JPS60160744A - Transmission start control system of local network - Google Patents

Abstract

PURPOSE:To decrease the probability causing an unnecessary collision for communication requiring the real time performance and the instantaneity by giving a transmission probability high in generation of an idle channel to a packet having a high priority. CONSTITUTION:When a transmission request from stations A and B is generated as shown in the thick arrows respectively during the transmission of the station A, the station B with low priority moves to a retest processing immediately after the channel is in use. On the other hand, even after the station C having a packet with high priority detects the busy state of the station A, the station C continues the supervision of the said channel. When the station A completes the transmission, the station C detects immediately that the channel is idle and starts the transmission. Moreover, the station B checks the idle channel at each retrial time and starts the transmission after the end of data transmission of the station C. Thus, the station C makes transmission with a minimum time delay without causing any collision.

Description

[Detailed Description of the Invention] (Field of Application) Regarding transmission start control methods, in particular, in a distributed system of multiple stations connected to communication cables, local This relates to a network transmission start control method.

(Prior art) Conventional path-based local network (hereinafter referred to as LAN)
FIG. 1 shows a schematic configuration of an example of a communication system.

In FIG. 1, a coaxial cable 3 installed as a transmission line has both ends connected to an impedance matching terminator 1.2 having a resistance value equal to the characteristic impedance.

Each station S, C, R... is connected to a T connector (
Tap) Connected to coaxial cable 6 through 4R ~ 4N. All these stations have basically the same configuration. Therefore, in FIG. 1, only the main parts of the eight stations connected to the T connector 4l are shown.

Each station is equipped with a user device 5 including a computer and a telephone.

The user device 5 includes a transmitter (encoder) 51 for transmitting a digital signal in units of packets to other stations, and a receiver for receiving digital signals in units of tents sent from other stations. A decoder 52 and a terminal controller 56 for controlling the terminal are provided.

The signal output from the transmitter 51 is temporarily stored in a transmission buffer memory 61.

Then, they are read out all at once at a predetermined time using a clock signal equal to the transmission speed on the coaxial cable 6, which is the transmission medium.

The read signal is from the transmission logic circuit 62 Fc.

It is converted into a predetermined packet. After passing through the transmission node core amplifier 63, the signal is sent out onto the coaxial cable δ through the T connector 4l.

On the other hand, all the packet signals transmitted on the coaxial cable 3 are received by the reception buffer amplifier 64 through the T connector 4l, and input to the reception logic circuit 65.

The reception logic circuit 65 selects only the packet addressed to its own station from the received notifications, and temporarily stores it in the reception buffer memory 66. The stored signals are continuously read out in the receiver 52 using a predetermined clock. This provides a received output signal.

Signals are transmitted and received as described above, and the transmission clock used in this case is transmitted by the transmission clock oscillator 67.
generated from.

The transmission control circuit 69 controls the terminal controller 56 based on the received signal addressed to the own station obtained from the reception logic circuit 65, and also controls the transmission logic circuit 6 according to instructions from the terminal controller 56.
Control 2.

Further, the collision detection circuit 74 checks whether a collision occurs with a packet signal transmitted from another station when the first packet signal is transmitted from the own station.

When a packet collision is detected, the transmission logic circuit 62 interrupts sending the packet signal to the coaxial cable 6 in response to a control signal from the collision detection circuit 74. Thereafter, retransmission is attempted according to a predetermined procedure.

In the conventional bus type LAN as mentioned above, multiple stations are connected to a common transmission medium (channel) K, and the right to use the channel (transmission right, access right) is acquired by C8MA/C.
D (Carrier Sense Multiple
Acc@ss with Co11isionDate
This is done using a distributed method called ction.

In this system, a station attempting to acquire a transmission right monitors the usage status of the channel and starts transmission after confirming that the channel is unused, in which no station is transmitting or receiving signals.

In this case, since there is a propagation delay in the transmission path, two or more stations may detect an unused channel state and start transmission almost simultaneously. In such a case,
This causes what is called a packet collision.

There are three ways to control the timing of the transmission start procedure when the transmission path is in use or when a packet collision occurs.Generally, there are 1-persistent C8MA/ The CD method is the most widely used.

1) Constantly monitor channel availability, and when a channel is detected to be available, start transmission with a probability of 1 (in other words, whenever a channel is detected to be available), and if a packet collision occurs, the scheduled transmission will be started. 1-Purge stent method that returns to channel monitoring mode after waiting the retry time.

2) Constantly monitor the availability of channels, and when a channel is detected to be vacant, (a) start transmission with probability p (p is a positive number less than or equal to 1), and (i) start transmission with probability (1-p) , it waits for the maximum round-trip propagation delay time, and when said delay time has elapsed, if the channel is unused, it starts transmitting (/→ If a packet collision occurs, the scheduled retry time p-purge stent method that returns to channel monitoring mode after waiting for

5) When the free state of the channel is not detected and when a packet collision occurs, a retry algorithm retries the start of transmission after a predetermined retry time (non-psrsiat).
snt) Method O FIG. 2 shows a control flowchart for the p-purge stent method, and each step thereof will be explained below.

