JPH0918510A - Network system data transmission method and device - Google Patents

Abstract

(57) [Summary] [Purpose] To limit the transmission of new data packets in situations where the network load is judged to be excessive, and to contribute to solving congestion and speeding up data transfer. [Structure] This is a data transmission method in which a transfer unit is sent out from a node and transferred on a network, and the estimated movement time (Tt) of a movable route of the transfer unit as seen from the node is stored,
The required arrival time (Tc) is read from the transfer unit, and the estimated arrival time (Tc) is added to the value obtained by adding the current time to the estimated travel time (Tt) of the expected travel route from the node to the node of the data receiving destination. Arrival time (Tp) for each transfer unit is calculated, and the arrival time (Tp) is compared with the preset identification time (Tps) to determine the load state of the network system, and the load state is excessive. When it is determined that the data transmission request is rejected.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data transmission method and device in a network system in which nodes communicate with each other via a transmission line to transmit data, and in particular, whether or not to accept a data delivery request. The present invention relates to a data transmission method and an apparatus for determining data.

[0002]

2. Description of the Related Art In recent years, LA has been used for various purposes and applications.
Networks such as N and WAN are established. FIG.
An example of a conventional network system is shown in FIG. The network system shown in the figure includes data delivery devices Node # 1 to #N that perform control for data communication, programmable controllers PC # 1 to PC # N that are host devices connected to the data delivery device, and data It is configured by a combination with communication paths Link # 1 to #N formed of communication cables that connect the delivery devices to each other.

In the above network system, when the programmable controllers PC # 1 to #N need to communicate with each other, information such as destination information, source information, the type of data to be transmitted, and the processing method at the destination is received. A header, which is information, is attached to the transmission data body, put together in a data packet, and a transmission request is issued to the data delivery device connected to the programmable controller.

Upon receiving the transmission request from the programmable controller, the data delivery device finds a vacant transmission line and sends the data packet to the network. The data packet is delivered to the programmable controller PC of the final destination via the data delivery device or the communication path on the route to the data delivery device connected to the programmable controller PC of the final destination.

Such a network system is used, for example, in an automated manufacturing site, in which a controller group for controlling and monitoring manufacturing machines distributed and installed in a manufacturing line communicates between devices on the site. Can be used for

In addition to the network system shown in FIG. 17, for example, a bus type ETHERNET (ISOIS8
802.3), token bus (ISOIS8802.4) LAN, FDDI (ISOIS9314) and token ring (ISOIS8) for loop type
802.5) There are various types of LAN and standard LAN.

In order to speed up the network system and effectively use the transmission line, various types of data, for example,
Next-generation high bandwidth as a transmission method that can integrate and communicate voice information that requires continuous periodic transfer, image information that requires a large amount of information, text and file data, etc.
The ATM (Asynchronous Transfer Mode) system for ISDN is drawing attention, and standardization work is currently being actively carried out.

[0008] In the ATM system, data integration is realized by performing communication with a 53-byte fixed length data packet (cell) consisting of a header and a 48-byte user information area. In this ATM system, the data delivery device is composed of a switch, and the exchange of the communication path is conducted based on the communication path information included in the header information so that the cell flowing from the input port to the switch is guided to the corresponding output port. Switch the switch. By repeating such an operation for each switch of each data delivery device, the cell is delivered to the final destination.

[0009]

In a conventional network system, when a host device tries to perform data communication, it is not always possible to obtain an expected time depending on the generation condition of a data delivery request over the entire network, which changes from moment to moment. In addition, there is a problem that the data packet is not always sent to the network system, and the sent data packet is not always delivered to the final destination of the network system within the expected time.

This is because the application processes on the network system share the resources involved in the communication of the network system while being related to each other, and
Since the individual data packets are exchanged with each other by using each other, if the delivery requests of the data packets are overlapped with each other, the network system is congested waiting for the allocation of communication resources.

Therefore, in order not to overlap the data packet delivery requests requested by the application process groups beyond the processing capability of the communication resources, resources are allocated, and the generation timing and sequence of the data packet delivery requests are excessive. An attempt can be made to limit them so that they do not overlap.

For example, in the ATM system, when there is a data communication request from the host device, the amount of data per unit time that the host device can send is determined by negotiation. And
If the host device violates this and tries to cause an excessive cell to flow into the network, discard the excessive cell as a violating cell to prevent network congestion,
As a result, the utilization efficiency of the network is improved for a large number of connected host devices.

However, as described above, a system that decides the communication capacity that can be used in advance by negotiation and suppresses traffic congestion by each host device protecting it, that is, a system that delivers data within the communication performance determined by negotiation, Since it is necessary to set the negotiation rule by using the average value or the peak value of the communication amount, efficient operation cannot be performed unless the data generation state is known to some extent. Communication between controllers used in automated manufacturing is
The generation of communication requests is random, and it is difficult to support a system such as ATM whose communication performance is determined in advance.

Further, it is necessary to predefine the generation timings and sequences of the data packet delivery requests so that they do not excessively overlap each other. In the actual application process, the data packet delivery requests are usually randomly generated in association with the operating status. To do.

Therefore, it is difficult to clearly define and control the generation timing and sequence in advance. Even if this is explicitly defined once by some means, in the actual operation stage, application process groups repeatedly join and leave the network system dynamically. Furthermore, when a new application process is added, the application process that has been running until now is reworked or stopped, and the network system is expanded or reworked, the sequence so far becomes meaningless. That is, considering the actual situation, it is impossible to define the sequence of data packets exchanged by the application process group.

The present invention has been made in consideration of such a situation, and monitors the delivery state of data packets in the operating state of the network system, and the delivery request of the data packets requested by the application process group is a communication resource. To provide a data transmission method and device for limiting acceptance of a delivery request and limiting inflow of a data packet to a network system in a data delivery device of a transmission source so as not to be overlapped with each other by exceeding a processing capability of the data delivery device. With the goal.

According to the present invention, when a new application process is added to the currently operating network system, the data packet exchanged in the current application system is replaced by the data packet exchanged by the newly added application process. Accepting a data packet delivery request issued by a newly added application process when there is a risk of failure while monitoring the congestion status of the network system to prevent abnormalities from being restricted by the delivery request It is an object of the present invention to provide a data transmission method and device that avoids a traffic jam state by rejecting the above and can maintain the employment of the current system.

[0018]

In order to achieve the above object, the present invention takes the following measures. The present invention corresponding to claim 1 or claim 9 transmits a transfer unit including data of a length as necessary from a node to a network in which a plurality of nodes are connected to each other so that data can be transmitted. In the data transmission method in which the transfer is performed by storing the transfer estimated time (Tt) of the transferable route of the transfer unit on the network viewed from the node, the transfer unit passing through the node is registered in the transfer unit. The required arrival time (Tc) at the data receiving destination is read, and the value obtained by adding the current time to the estimated travel time (Tt) of the expected travel route from the node to the data receiving node and the required arrival time (Tc). ) And arrival time (Tp) of the transfer unit is calculated, and the arrival time (Tp) is compared with the preset identification time (Tps) to determine the load state of the network system, When it is determined that the load condition of the network system is excessive, the node is refused to accept the requested data transmission request.

