JPS6130835A - Digital signal transmission system - Google Patents

Abstract

PURPOSE:To make both digital signals available in mixture in one cable by not stopping transmission or reception of a digital signal of an Faxnet of the own station even if a signal collides with a digital signal of the CSMA/CD system. CONSTITUTION:A transmission logical circuit 52' does not stop the transmission of a signal for a prescribed time from the start of packet transmission even if a collision detection signal is outputted from a collision detection circuit 62. When the transmission logic circuit 52' receives a transmission request signal from a transmission timeout detection circuit 70, the circuit 52' transmits a packet even if the packet order is not the order of own station transmission and a collision detection signal is outputted from a collision detection circuit 62. A reception logic circuit 55' starts the reception of packet and then does not stop the reception of packet even if collision is detected from the collision detection circuit 62. Further, when the reception logic circuit 55' receives a reception request signal from a reception timeout detection circuit 71, the reception operation is attained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a digital signal transmission system for transmitting data by time division multiplexing using communication cables.

[Conventional technology]

With the spread of electronic computers and the development of digital signal processing technology, data communication, which processes information online by combining a communication system and a data processing system, has been in the spotlight. In particular, for small-scale communication systems such as in-house communication conducted within the premises of government offices and companies, packet-based communication methods using communication cables such as coaxial cables are preferred due to their economic efficiency, reliability, and high transmission efficiency. is attracting particular attention.

In this packet-based communication system, a communication cable for bidirectional transmission is installed in a research institute, etc., and a large number of stations (personal stations) are connected to this cable. And each station, for example, ioo.

A message divided into data blocks of ~2000 bits is transmitted. Headers such as destination, serial number, etc. are added to the message. In this communication system, the network itself is a passive transmission medium with no control functions, and control is completely distributed to each station. Therefore, each station checks the availability of the transmission line and starts sending messages.
If a collision occurs with another packet being transmitted, both stations stop transmitting. A station that has stopped transmitting will attempt to retransmit the message after a random waiting period. This is called a CSMA/CD type packet switching communication system.

With this type of communication method, users at each location can of course access one computer;
It is possible to mutually utilize hardware such as storage devices and software such as programs that are distributed in various locations.

In other words, with this communication method, devices such as high-speed or high-precision printers and large-capacity files, which were concentrated on a large central computer in TSS (time sharing system), can be used in a distributed state at various locations. can. Therefore, in addition to saving resources and improving usage efficiency, it is also possible to develop large software systems by accommodating programs and data. In addition, in this communication method, each user (personal station)
There is no priority order (equal) in the use of the transmission path between the stations. Therefore, there is no master-slave hierarchy between stations that is often seen in other communication systems, and communication can be performed between any connected stations. Furthermore, since the transmission line such as a coaxial cable is completely composed of passive circuits, a highly reliable system can be easily created.

Although this communication method has various features as described above,
Since each station starts transmitting data arbitrarily, there is a possibility that packets will collide on the same transmission path. Naturally, this collision between packets becomes more prominent as the system becomes more efficient in using the transmission path.

To solve these problems, a signal transmission system called Priority Ethernet or Reservation Ethernet has been proposed. Of these, Priority Ethernet) (Pr1ority E)
In a system called ``thernet'', signal transmission from each station is prioritized using a preamble within a packet. If packets collide, the packet with higher priority is transmitted preferentially. Also, Reservation Ethernet
ernet), a station for mode specification (master station) is permanently installed, and in reservation mode, each station (personal station)
Check whether there is a signal to be transmitted each time and the amount of information to be transmitted. Based on this result, the order of packets to be transmitted by each station is determined for each frame, and signal transmission is performed in a time-division manner in the transmission mode.

However, according to the former proposed signal transmission method, there still remains the problem of variations in transmission delay time due to collisions between packets with the same priority. Therefore, it is unsuitable for real-time transmission, such as conversational voice communication, where the correspondence between transmission and reception in real time is important. Furthermore, according to the latter signal transmission method, the above-mentioned equality between stations is lost due to the presence of the master station. That is, in this system, if a failure occurs in the master station, data communication becomes impossible, and in this sense, the reliability of the system decreases.

As an improvement on the above points, a line-switched communication system called F axnet as shown in FIG. 15 has been proposed.