Step S1: Determine whether a transmission request has occurred. When a transmission request occurs, the process advances to step S2.

Step S2... It is detected whether the transmission line is in use, that is, whether the channel is vacant. When the channel is vacant, the process proceeds to step S3, and when the channel is not vacant, that is, when the transmission path is in use, the channel is continuously monitored for availability.

Step S3...The probability p is calculated, and when the probability p is larger than the planned value, the process proceeds to step S4, and on the other hand, when the probability p is smaller than the planned value, the process goes to step 87''4. In this case, the comparison calculation of the probability p can be performed by, for example, taking a random number from 0 to 1 and comparing it with a value set in advance for each station.

Step S4: When the determination in step s3 before the song is established, a packet is sent to the transmission path.

Step S5: It is determined whether a packet collision has occurred as a result of the transmission in the previous step s4.

If no collision occurs, the acquisition of the transmission right is successful. Further, when a collision is detected, the process advances to step S6.

Step S6: Wait for a time corresponding to a retry interval set in advance, and return to step s2 after that time has elapsed.

Step S7: Waiting for a time corresponding to the maximum round-trip propagation delay time. If the time has elapsed, the process advances to step S8.

Step S8: It is determined whether the transmission line is in use, that is, whether the channel is free or not. If it is vacant, the process advances to step s4 and a data packet is sent. Furthermore, when the channel is not available, the process returns to step s2.

FIG. 3 is a control flowchart for the 1-barge stent system. In this figure, the steps with the same numbers as in FIG. 2 perform the same processing. This method uses the previously mentioned p
- The difference from the purge stent method is that when an empty channel is detected in step S2, data packets are immediately sent out without calculating or determining the probability.

FIG. 4 is a control flowchart for the non-verge stent method. In this figure, the steps with the same numbers as in FIG. 2 perform the same processing.

This method differs from the 1-purge stent method shown in FIG. 3 in that when it is determined in step S2 that the channel is in use, the process always returns to step s2 after waiting for a time corresponding to the retry interval.

FIG. 5 is a timing chart showing the above-mentioned 1-barge stent system new channel access.

Now, while station A is transmitting, transmission requests from stations B and 0 are
Assume that trap 1 occurs as shown by the thick arrow in the figure. In addition,
In the figure, (L) and (H) indicate the priority of each request, with L representing "low priority" and H representing "high priority", respectively.

In the 1-verge stent C8MA/CD method.

Since priority control for acquisition of transmission rights is not performed, "A"
Simultaneously with the end of the station's transmission, both B and 0 stations begin transmitting. This causes packet collisions, as indicated by the X mark in the figure.

Note that in this figure, the reason why there is a gap after the transmission of station A ends is due to the propagation delay of the channel (transmission line).

As can be easily understood from the above explanation, regardless of which method is used to control the transmission start timing, the same transmission start control method is assigned to all stations. If this happens, there will be variations in the time from the generation of a transmission request to the successful transmission.

Therefore, especially in the case of a communication device that emphasizes real-time performance, such as conversational voice communication, there is a drawback that becomes an extremely large obstacle.

(Purpose) The present invention was made to eliminate the above-mentioned drawbacks, and its purpose is to minimize the probability of collision occurring in communications (or stations) that require real-time performance and immediacy. Another object of the present invention is to provide a local network transmission start control method that can reduce variations in transmission delays.

(Summary) In order to achieve the above object, the present invention provides a system that gives higher priority to communications or stations that require real-time performance or immediacy compared to communications or stations that do not require so much immediacy. For example, as a transmission start control method,
The 1-purge stent method is assigned to the former, and the non-purge stent method is assigned to the latter, thereby minimizing the probability that the former will cause an unnecessary collision.

(Example) The invention will now be described in detail with reference to one drawing.

FIG. 6 is a timing chart showing the operation of one embodiment of the present invention, that is, channel access. In the figure, the same reference numerals as in FIG. 5 represent the same or equivalent parts.

In this example, for high priority packets.

1-purge stent method is assigned, while non-purge stent method is assigned to packets with low priority, and channel acquisition, that is, transmission start control is performed for each packet.

Therefore, in FIG. 6, as in the case of FIG. 5, transmission requests from stations A and B occur while station A is transmitting, as shown by thick arrows.

(1) After detecting that the channel is in use, the low-priority B station immediately moves to retry processing. That is, the state of the channel is not monitored until a predetermined retry time has elapsed, and it is detected whether the channel is empty or not only after the retry time has elapsed (see FIG. 4). On the other hand, (2) Station 0, which has a high-priority packet, continues to monitor the channel even after detecting that station A is in use (the third
(see figure).

Therefore, when station A finishes transmitting, station 0 can immediately detect this (that the channel is empty) and start transmitting. Therefore, station 0 can successfully acquire the transmission right without causing a collision and with a minimum time delay as shown in FIG.

