JPS59139743A - Information transfer system - Google Patents

Abstract

PURPOSE:To transfer assuredly plural packets by transmitting the data on the time to be occupied at the own node to a bus line and meanwhile discontinuing the requests for transmission at other nodes. CONSTITUTION:When the continuous transmission of plural packets is desired at a node C, a JAM signal is transmitted after detecting an idle state of a bus line BL to inform the monopolization of the line BL to other nodes. Then the monopolized time T of the line BL is put on the line BL in the form of the data TD. Then the node C can use exclusively the line BL as long as no collision occurs. While other nodes receive the data TD and sets a timer at time T to hold a transmission request until the timer is turned off. As a result, the node C can advance consecutively the packets for the time T.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an information transmission system that enables continuous transmission of multiple packets.

Conventionally, TeC3 was used as an access method for controlling transmission rights in the form of a bus.
MA/CD (Carrier-3enseMult
iple Access /Co11ision D
) is known. That is, this access method (i) determines the presence/absence of transmission data from other nodes by detecting carriers;

(ii) Start transmission with carrier OFF. If a collision is detected before the transmission is complete, the transmission is stopped, waits for an appropriate amount of time, and then retransmitted.

This method allows all nodes to hang on the bus equally and acquire the transmission right for one packet under the same conditions, but on the other hand, if you want to send multiple packets continuously, that is, if you want to monopolize the bus for a certain period of time. It is not possible to respond to requests for freely transferring multiple packets.

This invention has been made to solve the above problems. The invention will be explained below with reference to the drawings.

FIG. 1 shows a block diagram with nodes hanging from a bus. In this figure, BL is a pass line, T1 and T2 are terminators, A, B, and C are mates, and as shown in FIG. 4, it has a ROM for storing control procedures. Further, M indicates a master node that monitors the network.

FIG. 2 shows a timing chart during data transmission (normal), and FIG. 3 shows a timing chart during data transmission (bus exclusive use).

FIG. 4 is a general flowchart at the time of transmission in those nodes.

In Figure 2, during normal data transmission (normal CSMA
/CD method), when a transmission request Q occurs to node B while node A is transmitting data A1, node B
The upper carrier is monitored P, and data B1 is transferred after the carrier is turned off. Next, when a transmission request occurs for both nodes A and C, they wait until the carrier runs out, as in the previous explanation, but when nodes A and C start transmitting, a collision K occurs on the path line BL. Ru. Both nodes A and C constantly monitor collision K during data transfer, and when collision K occurs, -
Interrupt data transfer. Then, wait for a while and then try to transfer the data again. This time interval is determined by random number generation to prevent collisions from occurring again.

In FIG. 2, tA2' and tc+ are respective time intervals, and it is possible to transfer data A2 and data C1, respectively, without causing a collision K again.

However, if a large amount of data is to be sent continuously, and if all the data cannot fit into one packet due to the buffer φ size or the like and must be divided into multiple packets and sent, this method is inconvenient.

This will be explained with reference to FIG.

In Fig. 3, when node C wants to send multiple packets consecutively, it first finds a free space on the path line BL and sends JA.
Give the M signal. This is treated as a control command that informs other nodes that the path line BL will now be occupied.

The node C subsequently places the time T during which it occupies the pass line BL as data To on the pass line BL. Only if there is no collision at this time will node C be able to exclusively occupy the path line BL. Immediately upon receiving the data TD, the other node sets a time T to a timer and turns on this timer. While the timer is on, other nodes hold it until the timer turns off even if there is a transmission request. Node C can continuously send packets during this time T.

When the timer is turned off, the normal method shown in FIG. 2 is resumed.

The flow at the time of the above transmission is summarized as shown in FIG. tIIIn in FIG. 4 is the waiting time after the collision in normal conditions.

Note that the pass line BL is occupied by each node A-C.
There is a way to do it before, but usually master node M
It is carried out under the management of

As explained above, in a bus-type network system, the present invention sends data to the path line for the time it wants to occupy at its own node, and prevents other nodes from issuing transmission requests during that time. It has the advantage of being able to occupy the pass line and therefore ensuring the transfer of multiple packets.

[Brief explanation of the drawing]

Figure 1 is a block diagram showing the relationship between the bus and each node, Figure 2 is a timing chart during normal data transmission, and Figure 3 is a block diagram showing the relationship between the bus and each node.
The figure also shows a timing chart when the bus is occupied, and Figure 4 is a general flowchart when transmitting at each node.In the figure, BL is a pass line, A, B, and C are nodes, M is a master node, and TI and T2 are It's the Terminator. Figure 1 Figure 2 Figure 3 Figure 4

Claims (2)

[Claims] (1) In the /<su type network system,
An information transfer method characterized by transmitting data to a path line for a time that the own node wants to occupy, and monopolizing the path line by preventing other mates from issuing transmission requests during that time. (2) The information transfer system according to claim (1), wherein the time to be occupied is determined by a timer included in each node, and the timer value is set by a command.