The present invention corresponding to claim 2 or claim 10 provides a transfer unit including data of a length as required on a network in which a plurality of nodes are connected to each other so that data can be transmitted. In the data transmission method for transmitting and transferring from the node, the estimated movement time (Tt) of the movable route of the transfer unit on the network viewed from the node is stored, and the transfer unit passing from the node to within the transfer unit is stored. The transmission source node registered in the node reads the reception time (Ta) at which the data transmission request is received, and the reception time (Ta) is set to the estimated movement time (Tt) of the transfer unit from the transmission source node to the node. The retention time (Tw) in which the transfer unit was retained in the network system was calculated from the added value and the current time, and the retention time (Tw)
And the preset identification time (Tws) are compared to determine the load status of the network system, and when it is determined that the load status of the network system is excessive, the data transmission request requested to the node is I refused to accept it.

The present invention corresponding to claim 3 or claim 11 provides a transfer unit including data of a length as necessary on a network in which a plurality of nodes are connected to each other so that data can be transmitted. In the data transmission method for transmitting and transferring from the node, the estimated movement time (Tt) of the movable route of the transfer unit on the network viewed from the node is stored, and the transfer unit passing from the node to within the transfer unit is stored. The required arrival time (Tc) at the data receiver registered in and the reception time (Ta) at which the transmission source node has received the data transmission request are read, and the required arrival time (Tc) and the reception time (Tc) of the transfer unit are read. Ta), the time (Tr) at which the transfer unit is identified, and the expected movement time (Tt) of the transfer unit to the node, the normalized transmission time (Tn) of the transfer unit to the network system is expressed by the following formula. More calculated, Tn = (Tr-Ta) / (Tc-ta-Tt) the normalized transmission time (Tn) and a preset identification time (Tn
s) is used to judge the load status of the network system, and when it is judged that the load status of the network system is excessive, the node is refused to accept the requested data transmission request. .

According to the present invention corresponding to claim 4 or claim 12, a transfer unit including data of a length as required is provided on a network in which a plurality of nodes are connected to each other so that data can be transmitted. In the data transmission method for transmitting and transferring from the node, the estimated movement time (Tt) of the movable route of the transfer unit on the network viewed from the node is stored, and the transfer unit passing from the node to within the transfer unit is stored. The required arrival time (Tc) at the data receiver registered in and the reception time (Ta) at which the transmission source node has received the data transmission request are read, and the required arrival time (Tc) and the reception time (Tc) of the transfer unit are read. Ta), the time (Tr) at which the transfer unit is identified, and the expected movement time (Tt) of the transfer unit to the node, the normalized transmission time (Tn) of the transfer unit to the network system is expressed by the following formula. Tn = (Tr−Ta) / (Tc−ta−Tt) The distribution state (Pt) of the normalized transmission time (Tn) of the transfer unit passing through the node during the measurement time (Tm) is calculated as follows: Obtain the cross-border area (Pa) where this distribution state (Pt) exceeds the boundary of the preset distribution identification pattern (Pts), and compare this cross-border area (Pa) with the preset cross-border area identification value (Pas). Then, the load state of the network system is determined, and when it is determined that the load state of the network system is excessive, the node is refused to accept the requested data transmission request.

The present invention corresponding to claim 5 or claim 13 provides a transfer unit including data of a length as necessary on a network in which a plurality of nodes are connected to each other so that data can be transmitted. In the data transmission method for transmitting and transferring from the node, the estimated movement time (Tt) of the movable route of the transfer unit on the network viewed from the node is stored, and the transfer unit passing from the node to within the transfer unit is stored. The required arrival time (Tc) at the data receiver registered in and the reception time (Ta) at which the transmission source node has received the data transmission request are read, and the required arrival time (Tc) and the reception time (Tc) of the transfer unit are read. Ta), the time (Tr) at which the transfer unit is identified, and the expected movement time (Tt) of the transfer unit to the node, the normalized transmission time (Tn) of the transfer unit to the network system is expressed by the following formula. Tn = (Tr-Ta) / (Tc-ta-Tt) The distribution state (Pt) of the normalized transmission time (Tn) of the transfer unit passing through the node during the measurement time (Tm) is calculated. , The cross-border area (Pa) where this distribution state (Pt) crosses the boundary of the preset distribution identification pattern (Pts) is obtained, and this cross-border area (Pa)
The cross-border change rate (Pad) that indicates the time change for each measurement time (Tm) is calculated, and this cross-border change rate (Pad) is compared with the preset cross-border change rate identification value (Pads) to determine the network system. When it is determined that the load state of the network system is excessive, the acceptance of the requested data transmission request to the relevant node is refused.

In the present invention corresponding to claim 6, when rejecting acceptance of the data transmission request, the transfer unit including the rejection information of the data transmission request is presented to the requester of the data transmission request.

According to the present invention of claim 7, when the acceptance of the data transmission request is rejected, there is a transfer unit that has already accepted the data transmission request at the node and is waiting for transmission to the network. For example, the transfer unit is transmitted to the network after a preset delay time (Td).

According to the present invention corresponding to claim 8, the arrival required time (Tc) is read from the transfer unit which has already received the data transmission request at the node and is waiting for transmission to the network, and responds to the data transmission request. When rejecting the acceptance by the transfer, the estimated transfer time (Tt) and delay time (T
d) is added to the current time and the mandatory arrival time (Tc) is compared, and when it is determined that the transfer unit waiting for transfer does not arrive at the data receiving destination by the mandatory arrival time (Tc), The standby transfer unit is discarded, and at least the transfer unit of the data including the discard information of the transfer unit is confirmed and presented to the requester of the transfer unit.

By the way, in this specification, a request when a response is required for various "requests" including a data transmission request is called "delivery request". The case of simply "transmitting" includes the case where no response is required.

[0027]

The present invention has the following effects by taking the above measures. In the present invention corresponding to claim 1 and claim 9, the essential arrival time (Tc) at the data receiver is read, and the essential arrival time (Tc) and the movement prediction of the movement prediction route of the transfer unit stored in advance. Time (T
From t) and the current time, a margin (Tp) of arrival at the data receiver of the transfer unit is calculated. In addition, the allowance time (T
If p) is smaller than the preset identification time (Tps), it is judged that the load on the network system is excessive and that the margin time has decreased. Refuse to accept transmission requests.

