JPH04286436A - Node equipment for independent synchronization multi-mediam lan - Google Patents

Abstract

PURPOSE:To prevent increase of the delay of relaying in each node by relaying a synchronization data relayed to a succeeding node not though a synchronization buffer. CONSTITUTION:A reception data is relayed to a succeeding node equipment via a relay path 54 and a selector 53 in a node equipment and inputted simultaneously to a bufer 51 for independent synchronization. The buffer 51 outputs the reception data according to a network synchronization clock to a synchronization terminal equipment 6. A transmission data from the terminal equipment 6 is inputted to a network synchronization buffer 52, from which the data is outputted according to the independent synchroniation clock, fed to a frame processing section 32 via a selector 53, superimposed on a synchronization data area of an information frame and sent to a succeeding node equipment. Thus, the connection of a synchroniging terminal equipment 6 to the node equipment is attained and jitter caused accompanying the regeneration of the network synchronization clock is absorbed without increasing the delay of relaying.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an independent synchronous multimedia LAN that accommodates synchronous and asynchronous data, and more particularly to a node device thereof.

0002

2. Description of the Related Art Conventionally, a dependent synchronization method, which is easy to configure, has been used as a multimedia LAN method for accommodating synchronous data and asynchronous data. Here, synchronous data refers to data that needs to be periodically transmitted in a predetermined amount. Furthermore, asynchronous data refers to data that does not need to be sent periodically.

[0003] A typical known example is “A1.2 G
bps optical loop LAN for
wideband office communica
IEEE Global Teleco
Communication Conference 1
There is 985,15-4. In a slave synchronization type LAN, clocks are distributed by each node device reproducing and relaying a clock generated by a master node. In this method, the entire LAN operates with a common clock,
LA by synchronizing the clock of the master node with the network.
Since the entire N network is synchronized, it is easy to connect synchronous devices. However, the jitter that occurs during clock recovery and relaying accumulates each time it passes through a node, and as the number of nodes increases, the jitter also increases, making it difficult to receive data, which limits the number of nodes that can be connected. be.

As a countermeasure against jitter accumulation during clock recovery and relay, there is a method of using an independent synchronization system for the LAN. For example, FFDI-II (FDDI Hybrid Ring
Control, Draft Proposed
American Standard, January
y 20, 1989), the use of an independent synchronization method is being considered. This draft standard adjusts the oscillation frequency deviation of each node by increasing or decreasing the number of bits in the control information area within a transmission unit (frame) of a fixed period.
And by introducing a frame structure, periodic data transfer is realized.

[0005]

[Problem to be solved by the invention] Multimedia LAN
As a synchronization method and a method for storing synchronized data,
When the above-described slave synchronization method and FDDI-II method are adopted, the following problems arise.

[0006] In the dependent synchronization method, as mentioned above, the problem is that the number of nodes is limited due to the accumulation of jitter.

[0007] Using an independent synchronization system such as the FDDI-II system can prevent the accumulation of jitter, but the frame length becomes undefined because the number of bits in the control information area within the frame increases or decreases. When accommodating synchronized data for transmitting and receiving frames at regular intervals, frame processing becomes complicated, a buffer is required, and relay delay increases.

Furthermore, when accommodating synchronous data, an elastic store (ES) is required to transfer the clock to the synchronous clock, which increases the relay delay for each node, making it difficult for a LAN system with a large number of nodes. Then, there is a problem that system transmission delay increases in proportion to the number of nodes.

[0009] An object of the present invention is to
An object of the present invention is to provide a node device for an independent synchronous multimedia LAN in which relay delay for each node does not increase.

0010

[Means for Solving the Problems] In order to achieve the above object, a multimedia LAN node device according to the present invention uses reference clock information generated by a node device (master node) having a clock source for network synchronization. Each node device that receives the transfer frame containing the transfer frame via the ring-shaped transmission path regenerates the network synchronization clock based on the reference clock information, synchronizes it with the network synchronization clock of the master node, and accommodates the synchronized data. Multimedia LA
In the N node devices, each node device operates with an independent synchronous clock and is connected to the transmission path, and a synchronous terminal interface portion which operates with the network synchronous clock and is necessary for connection of synchronous terminals. , a first synchronization buffer that inputs data from the common part in synchronization with the independent synchronization clock and outputs it to the synchronization terminal interface part in synchronization with the network synchronization clock; a second independent synchronization buffer inputting data in synchronization with the network synchronization clock and outputting data to the common part in synchronization with the independent synchronization clock; And processing is performed in the common portion without going through the second synchronization buffer.