Note that in FIG. 6, station B checks the availability of the channel at every retry time, and successfully starts transmission after the data transmission of station 0 is completed.

As described above, according to the present invention, it is possible to prevent collisions between packets with different priorities and to minimize the transfer delay time of packets with high priority.

However, even in this embodiment, collisions between packets with the same priority may occur.

The second embodiment of the present invention is an extension of the first embodiment described above, in which NFL priority is assigned to each packet (or station).

In that case, the 1-purge stent method is assigned to the highest priority, the non-purge stent method is assigned to the lowest priority, and the p-purge stent method is assigned to those with intermediate priorities. It subsides.

If the value of the probability p is set differently, an arbitrary number of priorities can be set for a packet (or station). At this time, as is clear, the closer the value of probability p is to 1, the higher the priority is (and conversely, the closer the value of probability p is to 0, the lower the priority is).

(Modified Example) The above describes an example in which the present invention is applied to a LANK in which a large number of terminal stations are connected to one transmission line (cable, etc.), but other LANs - for example, using optical fibers etc. Even in star format LAN, C8MA/CD
If this method is adopted, the present invention can be applied as is.

FIG. 7 shows an outline of a configuration example in this case. In FIG. 7, a plurality of stations A, B, C, and D are coupled to a star coupler Q by transmission lines La, Lb, Lc, and Ld K, such as optical fibers.

(Effects) As is clear from the above description, according to the present invention, the following effects are achieved.

(1) It is highly reliable because it is completely distributed control.

(2) Priority control reduces packet collisions, so communication delays are shortened and variations in communication delays are reduced. Therefore, it is possible to handle highly immediate communication.

(3) If a channel is vacant, transmission is started immediately after this is detected, regardless of the priority, so there is no increase in delay when the load is low.

(4) Priority (probability p
) and 'C'P.
(transmission line extension, etc.) is easy.

(5) Existing systems (e.g. Ethernet
) can be realized by simply adding (or changing) the control algorithm without changing the hardware logic at all.

(6) High load characteristics can be improved without sacrificing the advantages of the current system.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing an example of the configuration of a general conventional local network, Fig. 2 is a flowchart showing the transmission start control operation in the case of the p-purge stent method, and Fig. 3 is a flowchart showing the transmission start control operation in the case of the p-purge stent method. Fig. 70 is a flowchart showing the transmission start control operation in the case of the non-purge stent method, and Fig. 5 is a time chart showing the state of channel access according to the conventional control method. , FIG. 6 is a time chart showing an example of channel access according to the control method of the present invention, and FIG. 7 is a block diagram showing another example of the configuration of a local network to which the present invention can be applied. 3...Coaxial cable, 41~4N...T connector (
tap), 5... User device, 51... Transmitter (encoder), 52... Receiver (decoder), 53...
Terminal controller, 61... Transmission buffer memory, 62...
- Transmission logic circuit, 66... Transmission buffer amplifier, 64
. . . Reception buffer amplifier, 65 . . . Reception logic circuit,
66... Reception backup memory, 67... Transmission lock oscillator, 69... Transmission control circuit, 74 Collision detection circuit Patent attorney Michi Hiraki 1 non-individual Figure 6 Figure 4 O 5 Figure 26 Figure (L) (H) 27 Figure

Claims (1)

[Scope of Claims] (1) Digital signals transmitted over transmission paths such as communication cables are given opportunities to transmit digital signals to each station using a competitive method, and the signals are sent and received in a time division multiplexed manner using a packet format. In the local network transmission start control method, when each station requests packet transmission onto a transmission path, the station monitors the usage status of the transmission path and detects whether the transmission path is not in use or the channel is empty. When detected, packet transmission is executed with a preset probability (positive number less than 1) for each station or each packet, and when the usage status of the transmission path is detected, transmission is performed for each station or each packet. A transmission start control method for a local network, characterized in that the usage status of the transmission path is monitored after waiting for a preset time (including 0). ? ) When the preset probability is P, (1-
P), wait for the maximum round-trip propagation delay time, and once the waiting time has elapsed, monitor the usage status of the trap transmission line again;
81. The local network transmission-start control method described in Patent No. 81, characterized in that when a channel free state in which a transmission path is not used is detected, packet transmission is immediately executed. (3) A 1-persistent method is assigned to a station or packet with a high priority, and a non-persistent method is assigned to a station or packet with a low priority. Local network transmission start control method described in Section 1. (4) Start of transmission of a local network of a patent characterized in that the 1-persistent method is assigned to stations or packets with high priority, and the P-purge stent method is assigned to stations or patents with low priority. control method. (5) The P-persistent method is assigned to stations or packets with high priority, and the non-persistent method is assigned to stations or packets with low priority. Local network transmission start control method described in Section 1. (6) The station or packet with the highest priority shall have 1
- A persistent method is assigned, a non-persistent method is assigned to a station or packet with the lowest priority, and a P-persistent method is assigned to a station or packet with a medium priority. A local network transmission start control method according to claim 1.