In the communication system shown in FIG. 15, a coaxial cable 1 installed as a transmission path has both ends connected to an impedance matching terminator 2 having a resistance value equal to the characteristic impedance. Each station has a T connector (
Tap) 3. It is connected to the coaxial cable 1 through ~3 circles. Since all these stations have basically the same configuration, in the figure the Δ
Only the main parts of the station are shown.

Each station is equipped with a user device 4 including a computer and a telephone. The user device 4 includes a transmitter (encoder) 41 for transmitting digital signals in packet units to other stations, and a receiver (encoder) 41 for receiving digital signals in packet units sent from other stations. A decoder) 42, a terminal controller 43 for controlling the terminal, and the like are provided.

Of these, the signal output from the transmitter 41 is temporarily stored in the transmission buffer memory 51. The information is then read out at a predetermined time using a clock signal equal to the transmission speed on the coaxial cable l, which is the transmission medium. This read signal is converted into a predetermined packet by the transmission logic circuit 52. After passing through the transmission buffer amplifier 53, T
It is sent out onto the coaxial cable 1 through the connector 31 .

On the other hand, all packet signals being transmitted on the coaxial cable 1 are sent to the T connector 3. The received signal is received by the receiving buffer amplifier 54 through the receiving buffer amplifier 54. The reception logic circuit 55 selects only the packet addressed to its own station from the received packets and temporarily stores it in the reception buffer memory 56. This stored signal is continuously read out in the receiver 42 using a predetermined clock. This provides a received output signal.

The transmission and reception of signals is performed as described above, and the transmission lock a used for these is the transmission lock oscillator 57.
generated from. The transmission control circuit 61 is connected to the reception logic circuit 5
The terminal controller 43 receives the received signal addressed to its own station obtained from 5.
It also controls the transmission logic circuit 52 according to instructions from the terminal controller 43. Further, the collision detection circuit 62
When transmitting a packet signal, it is detected whether or not there is a collision with a packet signal of another station.

Each station also has a frame control circuit 80, tl
A timer circuit 81 and a packet management circuit 82 are provided.

As shown in FIGS. 19 and 20, the frame control circuit 80 receives the bit clock a (which has a frequency approximately equal to the transmission lock of the packet on the transmission path) from the transmission lock oscillator 57. It also receives the master detection signal f and master reset signal g from the reception logic circuit 55, and generates a frame clock h and a master window signal 1. In this circuit, when the master detection signal f is on, the frame tarot clock h and master window 1 are generated by the master reset signal g, and when the master detection signal f is off, the trigger pulse generated inside the circuit is generated. The frame clock h and master window 1 are generated.

1. The timer circuit 81 is a timer circuit for detecting that there is a station that has finished transmitting a packet, and uses the received signal j and frame clock h as input signals, and the timer circuit 81 is a timer circuit for detecting that there is a station that has finished transmitting a packet. Start the timer and reset it when the reception signal turns on. If the reception signal does not turn on even after tl time has elapsed after starting the timer, a timeout signal k is output. Note that when the time-out signal k is output, the tl timer is automatically restarted and checks for the existence of the next packet.

The packet management circuit 82 receives frame clock h1 received signals J and tl as shown in FIGS. 22 and 23.
It is supplied with a time-out signal from a timer circuit 81. It detects the packet transmission order of its own station from these human signals and outputs the transmission control signal 1. This output signal is supplied to the transmission control circuit 61, and this transmission control circuit 6
1, a transmission request signal to the transmission logic circuit 52 is generated under control together with a transmission signal from the controller of the terminal device.

FIG. 119 shows details of the frame control circuit 80. Referring to this figure, the frame control circuit 80 is composed of a bit counter 90, a flip-flop 91, and several gate elements. The bit counter 90 is a counter that counts bit clocks, and generates a frame clock signal according to the states of the master detection signal f and the master reset signal g, as described above. This frame clock signal is input to a flip-flop 910 set S, and a master window 1 is output from the Q output of this flip-flop 91.

The operation of this frame control circuit will become clear from the timing chart of FIG.

FIG. 12 shows the relationship between the frame clock output from this circuit and the packet signal on the transmission path.