Therefore, in a situation where it is determined that the load condition of the network system is excessive, transmission of a new data packet or the like as a transfer unit is restricted, which leads to the elimination of network congestion. Further, the load on the network is limited, which contributes to speeding up of data transfer. Furthermore, if the rejected data transmission request corresponds to a new application that is added to the currently operating network system, it is avoided that the rejection causes a traffic jam condition. Therefore, there is an effect that the system, which is often experienced when adding a new application, can be prevented from falling into an unstable state and normal operation can be maintained.

In the present invention corresponding to claims 2 and 10, the reception time (Ta) at which the data delivery request for the transfer unit is accepted from the transfer unit is read at the transmission source node. Next, the transfer time stays in the network system by adding the reception time (Ta) to the pre-stored estimated transfer time (Tt) of the transfer unit to the node and subtracting it from the current time. The retention time (Tw) is calculated. When the retention time (Tw) is larger than the preset identification time (Tws), it is judged that the load condition of the network system is excessive and the retention time has increased, and the data transmission request requested to the relevant node. Refuse to

In the inventions corresponding to claims 3 and 11, the arrival required time (Tc) and the acceptance time (Ta) are read from the transfer unit. And the required arrival time (Tc)
And the reception time (Ta) and the time when the transfer unit is identified (T
r) and the stored estimated transfer time (Tt) of the transfer unit to the node, the normalized transmission time (Tn = Tr-Ta / Tc-Ta-T) of the transfer unit to the network system.
t) is calculated. Furthermore, if the normalized sending time (Tn) is compared with the preset identification time (Tns) and is large, it is determined that the normalized sending time has been extended because the network system is overloaded, and the request was made. Reject the data transmission request.

In the inventions corresponding to claims 4 and 12, the normalized sending time (Tn) is obtained in the same manner as the inventions corresponding to claims 3 and 11. And
Normalized sending time (Tm) over the set measurement time (Tm)
n) The distribution state (Pt) is calculated and the cross-border area (Pa) that exceeds the boundary between the distribution state (Pt) and the preset distribution identification pattern (Pts) is pre-determined. It is compared with the value (Pas). As a result of the comparison, if it is larger than the cross-border area identification value, it can be judged that the load is excessively larger than the assumed load state of the network system. Therefore, the acceptance of the transmission request is rejected. As a result, it becomes possible to limit the delivery request for each transfer unit in consideration of the constant network load state rather than the momentary traffic congestion state.

In the inventions corresponding to claims 5 and 13, in the same manner as above, the cross-border area (Pa) in which the distribution state (Pt) of the normalized sending time (Tn) exceeds the boundary of the distribution identification pattern (Pts) ) Is calculated, and the cross-border change rate (Pad) that indicates the temporal change of the cross-border area (Pa) is calculated. Further, the cross-border change rate (Pad) is compared with a preset cross-border change rate identification value (Pads), and when the cross-border change rate (Pad) is large, it is determined that the load fluctuation state of the network system is excessive, Refuse to accept delivery requests. As a result, it is possible to determine the progress speed of the traffic jam condition as well as the instantaneous traffic jam condition, so that it is possible to limit the delivery request in consideration of this.

In the invention according to claim 6, when acceptance of the data transmission request is rejected by the node, the data transfer unit including the rejection information is presented to the requester of the data transmission request. Therefore, the request source can take the necessary measures in response to the refusal to accept the data transmission request.

In the invention according to claim 7, when the acceptance of the data transmission request is rejected, the transfer unit which has already accepted the data transmission request and is waiting to be sent to the network system is in advance. After the set delay time (Td), the transfer unit is sent to the network system.

In the invention corresponding to claim 8, the arrival required time (T
c) is read, and the estimated movement time (Tt), delay time (Td) and current time of the movable route to the data receiving destination stored in the storage means are added to arrive at the data receiving destination of the transfer unit. Determine if it is in time for the required arrival time (Tc). When it is judged that the arrival time (Tc) is not in time, by discarding the transfer unit, meaningless transfer unit is eliminated from the network, leading to the elimination of network congestion and contributing to speeding up data transfer. .

[0036]

Embodiments of the present invention will be described below. (First Embodiment) FIG. 1 is a diagram showing the configuration of a network to which the data delivery device according to the present embodiment is applied. It has the same network configuration as the network shown in FIG. That is, the data delivery devices Node # 1 to # 5
And host devices PC # 1 to PC connected to the data delivery device
It is configured by a combination of # 7 and communication paths Link # 1 to #N formed of communication cables connecting the data delivery devices to each other.

The data delivery device Node # 3 is connected to its own node via the transmission line, and the input port control unit 11 for inputting data from another data delivery device is connected to the own node via the transmission line. An output port control unit 12 that sends data to another data delivery device, a transfer unit 13 that transfers the data entered in the input port control unit 11 to a specific output port of the output port control unit 12, and an output port at the transfer unit 13. The storage unit 14 stores the information necessary for determining. The data delivery device Node # 3 further determines a delivery request acceptance determination unit 15 that determines whether to reject the acceptance of the data delivery request from the host, and necessary information for determining whether to reject the acceptance of the data delivery request. And a receiving unit 1 that receives a data delivery request from the host device, notifies the determination result, and inserts the received data packet of the data delivery request into the output port control unit 12.
7 is provided. The data delivery devices other than the data delivery device Node # 3 have the same configuration.

FIG. 2 shows specific functional blocks of the data delivery device Node # 3. The signal reception input circuit 21 is connected to a communication path from another data delivery device. Although not shown, a plurality of signal reception input circuits 21 are provided. The data packet from another data delivery device output from the signal reception input circuit 21 is input to the input buffer element 22.
To accept and buffer. Input buffer element 2
The output from 2 is put into the input queue element 23 having a queuing-capable storage function, and the data packet output from the input queue element 23 is supplied to the transfer element 25 via the input port element 24. The input port element 24 performs interface processing with the transfer element 25. The signal transmission output circuit 26 is connected to a communication path to another data delivery device. Although not shown, a plurality of signal transmission output circuits 2
6 are provided. In order to send the data packet to be transferred by the data delivery device to the signal output circuit 26 connected to the next data delivery device on the transfer destination path, the output port element 27 temporarily buffers the data packet.

The storage element 28 forms a memory area of the storage unit 14. The storage element 28 holds output port information connected to a communication path on the route from the data delivery device to the target data delivery device to the next data delivery device. FIG. 4 shows specific storage contents of the storage element 28.

The identification element 29 extracts information necessary for network monitoring and information necessary for data packet transfer from a specific area of the data packet received by the data delivery device. Specifically, "destination information" that specifies the final destination included in the received data packet, "completion time information" that specifies the arrival required time (Tc) at the final destination, and the transmission of the packet. In addition to the “source information” indicating the source, other information such as priority processing and information designating the execution of dynamic route selection in the transfer processing of the data packet is extracted.

