as a data transmission function. When transmitting a response by the response transmission function, the transmission timing of the response is delayed according to the unique value, and when transmitting the request by the request transmission function, the request is transmitted without delaying the transmission timing. Sending enables quick sending of requests.

According to a fourth aspect of the present invention, in the response-distributed communication method according to the first aspect, each of the plurality of nodes, when transmitting the response, transmits the transmission speed of the transmission medium and the response speed of the response. Calculate the residence time of the response in the transmission medium from the data length, calculate the transmission time interval of the response required to avoid congestion in the communication system from the residence time, and calculate the unique value held by the node. , The transmission order of the response is determined, and the response is transmitted according to the transmission order and the transmission time interval. Therefore, it is possible to return the response with a minimum delay.

According to a fifth aspect of the present invention, in the response-distributed communication method according to the first or third aspect, a load on the communication system is calculated from a data reception time, and the load on the communication system is calculated from the load. It is determined whether or not the request has been increased, and if the load on the communication system is increased, the transmission time interval of the response is lengthened, so that the probability of data loss of the request due to signal collision or reception buffer overflow is suppressed. be able to.

According to a sixth aspect of the present invention, in the response-distributed communication method according to the second aspect, each of the plurality of nodes has a timer and stores in advance a list of nodes that are expected to return a response. When transmitting the request, the timer is started, and the header information includes designation information of at least one destination node to which the request is to be transmitted using the list, and the request is transmitted together with the header information. Thereafter, before receiving a response from all of the nodes expecting the return of the response included in the header information before the timeout occurs due to the counting of the timer, while stopping the timer, If a time-out occurs due to the time measurement of the timer before receiving responses from all the nodes that are expected to return the Since the retransmission of the request only to the nodes that have not returned the response of the nodes,
Redundant traffic can be suppressed, and the probability of data loss of a request due to signal collision, reception buffer overflow, or the like can be suppressed.

According to a seventh aspect of the present invention, in the response-distributed communication method according to the sixth aspect, each of the plurality of nodes receives a response from the retransmission destination even if the request is retransmitted a predetermined number of times. Is not returned, the partner node is treated as a failed node, and the number of retransmissions for the failed node is made smaller than the number of retransmissions for the node not treated as the failed node. Request can be sent quickly.

According to an eighth aspect of the present invention, in the response-distributed communication method according to the third aspect, at least one of the plurality of nodes, which has a request for data transmission larger than other nodes, A high-priority node, the communication system is set such that a response from the high-priority node is transmitted before a response from the other node, and the high-priority If no response is returned from the node, the collection of the response is interrupted and the request to the high priority node is retransmitted, so that the delay of the request to the high priority node can be minimized.

According to a ninth aspect of the present invention, in the response-distributed communication method according to the fourth aspect, each of the plurality of nodes has a timer, and each of the nodes has a timer according to the number of nodes connected to the communication system. Since the length of time measured by the timer is set to a value that can avoid congestion in the communication system, congestion in the communication system can be avoided even if the number of nodes connected to the communication system changes. Further, it is possible to minimize the response delay according to the number of nodes connected to the communication system.

According to the tenth aspect, the ninth aspect is provided.
In the distributed response communication method described above, each of the plurality of nodes sets a timer setting time for response collection according to the number of the failed nodes, thereby suppressing redundant delay due to waiting for a response from the failed node. In addition, the response delay can be minimized.

According to the eleventh aspect of the present invention, there is provided a communication system in which a plurality of nodes mutually transmit and receive data, wherein each of the plurality of nodes holds an assigned unique value and transmits a response. When transmitting, the transmission timing of this response is delayed according to the unique value, so that it is possible to avoid congestion of the communication system such as when traffic exceeding the communication capacity or data receiving capacity of the transmission medium occurs.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a distributed response communication method according to a first embodiment of the present invention.

FIG. 2 is a diagram illustrating a configuration example of a communication system.

FIG. 3 is a diagram illustrating an example of a frame format of data transmitted and received in a communication system.

FIG. 4 is a hardware configuration diagram of each node shown in FIG. 2;

FIG. 5 is a functional block diagram provided on the hardware shown in FIG. 4, particularly on a control unit.

FIG. 6 is an explanatory diagram of a distributed response communication method according to a second embodiment of the present invention.

