JPH02237251A - Packet multiplexing system - Google Patents

Abstract

PURPOSE:To efficiently multiplex packets by separating an abolishment disable packet from an abolishment enable packet and detachedly transmitting the former to a common buffer and the latter to a common buffer with abolishment function. CONSTITUTION:The common buffer means 110 multiplexing and outputting abolishment disable packets outputted from individual buffer means 101-10N, and the common buffer means with abolishment function 120 multiplexing and outputting the abolishment enable packets are provided. Since the common buffer with abolishment function 120 does not need to select the packets even if an abolishment control signal is received from a controller 140, the common buffer with abolishment function can easily be constituted. Since only the traffic quantity of the abolishment disable packets is needed to consider at the time of deciding the size and the transfer capacity of the common buffer 110 and the common buffer with abolishment function 120, it becomes facile, and an operation for making the packets much is set efficient, whereby the traffic storage capacity can be increased.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a Nokket exchange, and particularly to an i<'ket multiplexing method of a Nokket multiplexing device.

(Prior Art) Research on packet multiplexing devices in packet switching equipment is progressing.

Conventional technologies in this field include the IEICE Technical Research Report (1988) published by the Institute of Electronics, Information and Communication Engineers.
7 December 18) SE87-138 r Study of statistical multiplex control method in high-speed packet switching” Naoya Watanabe et al. P4
3 to 48, and various studies have been made.

An embodiment of a conventional packet multiplexing device for multiplexing packets is shown in FIG.

In Fig. 2, 201 to 20N are individual buffers 1 to N,
210 is a common buffer with a discard function, 220 is a controller, a
lxaN is a packet input input from a terminal etc., il-
iN is the packet output output from the individual buffers 1 to N, f2 is the multiple packet output, g2 is the usage rate information of the common buffer with a discard function, and h2 is a discard control signal.

Both discardable packets and non-discardable packets generated in a burst manner from a terminal etc. are input to individual buffers 201-20N as inputs a1-aN, and smoothed.

The packet output H-iN output from each individual buffer is input to the common buffer with discard function 210 and multiplexed. The common buffer with discard function outputs the multiplexed packet output f2 to the next stage.

The common buffer with discard function 210 also outputs usage rate information g2 of the common buffer with discard function to the controller 220.

The controller 220 identifies the congestion state of the common buffer with discard function 210 based on the information g2, and outputs the discard control signal h2 to the common buffer with discard function 210 according to the state.

The common buffer with discard function 210 selects only discardable packets from the accumulated discardable packets and non-discardable packets and discards them, in response to the discard control signal h2.

(Problem to be Solved by the Invention) However, in the device having the above configuration, in the common buffer with a discard function 210, whether discard is possible or not is possible. 4 kets that cannot be discarded and 4 kets that cannot be discarded. The problem is that the configuration of the common buffer with discard function 210 becomes complicated because only packets have to be selected and discarded, and when determining the buffer size and packet transfer capacity of the common buffer with discard function 2lθ. Which packets can be discarded and which ones cannot be discarded? There is a problem in that it becomes complicated because it is necessary to take into account the traffic volume of the network.

This invention provides the common buffer 2lθ with a discard function described above.
The purpose is to provide an excellent packet multiplexing method that efficiently multiplexes Pakenotes by eliminating the problems of complicated configuration and difficulty in determining buffer size and transfer capacity. do.

(Means for Solving the Problems) The means adopted by the present invention to solve the above problems is that packets are input in bursts, and either discardable packets that can be discarded or non-discardable packets that cannot be discarded. A packet multiplexing device for multiplexing identifiable packets includes an individual buffer means for smoothing packets input in bursts and separating and outputting discardable packets and non-discardable packets; buffer means for multiplexing and outputting non-discardable packets and outputting the discardable packets output from the individual buffer means; A common buffer with a discard function for discarding packets, and a multiplex/S sofa that further multiplexes and outputs the multiplexed packets that cannot be discarded and the multiplexed packets that can be discarded, and also outputs its own buffer usage rate. and a control means for knowing the congestion state of the multiplex buffer means from the buffer utilization information outputted from the multiplex buffer means and outputting a discard control signal in accordance with the state.

(Operation) According to the present invention, since the packet multiplexing device is configured as described above, packets that cannot be discarded are stored in the common buffer, and packets that can be discarded are separated from packets that can be discarded in the common buffer with a discard function. Even when a discard control signal is received from the controller, the common buffer with discard function stores only packets that can be discarded, so there is no need to select four packets, so the common buffer with discard function can be easily configured. I can do it.