FIG. 22 shows details of the packet management circuit 82. Referring to this diagram, the packet management circuit 82 includes a packet order memo I 7100, a packet counter 101
, tl It is composed of a timer timeout output counter 102, a comparator 103, and an adder 104, and is further provided with one flip-flop 105 and one ant gate 106. The packet order memory 100 is a memory that stores the order of buckets 7) to be sent by the own station, the bucket counter 101 is a counter that counts the number of packets in one frame, and the 1+ timer timeout output counter 102 is a tl timer. This is a counter that counts the timeout output from. The adder adds the number of packets and the number of timeout outputs of the TL timer, and the comparator compares the added value with the stored packet number and determines whether the added value is one packet number before the packet number. When the value of frame clock h
generates an output signal according to the

The transmission control method based on this method will be described below.

(2) FIG. 24 shows the transmitter packet of the station that starts transmission first when the transmission path is silent.

In this case, the packet is sent out using the following procedure.

■ Monitor carrier sense for one frame (7. If there is no carrier sense output, transmit packets at a cycle according to the frame clock of the local station. At this time, write the master signal in the control field of the packet to be transmitted. A flag indicating that the packet is a packet may be set. Alternatively, an area may be provided in the control field to indicate the packet order, and the fact that the packet is #1 may be indicated there.

■ Send the first packet and if there is no collision with another station's packet, send the packet according to the frame clock.

■ When the first packet collides, perform backoff and monitor the carrier sense between one frame immediately before sending the packet, and if there is a carrier or not, resend the packet according to the respective procedure. do.

(2) If there is a signal on the transmission path, follow the steps below. (Refer to Figure 25) ■ Detects the master packet, counts the number of packets after the master packet including the master packet using the packet counter, and stores the data in the packet counting memory when the master window in the next frame is detected. to be memorized. (The master window will be described later.) ■ Detect an empty area and determine whether the following conditions are satisfied.

Empty area ≧ (Packet size + 2a + α) (Detection of empty area will be explained later) ■ If the judgment in ■ is OK, count the packets from the master packet in the next frame, and calculate the packets stored in the previous frame. When a few people arrive, check the carrier of the packet.
Sense-off is detected, and a packet is sent after α time from that point.

For example, in the case of FIG. 25, packet #3 of every frame is detected, and the packet is sent out a time α after the carrier sense of the packet is turned off.

(Own station - 0 station) ■ If a packet is sent and there is a collision, perform a back-up and retry.

■ When your own station sends a packet, if there is a station that has finished transmitting, follow the procedure in ■.

(2) When there is a station that has finished transmitting a packet (see FIG. 26) (2) Each station has a frame clock and a timer for time t0 that starts when the carrier sense of each packet is turned off. This timer is reset by carrier sense on and the frame clock, and if it is not reset by t1 time, it will automatically restart after tl timeout.

t, = 2a+β (β〉α) ■ Each station has the number of packets from the master packet plus 1+
When it detects that the total value including the number of times the timer has timed out has reached the value one before the packet number that the local station should send, if there is a packet there, turn on the carrier sense. The packet is sent out α time after detection.

If there was no packet immediately before, the tl timer times out, so if this is detected, the packet is sent out after α time.

■ The packet counter of each transmitting station is 1 for each frame.
The number of packets including the master packet is counted from the master packet to the order in which the local station sends the packet. For example, in the F(n) frame shown in FIG. 26, the packet counter value of the B station is 2.0 and the packet counter value of the B station is 3.

If a station finishes transmitting between the master packet and its own station's turn to send packets, the value of the own station's packet counter will be less than the value in the previous frame by that amount. .

Therefore, in the next frame, packets are sent out in accordance with this updated packet sending order.

Note that the packet counter of stations that are not transmitting is the number of packets between master packets (including master packets).
are being counted.

■When the master packet disappears (see Figure 27) ■Each station monitors the master packet, and if the master packet cannot be detected at the position where it should appear in that frame, the own station will The existence of a master packet is monitored by the timing of sending a packet, and if there is no master packet, the immediately preceding tl is used.If the timer times out, α
After a certain period of time, it sets the master flag of its own sending packet and sends it out.

■ In the next frame, the station that became the master becomes the master at the #l packet position according to its own frame clock.
Send a packet.

■Detecting and checking master packets (see Figure 28) ■When a master packet is detected, the master window is Start the timer.

■ Generate a master window at the position where the master packet should appear in the next frame. The window width of this master window is determined so that the tl timer output can be detected and the master flag can be detected even in the worst case when the master packet position changes when the master window transitions.