Of the information extracted from the data packet by the identification element 29, information necessary for transfer, that is, "destination information" and "source information" is input to the determination element 31. The determination element 31 outputs the output port number corresponding to the route to the data delivery device to be transferred next to the insertion element 32 based on the output port information read from the storage element 28.

The inserting element 32 receives the output port number from the judging element 31, and embeds the output port number in the specific area of the data packet. Transfer unit 1 described above
The forwarding element 25 realizing the switch function of No. 3 delivers the data packet to the next data delivery device on the communication path. Transfer control of the data packet to the output port element 27 corresponding to the data delivery device.

The storage element 33 forms the memory area of the storage unit 16 described above. The storage element 33 holds delivery time information and the like regarding expected movement time (Tt) required for delivery on the route for delivering the data packet from the data delivery apparatus to the target data delivery apparatus. FIG. 3 shows the storage element 33.
3 shows the specific storage contents of. The data delivery device
Stores information such as route information from Node # 3 to other data delivery devices, estimated travel time (Tt) when each route is used, and identification time (Tps) for determining network load status. .

The "destination information" of the data packet extracted from the data packet by the identification element 29 is input to the other identification element 34, and the "completion time information" and "source information" extracted from the data packet are accepted. Input to element 35. The identification element 34 searches for delivery time information composed of estimated travel time (Tt) corresponding to a plurality of routes to the destination, which is held in the storage element 33 based on the “destination information”, and extracts the delivery time information to determine the acceptance determination element. Input to 35. The reception determination element 35 includes “delivery time information” corresponding to a plurality of routes from the identification element 34, “completion time information” and “sender information” extracted in the identification guideline 29, and “current time” from the clock guideline 36. Information "is entered. This acceptance determination element 35 is
From this input information, calculate the allowance time (Tp) for the data packet to arrive at the data receiver and compare the allowance time (Tp) with the identification time (Tps) of the corresponding route to determine whether or not the delivery request can be accepted. To judge.

It should be noted that the clock element 36 is synchronized among all the data delivery devices within an allowable error range.
The host device PC # 5, which directly communicates with the data delivery device Node # 3, is connected via the host interface circuit 37. The reception element 38 receives a data delivery request issued from the host device PC # 5 to the data delivery apparatus Node # 3.

The reception element 38 judges the data delivery request coming through the host interface circuit 37 and transfers it to the reception judgment element 35, and based on the judgment output of the reception judgment element 35 for the data transmission request, the host device PC # 5: Returns the acceptance permission or rejection information of the data delivery request in units of transfer. When the acceptance is permitted, a process of writing the time information from the clock element 36 in the specific area of the data packet is performed.

The data packet corresponding to the data delivery request accepted based on the determination result from the acceptance determination element 35 is stored in the transmission buffer queue 39. The output port insertion element 40 outputs the data packet at the head of the transmission buffer queue 39 to the corresponding output port 27 based on the destination information of the data packet. The reception unit 17 described above includes a host interface circuit 37, a reception element 38, and a transmission buffer queue 39.

Schedule elements 41 and 42 for controlling the timing relating to the data transfer processing between the above-mentioned respective elements in the data delivery apparatus and other non-essential elements attached thereto and controlling the generation of necessary timing signals are provided. I have it.

Next, the operation of this embodiment configured as described above will be described with reference to FIGS. Here, the network monitoring in the data delivery device Node # 3 and the acceptance refusal determination when there is a data delivery request from the host device PC # 5 to the data delivery device Node # 3 will be specifically described.

The data delivery device Node # 3 constantly monitors data packets passing through the data delivery device in order to grasp the load state of the network. Now, the data delivery device is transmitted from the host device PC # 4 connected to the data delivery device Node # 4, and the data delivery device Node # 1
It is assumed that a data packet whose final destination is the host device PC # 1 connected to is being viewed.

At the transmission source of the data packet, a mandatory arrival time (Tc) indicating by what time the host device PC # 1 has to arrive in a specific area of the data packet is registered. FIG. 6 shows a schematic data structure of a data packet received by the data delivery device. The mandatory arrival time (Tc) is registered in the specific area indicated by the diagonal lines of the data packet.

The identification element 29 extracts "destination information", "completion time information" specifying the required arrival time (Tc), "sender information", and other necessary information from the data packet. When the "destination information" is input, the identification element 34 searches for the estimated movement time (Tt) corresponding to a plurality of routes from the storage element 33 to the destination (PC # 1). There are two routes, (2,1) and (4,1), but the source is No.
Since it is de # 4, the estimated movement time (Tt = 20) and the corresponding identification time (Tps = 5) are read from the storage element 33 and input to the acceptance determination element 35 only for the route (2,1).

The acceptance judgment element 35 subtracts the estimated movement time (Tt) and the current time (Tr) of the expected movement route of the data packet from the required arrival time (Tc) as shown in FIG. Calculate the spare time (Tp) for arrival at the destination. The calculated margin time (Tp) is compared with the identification time (Tps).

Here, when the network load becomes large, it takes a long time to reach the data delivery device Node # 3 from the data delivery device Node # 4 which is the transmission source.
Since the period from the reception time (Ta) to the current time (Tr) shown in (3) becomes long, it is expected that the margin time (Tp) will become short. The time margin (Tp) is a parameter that expresses the load state of the network. Therefore, the load state of the network can be determined by comparing the identification time (Tps) and the allowance time (Tp) that are appropriately determined according to the expected movement route.

The margin time (Tp) calculated is the identification time.
If it is shorter than (Tps), it is judged that the load condition of the network system is excessive, and the margin time (Tp) is the identification time (Tp).
If it is longer than ps), it is judged that the network system is not overloaded.

The acceptance judgment element 35 constantly monitors the load state of the network by the above-mentioned arithmetic processing. The reception element 38 of the data delivery device Node # 3 is the host device PC #.
When the data delivery request is received from 5, the acceptance determination element 35
Notify that there is a data delivery request from the host device PC # 5. If the acceptance determination element 35 determines that the load state of the network system is excessive when the data delivery request is received, the acceptance determination element 35 returns an acceptance rejection determination result that rejects the acceptance of the data delivery request to the acceptance element 38. The acceptance element 38 notifies the host device PC # 5 of the determination result (rejection of acceptance) via the host interface circuit 37.

On the other hand, if the acceptance determination element 35 determines that the load state of the network system is the non-overloaded state when the data delivery request is received, the data delivery request is accepted. When the data delivery request is accepted,
The data packet transmitted from the host device PC # 5 is registered in the transmission buffer queue 39.

As described above, according to this embodiment, the margin time (Tp) for the arrival of the data packet at the data receiver is calculated, and the margin time (Tp) is compared with the identification time (Tps) of the corresponding route. Then, when it is determined that the load state of the network system is excessive, the acceptance of the data delivery request is rejected, so that the transmission of new data packets is restricted and the network congestion can be eliminated. Also,
By limiting the load on the network, it is possible to contribute to speeding up of data transfer.