When determining the size and transfer capacity of the common buffer and the common buffer with a discard function, it is easy to mainly consider the traffic volume of non-discardable packets, and the multiplexing of packets can be efficiently operated. In addition to being able to
Rahic capacity can be increased.

(Embodiment) FIG. 1 is a block diagram showing an embodiment of this invention,
lOl to 10N are individual buffers 1 to N, 1lo is a common buffer that stores packets that cannot be discarded, l2θ is a common buffer with a discard function 130 that stores packets that can be discarded, is a multiplex buffer, 140 is a controller, al-aN is a terminal, etc. (not shown) packet input to each individual buffer,
b1 to bN are non-discardable packets output from individual buffers, c1 to cN are 4 discardable packets output from each individual buffer, d1 is multiple non-discardable packets output from the common buffer 110, and e1 is a discard function multiple discardable packets I. outputted from the common buffer, f1 is multiplexed packets outputted from the multiplexing buffer 130, g1 is buffer usage rate information outputted from the multiplexing buffer 130,
hl is a discard control signal output from the controller 140.

The operation of FIG. 1 will be explained next.

Both discardable packets and non-discardable IR packets generated in a burst manner by a terminal etc. (not shown) are transmitted from each channel.
9-ket input a1 - separate buffer 10 as aN
1 to 1ON.

Input packets cannot be discarded in each individual buffer.
The packets are separated into packets and discardable packets, smoothed, and the non-discardable packets bl-bN are sent to the common buffer 110, and the discardable packets c1 to cN are sent to the common buffer with a discard function.

The non-discardable packets b1 to bN inputted from each individual buffer 101-ION are temporarily stored in the common buffer 110, then multiplexed and sent to the multiplex buffer 130.

In addition, the discardable bucket }cl-cN input from each individual buffer 101-ION is the common buffer with discard function 120
After being temporarily stored, the data is multiplexed and sent to the multiplex buffer 130.

The multiple discardable packets dl output from the common buffer 110 and the multiple discardable packets el output from the common buffer with discard function 120 are input to the multiplex buffer 130, where they are temporarily stored and then multiplexed into multiple packets.
It is output to the next stage as fl.

The multiplex buffer 130 constantly monitors its own packet storage amount, calculates the buffer usage rate, and sends it to the controller 140 as buffer usage rate information g1.

The controller 140 learns the congestion state of the nine multiplex buffers 130 based on the buffer usage rate information gl output from the multiplex buffer 130, creates a discard control signal h1 based on the congestion state, and sends it to the common buffer with discard function 120 described above. Send.

The common buffer l2θ with a discard function controls the packet traffic flow rate by discarding some or all of the low-priority discardable packets accumulated in the buffer 120 in response to a discard control signal sent from the controller. , it is possible to efficiently operate packet multiplexing, and the traffic capacity can be increased.

(Effects of the Invention) As explained in detail above, according to the present invention, packets that cannot be discarded and packets that can be discarded are separated in each individual buffer, and the former are sent out in a common buffer and the latter in a common buffer with a discard function. By doing so, it is expected that the configuration of the common buffer with a discard function will be simplified, and the buffer sizes and packet transfer capabilities of the common buffer and the common buffer with a discard function will be easily determined.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of a packet multiplexing device of the present invention, and FIG. 2 is a block diagram of an embodiment of a conventional packet multiplexing device. Explanation of symbols of main parts 101 to ION...Individual buffer, 110...Common buffer, 120...Common buffer with discard function, 13
0...Multiple buffer, 140...Controller, al-a
N...packet input, bl-bN...packet that cannot be discarded 1., cl=cN...packet that can be discarded, dl...
i9 packet that cannot be multiplexed, el...packet that can be discarded, fl...multiple packet, g1...buffer usage rate information, hl...discard control signal

Claims (1)

[Scope of Claims] A packet multiplexing device that inputs packets in bursts and multiplexes packets that can be identified as discardable packets or non-discardable packets, individual buffer means for smoothing the packets and separating and outputting discardable packets and non-discardable packets; common buffer means for multiplexing and outputting the non-discardable packets output from the individual buffer means; and the individual buffer means. common buffer means with a discard function for multiplexing and outputting discardable packets output from the control means and discarding the discardable packets in response to a discard control signal given from the control means; A multiplex buffer means that further multiplexes discardable packets and outputs the same and outputs its own buffer usage rate, and the congestion state of the multiplex buffer means is known from the buffer usage rate information outputted from the multiplex buffer means, and the congestion status is adjusted to that state. and a control means for outputting a discard control signal in accordance with the control means, and after the congestion state of the multiplex buffer means is identified by the control means, the common buffer means with a discard function is instructed,
A packet multiplexing method characterized by discarding discardable packets from a common buffer with a discarding function.