■ Each station checks the presence or absence of the master flag within this master window, the presence or absence of a carrier, and the timeout output of the tl timer.

■ Within the master window, t, no timer timeout output (, master flag or carrier sense
When ON is detected, it is regarded as a master packet, and the master window timer is reset and restarted at a certain reference point determined by the mark (■) of the master packet. (A in the figure) ■ When the timeout output of t and timer is detected within the master window, it is determined that there is no master packet, and in that case, the periodic timer is restarted from the end of the master window. .

(Middle opening in the figure) Also, the end of the master window must be aligned with a certain reference point defined in ■.

■ Detection of empty area (see Figure 29) ■ Each station detects a master packet, and when the carrier sense of this master packet turns off, starts the empty area detection counter (a in the figure) ■ Next If a packet is detected before the master window of the frame is detected, the counter is stopped when the carrier sense is turned on (b in the figure), and the counter is reset and restarted when the carrier sense is turned off. .

■ When the master window is detected, stop the counter and read the counter value at that time (C in the figure).

This value becomes the empty area.

Here, the value a is the propagation delay time between the farthest stations.

Further, the value α time is a value determined based on hardware performance in order to maintain an appropriate packet interval.

That is, the time it takes for the channel to recover to a silent voltage level after the last bit of a packet on the channel has passed, and for the other station's receive logic to process the packet it just received and restore it to a ready state. This value is determined based on the time required for

(2) is to prevent the packet from overlapping with the master packet at the beginning of the next frame when attempting to access the channel at the end of the frame.

In other words, when considering the distance between the farthest stations, the remaining frame time must be packet size+2a+α time as shown in FIG. The equation (2) in (2) is the limiting condition determined by this.

The 2a+β time of 1+ of ■■ is a value obtained from the following.

As shown in FIG. 3, consider a case where the end station of the cable pretends to be transmitting the m-th packet, and the end station on the opposite side is transmitting the m+1-th packet. The m+1st packet is 2a+ after detecting the mth packet.
The leading bit must be detected by all stations within α time.

If it is not detected, that is, if it remains silent even after 2a+α time, all stations may recognize that the m+1th packet has ended.

As a result, the m+2-th packet becomes the m+1-th packet earlier in order. Similarly, the order of subsequent packets is advanced one by one.

Also, β>α and 2a+β>2a+α, which is 2a+
This indicates that if it is silent during β time, it is also silent during 2a+α time.

Note that the packet size in this method is 2a + β - α
Must be above.

This regulation of the effective packet size is a limiting condition to ensure that the transmission timing of the next frame of a packet being transmitted in a frame is not delayed by one frame time or more from the current frame. For example, when the m-th packet ends as shown in FIG. 5, the m+1-th packet is sent out after confirming silence for 2a+β time. In the previous frame, from m to 1st packet (mth packet size)
It was sent after +α time.

From this, in order to prevent the sending timing of the (m+1)th packet from being delayed, the following equation needs to hold true.

Packet size +α≧2a+β→ Packet size≧2a+β−α [Problem to be solved by the invention] The CSMA/CD system is currently a stack of packet-switched communication systems in local area networks. Therefore, communication systems employing the CSMA/CD method are widely used. However, the CSMA/CD system has difficulty in real-time transmission and is considered disadvantageous for audio/image data transmission. On the other hand, Faxnet is a circuit-switched communication system that is advantageous for real-time transmission. Any type of data transmission is possible when both methods are used together. Additionally, reliability and communication quality can be provided depending on the type of data. Therefore, an attempt was made to mix and communicate both systems using a single cable, but since the transmission methods were different, packet collisions occurred as shown in FIG. 30, and real-time transmission could not be realized. Therefore, in order to use the CSMA/CD type packet-switched communication system and the Faxnet circuit-switched communication system together, two separate cables are required.

Therefore, an object of the present invention is to improve the conventional FAXnet transmission system and realize a line-switched communication system that can coexist with a C3MΔ/CD type packet-switched communication system on a single cable.