[0011]

[Operation] By providing each node device with an independent clock oscillation source to form an independent synchronous LAN, the problem of limiting the number of nodes due to accumulation of clock jitter can be solved. Furthermore, by using fixed length frames, frame synchronization can be achieved even with a simple device configuration, and it has compatibility with synchronization data and facilitates connection with synchronization terminals.

[0012] By inserting and transmitting the change point position information of the reference clock for network synchronization into the control information area of the information frame in the LAN and reproducing it in each node device, the L
The entire AN can be network synchronized.

Furthermore, in each node device, the only part that operates with the network synchronous clock is the synchronous terminal interface part, and the synchronization buffer (error By removing the stick store from the relay path of each node device, in a LAN system with a large number of node devices in a loop, the relay delay of the entire system does not increase, and the capacity of the synchronization buffer is reduced. Since it becomes possible to ensure a sufficient amount of clock jitter, the tolerance against clock jitter is improved.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows the configuration of a multimedia LAN to which the present invention is applied. Node devices 2, 3, and 4 are transmission path 1
are connected in a ring shape. Each node device is
As described later, it consists of a common part and a synchronous terminal interface part. The common portion is connected to the transmission line 1 and operates with independent synchronous clocks from independent clock oscillation sources 21, 31, and 41. The master node 2 has a clock oscillation source 24 for network synchronization, the clock oscillated from this source is sampled as reference clock information by a clock sampler 23, and the reference clock information is sent to the control information area of the transfer frame by the frame processing unit 22. insert. On the other hand, the frame processing units 32 and 42 of the node devices 3 and 4 other than the master node
Then, reference clock information is extracted from the control information area of the received transfer frame, and the clock regenerators 33 and 43 regenerate the network synchronization clock. Furthermore, all frame processing units 22, 32, and 42 have a stuffing function to adjust the clock frequency and jitter of received data.

FIG. 2 shows a schematic configuration of a transfer frame used in a multimedia LAN to which the present invention is applied. The transfer frame includes a fixed-length information frame and a stuffing area for keeping the transfer frame at a constant period. The information frame has a control information area and a normal information area. Furthermore, the control information area includes a reference clock information area, into which reference clock information is inserted and transferred to each node device. Both synchronous data and asynchronous data are stored in the communication information area, and are accessed by the terminal interface section of each node device.

FIG. 3 shows the configuration of a multimedia LAN node device according to an embodiment of the present invention. Elements that are the same as those shown in FIG. 1 are given the same reference numerals. This node device 3 is comprised of a common part 8 that operates with an independent synchronous clock from an independent clock oscillation source 31 and a synchronous terminal interface part 5 that operates with a network synchronization clock regenerated by a clock regenerator 33. . The frame processing unit 32 of the common part 8 performs transmission and reception processing of synchronous data and asynchronous data. Received data is relay route 5
4 and the selector 53 to the next node device, and simultaneously input to the independent synchronization buffer 51. The independent synchronization buffer 51 receives received data according to the independent synchronization clock and outputs this data according to the network synchronization clock, thereby sending data synchronized with the network synchronization clock to the synchronization terminal 6. Transmission data from the synchronous terminal 6 is input to the network synchronization buffer 52. The network synchronization buffer 52 accepts transmission data according to the network synchronization clock and outputs this data according to the independent synchronization clock. This transmission data is sent to the frame processing unit 32 via the selector 53, where it is placed in the synchronization data area (FIG. 2) of the information frame and sent to the next node device. Both synchronization buffers 51 and 52 constitute an elastic store that performs clock switching between the network synchronization clock reproduced by the clock regenerator 33 and the independent synchronization clock from the independent clock oscillation source 31. The configuration shown in FIG. 3 makes it possible to connect the synchronization terminal 6 to the node equipment, and to absorb jitter that occurs due to regeneration of the network synchronization clock without increasing relay delay. .