Further, assuming that the data delivery request whose acceptance is rejected corresponds to a new application to be added to the network system which is normally operating now, the data delivery request from the new application process is rejected. By doing so, it is possible to avoid in advance causing a traffic jam. Therefore, it is possible to prevent the system, which is often experienced when adding a new application, from falling into an unstable state, and it is possible to achieve an excellent effect of maintaining normal operation.

(Second Embodiment) Next, a second embodiment will be described. The basic configuration of this embodiment is the same as that of the first embodiment described above.

In this embodiment, the time (Ta) at which a data delivery request for a data packet sent to the network is accepted by the data delivery device of the transmission source.
Is registered. FIG. 8 shows an example of the data structure of a data packet. In the figure, the reception time (Ta) is registered in the specific area indicated by the diagonal lines.

The storage element 33 stores in advance the information shown in FIG. 3 as information for judging the load state of the network, but in the present embodiment, the identification time (Tps) for comparison with the margin time (Tp). ), The identification time (Tws) to be compared with the retention time (Tw) described later is registered. The load determination state of the network is determined by comparing the acceptance determination element 35, the retention time (Tw) and the identification time (Tws).

For example, if the data delivery device Node # 3 is No
It is assumed that the data packet shown in FIG. 8 transmitted from the host device PC # 4 of de # 4 and having the host device PC # 1 of Node # 1 as the final destination is received.

The identification element 29 identifies "destination information", reception time (Ta), "source information", and other necessary information from the data packet. The identification element 34 stores the storage element 3 based on the “source information” identified by the identification element 29.
3 to the transmission source (Node # 4) to the data delivery device (No.
Estimated travel time (Tt) for multiple routes up to de # 3)
Search for. There are two routes, (4) and (2,1,4), but since the destination is Node # 1, only the route (4) has an estimated travel time (Tt = 10) and the corresponding identification time (Tws). ) Is read from the storage element 33 and the acceptance determination element 35
Enter

In the acceptance judgment element 35, the time obtained by adding the expected travel time (Tt) to the acceptance time (Ta) as shown in FIG.
The residence time (Tw) is calculated by subtracting from the current time (Tr). As a result, the time (Tw) that the data packet stays in the network system is calculated.

Here, since the staying time (Tw) represents the time that the data packet stays in the network system as described above, if the staying time is longer than necessary, the network is congested. Means that. Therefore, register the appropriate identification time (Tws) for each route, and compare the calculated retention time (Tw) with the corresponding route identification time (Tws) to determine the load status of the network. Can be judged.

When the acceptance determination element 35 is notified by the acceptance element 38 that there is a data delivery request from the host device PC # 5, the load state of the network system determined from the information of the data packet being received at that time. According to the above, it is determined whether or not to reject the data delivery request. If the retention time (Tw) is larger than the identification time (Tws), it is judged that the network load is excessive and the acceptance is rejected. If the retention time (Tw) is smaller than the identification time (Tws), the network load is reduced. Accepts that there is a non-overload condition. The subsequent processing is the same as that of the first embodiment described above.

As described above, according to this embodiment, the load state of the network is judged by determining the retention time (Tw) of the received data packet on the network and comparing it with the identification time (Tws). Since the data delivery request is rejected when the load is excessive, the same effect as that of the first embodiment described above can be obtained.

(Third Embodiment) Next, a third embodiment will be described. The basic configuration of this embodiment is the same as that of the first embodiment described above.

In this embodiment, the time (T (T) at which the data delivery request of the data packet is sent by the data delivery device of the transmission source is accepted in the data packet sent to the network.
The required arrival time (Tc) at a) and the final destination is registered. FIG. 10 shows an example of the data structure of a data packet. In the figure, the reception time (T
a) and mandatory arrival time (Tc) are registered respectively.

The storage element 33 stores in advance the information shown in FIG. 3 as information for judging the load state of the network, but in the present embodiment, the identification time (Tps) for comparison with the margin time (Tp). ) Instead of normalized transmission time (Tn)
The identification time (Tns) for comparison with is registered. Reception determination element 35, normalized transmission time (Tn) and identification time (Tn)
The load state of the network is determined by comparing with s).

For example, the data delivery device Node # 3
Is sent from the host device PC # 4 of Node # 4, Node # 1
It is assumed that the data packet shown in FIG. 10 whose final destination is the host device PC # 1 of FIG.

The identification element 29 determines from the data packet “destination information”, acceptance time (Ta), arrival required time (Tc),
Identify "source information" and other necessary information. The identification element 34 is an estimated movement time corresponding to a plurality of routes from the transmission source (Node # 4) to the data delivery device (Node # 3) from the storage element 33 based on the “transmission source information” identified by the identification element 29. Search for (Tt). There are two routes, (4) and (2,1,4), but the destination is Node # 1, so the estimated travel time (Tt = 1) for route (4) only.
0) and the corresponding identification time (Tns) are read from the storage element 33 and input to the acceptance determination element 35.

As shown in FIG. 9, the acceptance judgment element 35 includes the arrival required time (Tc), the acceptance time (Ta), the time (Tr) at which the data packet was identified, and the data packet up to the data delivery device. From the estimated movement time (Tt), the normalized transmission time (Tn) of the data packet to the network system is calculated based on the following equation.

Tn = (Tr −T) a / (Tc 1 Ta −Tt) Normalized sending time (Tn) and preset identification time (Tns)
And is compared to determine the load state of the network system. As shown in FIG. 11, the larger the normalized transmission time (Tn), the closer the identification time (Tr) is to the arrival required time (Tc), indicating that the network is congested. Therefore, if the normalized transmission time (Tn) is larger than the identification time (Tns), it is determined that the normalized transmission time has increased because the network load is excessive, and the requested data delivery request is rejected.

According to the present embodiment as described above, the network load state is determined by calculating the normalized transmission time Tn indicating the network load state and comparing it with the identification time (Tns). When it is determined that there is a request to reject data delivery from the host device,
The same effects as those of the above-described first embodiment can be obtained.

(Fourth Embodiment) Next, a fourth embodiment will be described. The basic configuration of this embodiment is similar to that of the above-mentioned third embodiment. That is, as shown in FIG. 10, the data packet transmitted to the network has the reception time (Ta) of the data delivery request of the data packet and the arrival required time (Tc) at the final destination data delivery device registered. ing. Further, the information shown in FIG. 3 is stored in the storage element 33 in which information for monitoring the network is stored, and the identification time (Tns) is registered instead of the identification time (Tps).