[Means and operations for solving the problems] In order to achieve the above object, the digital signal transmission system of the present invention prevents the digital signal of the own station from colliding with the digital signal transmitted within the station in accordance with the CSMA/CD system. It is equipped with a transmission logic circuit that has a function that does not stop the transmission of the signal, and a reception logic circuit that has a function that does not stop the reception of the digital signal even if a collision is detected midway through the reception of the digital signal. A transmission timeout detection circuit that generates a transmission request signal that requests unconditional transmission if the signal cannot be sent for one frame after the signal is sent, and a transmission timeout detection circuit that generates a transmission request signal that requests transmission unconditionally, and if the signal addressed to the own station cannot be received for one frame after receiving the digital signal addressed to the own station. and a reception timeout detection circuit that generates a reception request signal that requests the start of reception unconditionally.

〔Example〕

FIG. 1 shows a configuration diagram of an embodiment of the present invention, and the differences from the conventional F axnet configuration are shown below.

In the figure, 52' and 55' are transmission logic circuits, respectively;
The reception logic circuit has the following functions in addition to those of the conventional transmission logic circuit 52 and reception logic circuit 55.

Transmission logic circuit 52' ■ Even if CDot= is detected within a time period of 2a+τI+τ2 from the start of packet transmission, the transmission is not stopped.

■ When a transmission request signal is received from the transmission timeout detection circuit, C3o is sent again even if the packet order is not the sending order of the local station.
The packet is sent even if x is detected.

Reception logic circuit 55 ■ Once packet reception starts, even if CDo++
Even if it detects this, it does not stop receiving the packet.

■ Performs a receiving operation when receiving a receiving request signal from the receiving timeout detection circuit. This is usually done from the time the CSO is reached.

The receiving operation here means detecting the rear end of the preamble, reading the next destination address, and taking in the packet if it is addressed to the local station. Also, CDo- means Co11isi
Abbreviation for on Detect on, collision detection circuit 6
2 means that a collision detection signal is output from CS,
, is an abbreviation for Can1er 5ense on, meaning that there is a signal on the cable. Furthermore, a
is the propagation delay time of the farthest station, τ1 is the time required for a CSMA/CD type station to detect a collision, and τ2 is the CSMA/CD type station.
/' indicates the jam signal time that the CD type station sends after a collision, and the time 2a+τ1 +τ2 is hereinafter referred to as the priority window.

Further, the communication system of the present invention is provided with a transmission timeout detection circuit 70 and a reception timeout detection circuit 71, and these detection circuits have the following functions.

Transmission timeout detection circuit 70 (2) Frame control circuit Counts empty frame locks and is reset to θ by a transmission request signal from the transmission control circuit.

(2) If more than one frame has elapsed from the time the transmission request signal comes until the next one, that is, if more than one frame has been counted, a transmission request is sent to the transmission logic circuit.

■ Repeat the above operations until there are no more packets to send.

Reception timeout detection circuit 71 ■ Counts frame locks from the frame control circuit, and is reset to θ by a packet detection signal directed to the local station from the reception logic circuit.

(2) When more than (1 frame - τ.) has elapsed since the arrival of the packet detection signal addressed to the local station until the next arrival, a reception request signal is sent to the reception logic circuit. Here, τ0 is the length of time required to determine the destination address after receiving the head of the packet.

■ Repeat the above operations until you no longer receive packets addressed to your station.

In this way, in the communication system of the present invention, Faxnet
Conventional Faxnet type stations are provided with a priority function in order to transmit packets with priority over C3MΔ/CD type packets. On the other hand, the C3M/CD type station is configured to stop packet transmission after τ1 + 2 hours if a collision occurs. In order to prevent fields after the preamble of the Faxnet packet of the present invention from being destroyed due to this collision, the length of the preamble must be
a+τI +τ2+τ3 (τ3: the minimum preamble length required to stabilize the receiving circuit and establish bit synchronization). (Figure 2) As a result, F of the present invention
The preamble of the axnet packet format is longer than that of the conventional one (Figure 3).

Next, FIGS. 4 to 6 show how Faxnet packets in the present invention are transmitted with priority over CSMA/CD type packets. In the figure, a station that sends out Faxnet type packets is called MST station, A station, B station, and a station that sends out CSMA/CD type packets is called C station.

In frame F (n) in FIG. 4, station C successfully sends a packet to the free area. In F(n+1) frame, C
The station sends a packet next to the MST station's packet and collides with the A station's packet. Station A continues to transmit and is successful. Station C stops transmission and retransmits again.

In frame F (n) in FIG. 5, station C sends a packet immediately before the MST station packet is sent.