FIG. 4 shows a synchronization buffer 51 of a node device.
(Fig. 3) shows the configuration. This synchronization buffer is a first-in-first-out (FIFO) type memory. The input data 61 synchronized with the independent synchronous clock 63 is
The output data is written sequentially from area a(0) of the synchronization buffer 51, and when it is written to the intermediate area a(n/2), the output data is synchronized with the network synchronization clock 64 in the order written from area a(0). It is read out as 62. The independent synchronous clock 63 and the network synchronous clock 64 have the same frequency over a long period of time, but over a short period of time there is a phase difference between the two clocks and jitter that occurs due to sample reproduction of the network synchronous clock. be. Independent synchronous clock 6
3 is higher than the frequency of the network synchronization clock 64, the data written to the synchronization buffer 51 per unit time is greater than the data read out, and initially up to half of the synchronization buffer 51 is written. The stored data gradually increases and becomes overflow when it exceeds the entire area a(n). Also, conversely, independent synchronous clock 6
When the frequency of 3 is lower than the frequency of the network synchronization clock 64, the data written per unit time is less than the data read, and the data that was initially written to half is gradually reduced and read out. Underfloat occurs when there is no more data. However, overflow and underflow occur when clock jitter is large, and as a countermeasure to this, clock jitter tolerance can be easily provided by increasing the number of regions (ie, capacity) of the synchronization buffer. The same applies to the synchronization buffer 52.

According to this embodiment, by increasing the capacity of the synchronization buffer, the delay time from when input data 61 is written to the synchronization buffer until it is output as output data 62 increases. As explained in connection with the node device in FIG. 3, since there is no relay delay in the node device, a synchronization buffer with sufficient capacity can be secured as a countermeasure against clock jitter.

[0019]

[Effects of the Invention] According to the present invention, since synchronized data to be relayed to the next node is relayed without using a synchronization buffer, a multimedia LAN can be realized without increasing the relay delay of data transmission at each node. becomes possible. Furthermore, since the capacity of the synchronization buffer is independent of relay delay, sufficient synchronization buffer capacity can be secured to counteract clock jitter.

[Brief explanation of the drawing]

FIG. 1 is a configuration diagram of a multimedia LAN to which the present invention is applied.

FIG. 2 is a schematic configuration diagram of a transfer frame used in the LAN of FIG. 1;

FIG. 3 is a configuration diagram of a node device according to the present invention used in the LAN of FIG. 1;

FIG. 4 is a configuration diagram of a synchronization buffer in the node device of FIG. 3;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1... Transmission path, 2, 3, 4... Node device, 5... Synchronous terminal interface part, 6... Synchronous terminal, 8... Common part, 21
, 31, 41...independent clock oscillation source, 22, 32, 4
2... Frame processing unit, 23... Clock sampler, 33, 4
3... Clock regenerator, 24... Clock oscillation source for network synchronization,
51, 52...Synchronization buffer, 63...Independent synchronous clock, 64...Network synchronous clock.

Claims (2)

[Claims] Claim 1: Each node device that receives a transfer frame containing reference clock information generated by a node device (hereinafter referred to as a master node) having a clock source for network synchronization via a ring-shaped transmission path uses the reference clock information. By regenerating the network synchronization clock based on the clock information, it is synchronized with the network synchronization clock of the master node, and an independent synchronization clock is provided to each node device in a multimedia LAN node device that accommodates synchronized data. a common part that operates and is connected to the transmission path; a synchronous terminal interface part that operates with the network synchronous clock and is connected to the synchronous terminal; and a synchronous terminal interface part that operates with the network synchronous clock and that receives data from the common part in synchronization with the independent synchronous clock; a first synchronization buffer that outputs data to the synchronous terminal interface part in synchronization with the network synchronous clock; a first synchronization buffer that receives data from the synchronous terminal interface part in synchronization with the network synchronous clock; a second independent synchronization buffer output to the common part, and data relayed from the previous node device to the next node device is processed in the common part without going through the first and second synchronization buffers. A multimedia LAN node device characterized by: 2. The common part includes a selector for selecting either the input to the first synchronization buffer or the output of the second synchronization buffer, and the output of the selector is connected to the next node. 2. The multimedia LAN node device according to claim 1, wherein the multimedia LAN node device transmits data to a multimedia LAN.