The acceptance judgment element 35 has a function of calculating the normalized transmission time Tn as in the third embodiment. Furthermore, the distribution state of the normalized transmission time (Tn) over a certain measurement time (Tm)
(Pt) is calculated, and a cross-boundary area (Pa) that exceeds the boundary of the distribution identification pattern (Pts) in the distribution (Pt) of the normalized transmission time (Tn) is provided. The acceptance judgment element 35 is a cross-border area (P
The load condition of the network is judged by comparing a) with the pre-determined cross-border area identification value (Pas).

FIG. 12 is a diagram showing the concept of this embodiment. When the network is not congested and the load is standard, the distribution status of the normalized transmission time (Tn) based on multiple data packets checked during an arbitrary measurement time (Tm) is the distribution identification pattern. (Pts). The limit point of the normalized sending time that is not included in the distribution identification pattern (Pts) is set as the boundary value.

On the other hand, if the distribution state actually created from the normalized transmission time (Tn) based on a plurality of data packets over a certain period (Tm) is Pt, as shown by the hatched lines in FIG. Right side (normalized transmission time (T
n) becomes larger), the cross-border area (Pt) of the distribution state (Pt)
The larger a) is, the more data packets the normalized transmission time (Tn) is. Normalized transmission time (T
Since n) is a parameter indicating the load state of the network as described in detail in the third embodiment, by comparing the cross-border area (Pa) with a predetermined cross-border area identification value (Pas), the instantaneous It is possible to judge the load condition of the network which is more constant than the traffic jam condition.

In this embodiment, as shown in FIG. 13, the distribution identification pattern (P
The acceptance determining element 35 is provided with a function that defines a boundary line as shown in FIG. In the acceptance determination element 35, the normalized transmission time (Tn) that changes from moment to moment with the period from the current time point to the predetermined time period as the measurement period (Tm)
The distribution state (Pt) is calculated, and the cross-border area (Pa) is calculated using a function that defines the boundary line stored in advance. FIG.
Shows the case where the distribution (Pt) of the normalized transmission time (Tn) is inside the boundary line and the cross-border area (Pa) does not occur.
FIG. 14 shows a case where a part of the distribution (Pt) of the normalized transmission time (Tn) is outside the boundary line and the cross-border area (Pa) occurs.

As a result of comparing the cross-border area (Pa) with a predetermined cross-border area identification value (Pas), the acceptance judgment element 35 determines that the cross-border area (Pa) is larger than the cross-border area identification value (Pas). It is determined that the load on the assumed network system is excessive. When the reception element 38 informs that there is a data delivery request while obtaining such a determination result, the acceptance of the data delivery request is rejected.

According to the present embodiment as described above, the distribution state (Pt) over the set measurement time (Tm) of the normalized transmission special time mark (Tn) is calculated, and the distribution state (Pt) is preset. Since the cross-border area (Pa) that exceeds the boundary of the distribution identification pattern (Pts) of the distributed state is compared with the pre-determined cross-border area identification value (Pas), the load condition of the network is judged. It is possible to obtain the same effect as that of the third embodiment,
Further, it becomes possible to limit the delivery request of the data packet in consideration of the constant network load condition rather than the momentary traffic congestion condition.

In the fourth embodiment, the cross-border change rate (Pad) showing the temporal change of the cross-border area (Pa) is calculated, and the cross-border change rate (Pad) is set to the preset cross-border change rate identification value. When the cross-border change rate (Pad) is larger than that of (Pads), it may be determined that the load fluctuation state of the network system is excessive, and the acceptance of the data delivery request may be rejected.

By monitoring the load condition of the network in this way, it is possible to determine the progress speed of the congestion condition as well as the instantaneous congestion condition, so that it is possible to limit the delivery request in consideration of this.

In each of the above embodiments, when the data delivery request is rejected, the data transfer unit including the rejection information is presented to the requester of the data delivery. As a result, the request source can take the necessary measures in response to the refusal to accept the data delivery request.

(Fifth Embodiment) Next, a fifth embodiment will be described. In the present embodiment, the following functions are added to the above-described first to fourth embodiments. In FIG. 15, the data delivery device Node # 3 determines whether or not to accept the data delivery request from the host device PC # 5, and the data packet for which the data delivery request has already been accepted is sent to the network system by the transmission buffer queue 39. Is waiting for.

In this embodiment, when the acceptance judgment element 35 rejects the acceptance of the data delivery request, the data delivery request has already been accepted and the transmission buffer queue 39 of the data delivery apparatus waits for transmission to the network system. If there is a data packet that has been transmitted, even if such a data packet is not immediately sent from the output port 27 to the network system, the delay time (T
The data packet is sent to the network system after d) has elapsed.

According to the present embodiment as described above, when the acceptance determining element 35 determines that the network load is excessive, not only the data delivery request is rejected, but also the data delivery request is already accepted and the transmission waits. Even if the data packet is being transmitted, the data packet is not transmitted to the network until a predetermined time has elapsed, so that the data packet can be prevented from being transmitted to the congested network.

(Sixth Embodiment) Next, a sixth embodiment will be described. In the present embodiment, the following functions are added to the above-described first to fifth embodiments. In the data delivery device Node # 3 shown in FIG. 2, a host transmission identification element 50 shown by a dotted line is arranged between the transmission queue buffer 39 and the output port insertion element 40.

FIG. 16 is a flowchart showing the operation contents of the host transmission identification element 50. When the acceptance determination element 35 rejects acceptance of the data delivery request, it determines whether or not a data packet is waiting in the transmission buffer queue 39. If there is a waiting data packet, the required arrival time (Tc) and the destination information are extracted from the specific area of the data packet, the expected movement time from the storage element 31 to the destination is searched, and further the clock element 36 is used. Read the current time of. Then, the time required for the data packet to reach the destination is calculated, and comparison with the mandatory arrival time (Tc) is performed. As a result of the comparison, if the time required to reach the destination is longer than the mandatory arrival time (Tc), it is clear that the packet does not arrive before the mandatory arrival time (Tc) even if it is sent as it is. The data packet is discarded without being sent to the. Then, the discard information indicating that the data packet has been discarded without being sent is notified to the requesting host device PC # 5.

If the time required to reach the destination is smaller than the required arrival time (Tc), it is sent to the output port insertion element 67 after the delay time elapses and sent to the network.

As described above, according to the present embodiment, the arrival time (T
c) is read, and the arrival time at which the arrival of the data packet at the data receiving destination is the mandatory arrival time (Tc) based on the estimated travel time (Tt), delay time (Td), and current time of the movable route to the data receiving destination. When it is determined that the question is not answered, by discarding the data packet, meaningless data packets are excluded from the network, the congestion of the network can be eliminated, and the data transfer speed can be increased.