Since the MST station sends out each frame regardless of C3o=10F-, a collision occurs in the middle of the packet sent by station 0. The MST station continues to transmit and is successful. Station C stops sending and retransmits. Assume that in the F(n+1) frame of FIG. 9, when station A finishes transmitting and the area occupied by station A becomes vacant, a 0-station packet is sent there. 0
There is no problem if the station packet is short, but if it is long, a collision occurs in the middle of the packet sent by station 0 because station B sends out each frame regardless of the packet order and C3oN10FF. Station B continues to transmit and is successful. Station C stops sending and retransmits. At that point, the packet count of station B is 2, so F(n+1)
In the frame, station B transmits at the 2nd and 4th positions.

The transmission control method according to this method will be described in detail below.

(2) FIG. 7 shows the transmission packet of the station that starts transmission first when the transmission path is silent.

In this case, the packet is sent out using the following procedure.

■ Carrier sense is monitored for one frame, and if there is no carrier sense output, packets are sent out at a cycle according to the own station's frame clock. At this time, a flag indicating that it is a master packet is set in the control field of the packet to be sent. Alternatively, an area indicating the packet order may be provided in the control field, and the fact that the packet is #1 may be displayed there.

■ Send the first packet and if there is no collision with another station's packet outside the priority window, send the packet according to the frame clock.

■ When the first packet collides outside the priority window, perform backoff and monitor the carrier balance between one frame immediately before sending the packet, and if there is a carrier or not, follow the respective procedures. Resend the packet.

■ If there is a signal on the transmission path, follow the steps below. (Refer to Figure 28) II-1. When a master packet is detected ■ Detect a master packet, count the number of packets after the master packet including the master packet using a packet counter, and calculate the number of packets in the next frame. When a master window is detected, it is stored in the packet counting memory. (The master window will be described later.) ■ Detect an empty area and determine whether the following conditions are satisfied.

Empty area ≧ (packet size + 2a + α) (Empty area detection will be explained later) ■ If the judgment in ■ is OK, count the packets from the master packet in the next frame and store them in the previous frame. After receiving several packets, check the carrier of the packet.
Sense-off is detected, and a packet is sent after α time from that point.

For example, in the case of FIG. 2, packet #3 of every frame is detected, and the packet is sent out a time α after the carrier sense of the packet is turned off. (Own station - 0 station) ■ If a packet is sent and there is a collision outside the priority window, hack off and try again.

■ When your own station sends a packet, if there is a station that has finished transmitting, follow the procedure in ■.

■-2, If the master packet is not detected■ Monitor for one more frame, and if there is no carrier sense output, follow the I procedure, and if the master packet is detected, follow the Tl
If you want to detect the master packet according to the procedure in -1, follow the procedure below.

The reason why one additional frame is monitored here is that if the master station is monitored for only one frame when the master station transitions, the master packet may not be detected.

■ If there is no carrier sense output, packets are sent out at a frequency according to the own station's frame clock. At this time, a flag indicating that it is a master packet is set in the control field of the packet to be sent. Alternatively, an area indicating the packet order may be provided in the control field, and the fact that the packet is #1 may be displayed there.

■ Send the first packet and if there is no collision with another station's packet outside the priority window, send the packet according to the frame clock.

■ When the first packet collides outside the priority window, perform back-off and monitor the carrier sense for one frame immediately before sending the packet, and if there is a carrier or not, follow the respective procedures. Resend the packet.

(2) When there is a station that has finished transmitting a packet (see FIG. 9) (2) Each station has a frame tally and a tl timer that starts when the carrier sense of each packet is turned off. This timer is reset by carrier sense on and frame clock, and if it is not reset until time t8, it will automatically restart after tl timeout.

t, = 2a+β (β〉α) ■ Each station determines that the total value of the number of packets from the master packet and the number of times the tl timer times out is the value immediately before the packet number that the station should send. If a packet is detected, the packet is sent out a time α after detecting carrier sense on.

However, if a transmission request signal is received from the transmission timeout detection circuit, the packet is immediately transmitted. If there was no packet immediately before, the tl timer times out, so
When it is detected, the packet is sent after α time.

■ The packet counter of each transmitting station counts the number of packets, including the master packet, from the master packet every frame to the time when the station transmits the packet. For example, in the F(n) frame shown in FIG. 9, the packet counter value of the B station is 2.0 and the packet counter value of the B station is 3.