FIG. 7 is a diagram illustrating another example of response collection.

FIG. 8 is an explanatory diagram of a distributed response communication method according to a fifth embodiment of the present invention.

FIG. 9 is an explanatory diagram of a distributed response communication method according to a sixth embodiment.

FIG. 10 is an explanatory diagram of a problem such as loss of data when a plurality of nodes perform transmission.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Communication system 2 Network 3 Node 30 Transmitting unit 31 Receiving unit 32 Timer 33 Memory 34 Control unit

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Yutaka Nakao 1048 Kadoma Kadoma, Osaka Prefecture Matsushita Electric Works Co., Ltd. F-term (reference) 5K030 GA13 HC01 HC14 LA19 LD02 LD18 LE01 MB06 5K033 AA05 CB06 CB13

Claims (11)

[Claims] 1. A communication method for distributing a response for a communication system in which a plurality of nodes mutually transmit and receive data, wherein each of the plurality of nodes holds an assigned unique value, and When transmitting, a response-distributed communication method characterized in that the transmission timing of the response is delayed according to the unique value. 2. When transmitting a request, each of the plurality of nodes includes, in header information, designation information of at least one destination node to which the request is to be transmitted, and transmits the request as data together with the header information. When data is received, it is determined based on the header information obtained from the data whether or not the received data is data addressed to itself, and if it is data addressed to itself, a response is transmitted. 2. The response-distributed communication method according to claim 1, wherein: 3. Each of the plurality of nodes individually has a response transmission function and a request transmission function as a data transmission function, and when transmitting a response by the response transmission function, sets the transmission timing of the response to the unique transmission timing. 2. The distributed response communication method according to claim 1, wherein when the request is transmitted by the request transmitting function, the request is transmitted without delaying the transmission timing. 4. Each of the plurality of nodes, when transmitting the response, calculates a residence time of the response in the transmission medium from a transmission speed of the transmission medium and a data length of the response, The transmission time interval of the response required to avoid congestion of the communication system is calculated from the above, the transmission order of the response is determined from the unique value held by the node, and the transmission order of the response is determined according to the transmission order and the transmission time interval The response-distributed communication method according to claim 1, wherein the response is transmitted by using a response. 5. A load of the communication system is calculated from a data reception time, and it is determined whether or not the load of the communication system is increasing based on the load. 4. The response-distributed communication method according to claim 1, wherein a response transmission time interval is lengthened. 6. Each of the plurality of nodes has a timer and stores in advance a list of nodes that are expected to return a response. When transmitting the request, the node activates the timer and utilizes the list. The header information includes at least one designation information of the destination node to which the request is to be transmitted, and transmits the request together with the header information. If the response is received from all of the nodes expecting the return of the response included in, the timer is stopped, and before the response is received from all of the nodes expecting the return of the response, the timer is stopped. When a time-out occurs due to the above, the request is re-sent only to the nodes that did not return a response among the nodes that are expected to return the response. Response distributed type communication method according to claim 2, wherein the performing. 7. When each of the plurality of nodes does not return a response from a retransmission partner even after performing the retransmission of the request a predetermined number of times, treats the destination node as a failed node, 7. The response-distributed communication method according to claim 6, wherein the number of retransmissions for the failed node is made smaller than the number of retransmissions for a node not treated as the failed node. 8. A high-priority node at least one of which requests for data transmission are higher than other nodes, wherein a response from the high-priority node is received from the other node. The communication system is set so as to be transmitted before the response to the other node, if a response is not returned from the high priority node to the other node, the collection of the response is interrupted and the high priority 4. The distributed response communication method according to claim 3, wherein the request is retransmitted to the first node. 9. Each of the plurality of nodes has a timer, and according to the number of nodes connected to the communication system, a length of time measured by the timer is set to a value capable of avoiding congestion of the communication system. 5. The response-distributed communication method according to claim 4, wherein: 10. The response-distributed communication method according to claim 9, wherein each of said plurality of nodes sets a timer set time for response collection according to the number of said failed nodes. 11. A communication system in which a plurality of nodes mutually transmit and receive data, wherein each of the plurality of nodes holds an assigned unique value and transmits a response when transmitting the response. A communication system, wherein timing is delayed according to the unique value.