Since an ATM host or the like can queue data packets in the host device, the sixth embodiment described above may be applied to the host device. Further, the type and structure of the data packet to be delivered are not particularly limited, and it can be widely and generally applied. Further, the network configuration of this system has been described by way of example with respect to a network system having a star type configuration, but also for an ATM network having a star type configuration, as will be understood from the example, the transfer element is an ATM switch. By replacing it with the configuration, it is possible to easily add the function of this system to the ATM data delivery device. Furthermore, the present invention is not limited to a star-type network system in particular, and is broadly and generally applicable to those in which data delivery devices are linearly arranged and those in which ring-shaped data delivery devices are arranged. Easy to apply. The present invention is not limited to the above embodiments, and various modifications can be made without departing from the gist of the present invention.

[0095]

According to the present invention, in a situation where it is determined that the load state of the network system is excessive, transmission of new data packets is restricted, which leads to elimination of network congestion. Further, the load on the network is limited, which contributes to speeding up of data transfer. Furthermore, if the data delivery request whose acceptance is rejected corresponds to a new application to be added to the network system that is currently operating normally, it is avoided that the rejection causes a traffic jam condition. Therefore, there is also an effect that the system, which is often experienced when adding a new application, can be prevented from falling into an unstable state and normal operation can be maintained.

Further, as a determination method, since it is possible to determine the instantaneous network congestion state as well as the constant network load state and the progress rate of the network congestion state, it is possible to limit the delivery request in consideration of this. In addition, meaningless data packets can be discarded, and meaningless data packets are eliminated from the network, which leads to elimination of network congestion and contributes to speeding up data transfer.

[Brief description of the drawings]

FIG. 1 is a network configuration diagram of a network system according to a first embodiment.

FIG. 2 is a configuration diagram of a data delivery device provided in the first embodiment.

FIG. 3 is a diagram showing stored contents of a storage element in the data delivery device shown in FIG.

FIG. 4 is a diagram showing stored contents of another storage element in the data delivery device shown in FIG. 2.

FIG. 5 is a diagram showing a data packet transfer path in the first embodiment.

FIG. 6 is a configuration example of a data packet transferred on the network in the first embodiment.

FIG. 7 is a conceptual diagram for determining the load state of the network in the first embodiment.

FIG. 8 is a configuration example of a data packet transferred on the network in the second embodiment.

FIG. 9 is a conceptual diagram for determining the load state of the network in the second embodiment.

FIG. 10 is a configuration example of a data packet transferred on the network in the third embodiment.

FIG. 11 is a conceptual diagram for determining the load state of the network in the third embodiment.

FIG. 12 is a conceptual diagram for determining the load state of the network in the fourth embodiment.

FIG. 13 is a diagram for determining the load state of the network in the fourth embodiment and is a diagram showing the relationship between the distribution of the non-overload state and the boundary line.

FIG. 14 is a diagram for determining the load state of the network in the fourth embodiment, and is a diagram showing the relationship between the distribution of the overload state and the boundary line.

FIG. 15 is a conceptual diagram of the fifth embodiment.

FIG. 16 is a flowchart showing the operation contents of the sixth embodiment.

FIG. 17 is a diagram showing a star-type network configuration.

[Explanation of symbols]

Node # 1 to Node # 5 ... Data delivery device, PC # 1 to PC # 7
... host device, Link # 1 to #N ... communication path, 11 ... input port control unit, 12 ... output port control unit, 13 ... transfer unit,
14, 16 ... Storage unit, 15 ... Delivery request acceptance judgment unit, 17
... reception unit, 29, 34 ... identification element, 35 ... reception determination element, 39 ... transmission queue buffer, 50 ... host transmission identification element.

Claims (13)