If a station finishes transmitting between the master packet and its own station's turn to send packets, the value of the own station's packet counter will be less than the value in the previous frame by that amount. .

Therefore, in the next frame, packets are sent out in accordance with this updated packet sending order.

Note that the packet counter of stations that are not transmitting is the number of packets between master packets (including master packets).
are being counted.

■When the master packet is lost (see Figure 10) ■Each station monitors the master packet, and if the master packet cannot be detected at the position where it should appear in that frame, the own station will The presence of a master packet is monitored until the timing of sending out a packet, and if there is no master packet, the master flag of the sent packet of the own station is set when the own station's turn to transmit comes or a transmission request is received from the transmission timeout detection circuit. and send it.

■ In the next frame, the station that became the master starts mastering at the #1 packet position according to its own frame clock.
Send the packet even if it is C3olI.

■Detection and checking of the master packet (see Figure 11) ■When a master packet is detected, the master window timer is set from a certain criterion of the master packet (for example, the end of the control field, the end of the packet, etc.). Let it start.

■ Generate a master window at the position where the master packet should appear in the next frame. The window width of this master window is determined so that the master flag can detect the master packet position even in the worst case even when the master packet position changes when the master window transitions.

■ Each station checks whether there is a master flag in this master window.

■ When a master flag is detected within the master window, it is treated as a master packet, and the master window timer is reset and restarted at a certain reference point determined by ■ of the master packet. (A in the figure) ■ If the master flag is not detected within the master window, it is determined that there is no master packet, and in that case, the periodic timer is restarted from the end of the master window. (Exit in the figure) Also, the end of the master window is ■
It is necessary to match a certain reference point determined by.

■, Detection of empty area (see Figure 12) ■ Each station detects a master packet, and when the carrier sense of this master packet turns off, starts the empty area detection counter (a in the figure) ■ Next If a packet is detected before the master window of the frame is detected, the counter is stopped when the carrier sense is turned on (b in the figure), and the counter is reset and restarted when the carrier sense is turned off. .

■ When the master window is detected, stop the counter and read the counter value at that time (C in the figure).

This value becomes the empty area.

It is also possible to set the area where Faxnet packets can be sent within a predetermined period from the beginning of the frame, and set the remaining area up to the rear end of the frame as an area where only C3MΔ/CD type packets can be sent. This further adds a frame counter which counts the frame clock h output from the frame control circuit 80 and is reset to θ by the master detection signal from the reception logic circuit 55'. The above function is accomplished by sending out a transmission request prohibition signal to the transmission control circuit at a predetermined frame count value and canceling the signal each time the frame counter is reset to θ. By doing this, when packets become crowded, CSMA/CD
CSMA/C type packets have a higher probability of successful transmission.
It becomes possible to shorten the time (transmission delay time) until a D-type packet is successfully transmitted. (FIG. 13) If the master station is C3opp from a predetermined frame timing to the rear end of the frame, it is also possible to send out a dummy preamble to the rear end of the frame. Here, the predetermined frame timing is determined as follows.

Frame timing - 1 frame length - C5MA/CD type packet maximum length This allows CSMA from the above frame timing to the rear end (that is, the beginning of the next frame).
Even if a /CD type packet is sent out, it will immediately collide and the packet will immediately stop being sent. (Figure 14) Therefore, CSM
1a by master packet of A/CD type communication system
tecollision will disappear. 1ate co
llision is when a CSMA/CD type packet passes the collision window (a period during which a collision could occur if only CSMA/CD type packets were being transmitted on the cable) and a collision occurs. In the present invention, this 1
ate collision may occur when a master packet is sent and after a transmission end packet is dropped. However, the possibility of the former occurring is much greater than the latter. 1ate collision is C
This does not occur only in SMA/CD type communication systems. From this, CSMA/CD type communication system and Faxnet
1ate collision when mixing
It is better to prevent this from occurring as much as possible. If the master station sends out a dummy preamble as described above, 1
The probability of occurrence of ate collision becomes extremely low. The above can be easily realized by adding the above-mentioned frame counter and having the master station send a dummy preamble transmission request to the transmission control circuit with a predetermined frame count value.