[Claims] 1. A data transmission method in which a transfer unit including data of a length as required is transmitted from the node and transferred onto a network in which a plurality of nodes are connected to each other so that data can be transmitted. The expected movement time (Tt) of the transferable route of the transfer unit on the network viewed from the node is stored, and arrival at the data receiver registered in the transfer unit from the transfer unit passing through the node is mandatory. The time (Tc) is read, and the arrival allowance of the transfer unit is calculated from the value obtained by adding the current time to the estimated travel time (Tt) of the expected travel route from the node to the data receiving node and the required arrival time (Tc). Calculate the time (Tp) and compare the arrival time (Tp) with the preset identification time (Tps) to determine the load status of the network system and determine the load status of the network system. Data transmission method of a network system is characterized in that so as to refuse to accept the requested data transmission request to the node when it is determined that excessive. 2. A data transmission method in which a transfer unit including data of a length as necessary is transmitted from the node and transferred onto a network in which a plurality of nodes are connected to each other so that data can be transferred. The estimated transfer time (Tt) of the movable route of the transfer unit on the network viewed from the node is stored, and the transmission source node registered in the transfer unit from the transfer unit passing through the node transmits data. The acceptance time (Ta) at which the request is accepted is read, and the transfer unit is determined from the value obtained by adding the acceptance time (Ta) to the expected transfer time (Tt) of the transfer unit from the transmission source node to the node and the current time. Residence time (T
w) is calculated, the retention time (Tw) is compared with the preset identification time (Tws) to determine the load state of the network system, and when it is determined that the load state of the network system is excessive A data transmission method for a network system, wherein refusal to accept a requested data transmission request to the node. 3. A data transmission method in which a transfer unit including data of a length as required is transmitted from the node and transferred onto a network in which a plurality of nodes are connected to each other so that data can be transmitted. The expected movement time (Tt) of the transferable route of the transfer unit on the network viewed from the node is stored, and arrival at the data receiver registered in the transfer unit from the transfer unit passing through the node is mandatory. Time (Tc) and reception time (Ta) when the transmission source node accepted the data transmission request
From the required transfer time (Tc) of the transfer unit, the reception time (Ta), the time (Tr) at which the transfer unit was identified, and the estimated transfer time (Tt) of the transfer unit to the node. The normalized transmission time (Tn) to the network system is calculated by the following formula: Tn = (Tr-Ta) / (Tc-ta-Tt) The normalized transmission time (Tn) and the preset identification time (Tn)
s) is used to judge the load status of the network system, and when it is judged that the load status of the network system is excessive, the node is refused to accept the requested data transmission request. A data transmission method for a network system, characterized in that 4. A data transmission method in which a transfer unit including data of a length as necessary is transmitted from the node and transferred onto a network in which a plurality of nodes are connected to each other so that data can be transferred. The expected movement time (Tt) of the transferable route of the transfer unit on the network viewed from the node is stored, and arrival at the data receiver registered in the transfer unit from the transfer unit passing through the node is mandatory. Time (Tc) and reception time (Ta) when the transmission source node accepted the data transmission request
From the required transfer time (Tc) of the transfer unit, the reception time (Ta), the time (Tr) at which the transfer unit was identified, and the estimated transfer time (Tt) of the transfer unit to the node. The normalized transmission time (Tn) to the network system is calculated by the following formula: Tn = (Tr-Ta) / (Tc-ta-Tt) The normal transfer unit passing through the node during the measurement time (Tm). The distribution state (Pt) of the encrypted transmission time (Tn) is obtained, the cross-border area (Pa) over which the distribution state (Pt) exceeds the boundary of the preset distribution identification pattern (Pts) is obtained, and this cross-border area (Pa) And the preset cross-border area identification value (Pas) are compared to determine the load status of the network system. When it is determined that the load status of the network system is excessive, the data transmission requested for the node Refuse to accept request A data transmission method for a network system, characterized in that 5. A data transmission method in which a transfer unit including data of a length as necessary is transmitted from the node and transferred onto a network in which a plurality of nodes are connected to each other so that data can be transferred. The expected movement time (Tt) of the transferable route of the transfer unit on the network viewed from the node is stored, and arrival at the data receiver registered in the transfer unit from the transfer unit passing through the node is mandatory. Time (Tc) and reception time (Ta) when the transmission source node accepted the data transmission request
From the required transfer time (Tc) of the transfer unit, the reception time (Ta), the time (Tr) at which the transfer unit was identified, and the estimated transfer time (Tt) of the transfer unit to the node. The normalized transmission time (Tn) to the network system of Tn = (Tr-Ta) / (Tc-ta-Tt) of the transfer unit passing through the node during the measurement time (Tm) The distribution state (Pt) of the normalized transmission time (Tn) is obtained, and the cross-border area (Pa) where this distribution state (Pt) exceeds the boundary of the preset distribution identification pattern (Pts) is obtained. )
The cross-border change rate (Pad) that indicates the time change for each measurement time (Tm) is calculated, and this cross-border change rate (Pad) is compared with the preset cross-border change rate identification value (Pads) to determine the network system. The network system data is characterized in that when the load state of the network system is judged and the load state of the network system is judged to be excessive, the node is refused to accept the requested data transmission request. Transmission method. 6. When rejecting acceptance of the data transmission request, a transfer unit including refusal information of the data transmission request is presented to a request source of the data transmission request. A method for transmitting data in a network system according to any one of 1. 7. When the data transmission request is rejected, if there is a transfer unit that has already received the data transmission request and is waiting for transmission to the network, the transfer unit is set in advance. The data transmission method for a network system according to any one of claims 1 to 6, wherein the data is transmitted to the network after the specified delay time (Td). 8. The required arrival time (Tc) is read from a transfer unit that has already received a data transmission request at the node and is waiting for transmission to the network, and rejects the data transmission request. In this case, for the transfer unit waiting for transfer, a value obtained by adding the estimated travel time (Tt), the delay time (Td) and the current time of the transferable route to the data receiver and the mandatory arrival time (Tc) are compared. , When it is determined that the transfer unit waiting for transfer does not arrive at the data receiving destination by the arrival required time (Tc), the transfer unit waiting at the transfer is discarded, and the data including at least the discard information of the transfer unit is transferred. 8. The data transmission method of a network system according to claim 1, wherein the unit is confirmed and presented to a request source of the transfer unit. 9. In a data transmission device of a network system for transferring a transfer unit including data of a length as necessary, a predicted movement time (Tt) corresponding to a movable route of the transfer unit on the network system. ) Is stored, an identification unit that reads the destination information registered in the transfer unit and the arrival required time (Tc) at the data receiving destination from the transfer unit, and a destination that the identification unit reads from the transfer unit. Means for retrieving the estimated movement time (Tt) of the estimated movement route of the transfer unit from the storage means based on information, and a value obtained by adding the current time to the retrieved estimated movement time (Tt) of the transfer unit and the arrival Means for calculating a transfer unit margin time (Tp) from the required time (Tc), and the margin time (Tp) and a preset identification time (Tps)
A network load monitoring means for determining the load status of the network system by comparing with the above, and an acceptance rejecting means for rejecting acceptance of the data transmission request when the network load monitoring means determines that the load status of the network system is excessive. A data transmission device for a network system, comprising: 10. A data transfer device of a network system for transferring a transfer unit including data of a length as required, and a predicted movement time (Tt) corresponding to a movable route of the transfer unit on the network system. ) Is stored, an identification unit that reads the transmission source information registered in the transfer unit and the reception time (Ta) at which the transmission source node received the data transmission request from the transfer unit, and the identification unit, A means for retrieving the estimated movement time (Tt) from the transmission source node from the storage means based on the read transmission source information, and a value obtained by adding the acceptance time (Ta) to the retrieved movement estimated time (Tt) and the current A means for calculating the staying time (Tw) in which the transfer unit stays in the network system from the time, and the staying time (Tw) and a preset identification time (Tws) Network load monitoring means for comparing the load status of the network system with each other, and an acceptance rejecting means for rejecting acceptance of the data transmission request when the network load monitoring means determines that the load status of the network system is excessive. A data transmission device for a network system, comprising: 11. In a data transmission device of a network system for transferring a transfer unit including data of a length as required, a predicted movement time (Tt) corresponding to a movable route of the transfer unit on the network system. ) Is stored, the transmission source information registered in the transfer unit from the transfer unit, the reception time (Ta) at which the transmission source node receives the data transmission request, and the arrival required time at the data reception destination ( Tc) identifying means, means for retrieving the estimated movement time (Tt) from the source node from the storage means based on the source information read by the identifying means, and a mandatory arrival time (Tc) for the transfer unit. ), The reception time (Ta), the time (Tr) at which the transfer unit was identified, and the expected movement time (Tt) of the transfer unit to the node, and the transfer unit's correctness to the network system. Normalization means for calculating the normalized transmission time (Tn) by the following formula: Tn = (Tr-Ta) / (Tc-ta-Tt) The normalized transmission time (Tn) calculated by the normalization means is preset. Network load monitoring means for comparing the identification time (Tns) to determine the load status of the network system, and accepting a data transmission request when the network load monitoring means judges that the load status of the network system is excessive. A data transmission device of a network system, comprising: an acceptance refusal unit that rejects. 12. The network load monitoring means detects a distribution state (Pt) of normalized transmission times (Tn) of transfer units over a set measurement time (Tm), and the detected distribution state (Pt) is The network is obtained by calculating the cross-border area (Pa) that exceeds the boundary of the distribution identification pattern (Pts) of the preset distribution state and comparing the cross-border area (Pa) with the preset cross-border area identification value (Pas). The data transmission device of a network system according to claim 11, wherein the load condition of the system is judged. 13. The network load monitoring means calculates a cross-border change rate (Pad) indicating a temporal change of the cross-border area (Pa) for each measurement time (Tm), and the cross-border change rate (Pad).
And a preset cross-border change rate identification value (Pads) are compared to determine the load fluctuation state of the network system, and the reception refusal means is the requested data when the load fluctuation of the network system is excessive. 13. The data transmission device for a network system according to claim 12, wherein acceptance of the transmission request is rejected.