〔Effect of the invention〕

According to the digital transmission system of the present invention, the preamble length of the Faxnet type packet is made long, so that even if this packet collides with a CSMA/CD type packet, the Faxnet type packet will not be transmitted.
A transmitting station for et type packets does not stop transmitting packets. Also, the transmitting station always transmits one packet for each frame that does not complete transmission.

As a result, in the present invention, Faxnet type packets can be transmitted with priority over CSMA/CD type packets, and real-time transmission is possible even on a single cable.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing the outline of the communication system of the present invention, Fig. 2 is an explanatory diagram showing the situation when a packet of the invention collides with a CSMA/CD type packet, and Fig. 3 is a conventional communication system. A comparison diagram of the packet format of the system and the present invention, Figures 4 to 6 show that the packet of the present invention is C3MΔ/
An explanatory diagram showing how priority is given to CD-type packets. Figures 7 to 12 are explanatory diagrams for explaining the principle of the digital transmission system of the present invention. Figure 13 is an illustration of another implementation of the present invention. FIG. 14 is an explanatory diagram of yet another embodiment of the present invention. FIG. 15 is a block diagram showing a schematic diagram of a conventional communication system. FIGS. An explanatory diagram for explaining the principle of digital transmission system, Figure 19 is a frame diagram.
Detailed circuit diagram of the control circuit, Figure 20 shows the timing diagram of the frame control circuit.
Chart; Figure 21 is a diagram showing the relationship on the time axis between packet signals on the transmission path and the frame clock of each station; Figure 22 is a detailed block diagram of the packet management circuit;
The figure is a timing chart of a packet management circuit, and FIGS. 24 to 29 are explanatory diagrams for explaining the principle of a conventional communication system. FIG. 30 is an explanatory diagram showing a collision when a conventional communication system and a CSMA/CD type communication system coexist on one cable. 1 coaxial cable, 2 coax cable, 3 T connector, 4 user equipment, 41 transmitter, 42 receiver, 43 terminal controller, 51 transmission buffer memory, 52'...transmission logic circuit, 53...・・Transmission buffer amplifier, 5
4...Receive buffer amplifier, 55'...Receive logic circuit, 56...Receive buffer memory, 6
1...Transmission control circuit, 62...Collision detection circuit, 70...Transmission timeout circuit, 71...
Reception timeout circuit, 72...transmission logic circuit,
80... Frame control circuit, 81...
... t1 timer circuit, 82 ... Packet management circuit, 90 ... Bit counter, 91.105
... Flip-flop, 100 ... Packet order memory, 101 ... Packet counter,
102... tl timer timeout output counter, 103... Comparator, 104... Adder, 106... Ant gate.

Claims (4)

[Claims] (1) Digital signals transmitted on a communication cable are fixedly positioned within a frame that is periodically repeated on the time axis, and each station connected to this communication cable via a tap A multi-station communication network in which one or more stations that are actually transmitting signals occupy an area on the time axis that is further divided within the frame, and that transmits time-division multiplexed digital signals. A digital system in which the order of the areas occupied by each station is determined for each frame so that the area in use is continuous with one group within one frame, and signals are transmitted in the order of each area. In the signal transmission system, the station is equipped with a transmission logic circuit that has a function that does not stop the transmission of the digital signal even if it collides with the digital signal sent out according to the CSMA/CD system, and the reception of the digital signal is started. 1. A digital signal transmission system comprising: a reception logic circuit having a function of not stopping reception of the digital signal even if a collision is detected during the transmission. (2) A transmission timeout detection circuit that generates a transmission request signal that requests unconditional transmission if the signal cannot be transmitted for one frame after sending a digital signal, and a signal addressed to the own station for one frame after receiving the digital signal addressed to the own station. The digital signal transmission system according to claim (1), further comprising a reception timeout detection circuit that generates a reception request signal that requests to start reception unconditionally if the signal cannot be received. . (3) By prohibiting transmission requests from stations other than the master station from a predetermined timing to the end of the frame, digital signals can be sent only to the area from the beginning of the frame to the predetermined timing mentioned above. A digital signal transmission system according to claim (1). (4) The master station designated as the reference station for frame synchronization from the timing determined by the maximum digital signal length allowed by the CSMA/CD system to the end of the frame is the CS (Carrier Sensing) station.
e) The digital signal transmission system according to claim 1, characterized in that a dummy digital signal is sent from the time when off is detected until the end of the frame.