JP3089315B2 - Interrupt-reset type communication system using timer - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention is capable of sharing a transmission path among a plurality of communication nodes and performing multimedia communication of data, voice, image, etc. Area Network), MAN (Metropol)
An interrupt reset type communication method using a timer that can be used as an itan area network) or a WAN (wide area network). In particular, a timer interrupt capable of guaranteeing communication that satisfies an allowable delay time from high priority information. The present invention relates to a reset communication system.

[Conventional technology]

As a LAN communication technology, for example, as described in “Nikkei Electronics, 4-22 (No. 367), 1985, pp. 185 to 238”, a CSMA / CD system and a token system are generally known. ing. The CSMA / CD system is a system in which data is transmitted after confirming that a transmission path is unused, and when data transmitted by another node collides with the data, this is detected and retransmitted. Further, the token system is a system for controlling the network by circulating one data of a token representing a transmission right on the network, and in this system, only the node receiving the token can transmit the data, and the transmission ends. Then the talk is passed to the next node.

 These methods have the following problems.

(A) When there is a lot of traffic, the transmission efficiency does not increase.

(B) It is difficult to perform high quality voice / image communication in terms of delay time.

On the other hand, in order to improve transmission efficiency and delay characteristics, a slotted ring system in which a transmission path is divided into fixed-length cells and accelerator control is performed on each cell is being studied.

For example, as described in Japanese Patent Application No. 63-50310, as an example of the slotted ring system, a cyclic monitoring packet switching system has been proposed.

In this method, when information to be transmitted has a priority, a communication section common to all nodes is provided according to the priority, and when a communication section of information with a higher priority ends, information of a lower priority is sequentially assigned to information. Move to the communication section. In addition, the time between the communication sections of the same rank is measured by a timer, and when this time exceeds the allowable time, the communication section of the higher rank interrupts the communication section of the lower rank. This guarantees communication of higher priority information.

[Problems to be solved by the invention]

In the above prior art, some low-priority information
To ensure that high-priority information is not interrupted,
A common communication section is provided for all nodes, and only one type of priority information can be communicated within a specific communication section, thereby realizing priority control in units of the communication section. That is, a method in which information arriving discretely is accumulated in a node, and when a corresponding communication section starts, is swept at a stroke.

The upper limit of the information that can be transmitted within the communication section is determined in advance by the resource management control as the amount of information that can be transmitted between the start of a certain communication section and the start of the next same-rank communication section. .

Therefore, once the corresponding communication section ends once, the information with high priority cannot be transmitted until the next communication section starts even if the amount of transmitted information has not reached the upper limit. In this case, even if transmission is permitted in terms of resource management, transmission is actually prohibited, which leads to an increase in delay time and a reduction in the use efficiency of the transmission path.

An object of the present invention is to provide an interrupt reset type communication system using a timer which can solve such a problem and can guarantee communication satisfying an allowable delay time from high priority information.

[Means for solving the problem]

In order to achieve the above object, an interrupt reset type communication method using a timer according to the present invention, a node connected to a transmission line includes an upper limit of the number of usable cells and a counter for recording the number of already used cells, When a cell composed of control information and user information is circulated on a transmission path and information is transferred between arbitrary nodes, all nodes have used cells up to the upper limit, or there is no information to be transmitted. The node that has detected any of the states indicates a reset or sends a reset cell to the transmission line, and at the same time, transitions to a pruning waiting state for cutting the reset indication of the arriving reset cell. The node that has transitioned to the normal state due to mowing of the reset cell receives the reset cell. There, if the priority by transmission wait permissible delay time from the highest rank 1 to the information to be transmitted to the lowest priority N (N ≧ 2) is present, in the node,
N counters corresponding to priorities and their upper limit values, timers and their upper limit values, and N +
A communication state of 1 level is set, and a reset display of N + 1 level is provided in the cell, and the level i of the communication state is 1
≦ i ≦ N, priority 1 to priority i
If i = N + 1, it is possible to transmit information from priority 1 to priority N, and when reset cells of different levels occur simultaneously, a node that has transmitted a low-level reset cell When a high-level reset cell is received, the node transitions to a high-level normal state without cutting the high-level reset cell, and the node that has transmitted the high-level reset cell receives the low-level reset cell. Then, the low-level reset cells are reaped, and only when a high-level reset cell is received, a transition is made to the high-level normal state.

The node does not reset the counter and the timer when transmitting a reset cell of level 1, and transmits a reset cell of level j (2 ≦ j ≦ N + 1) from 1 to j−1. And resets the counter and timer to the communication state of the level j. At the time of information transmission, the transmission wait allowable delay time is compared with the upper limit value of the counter, and the upper limit value is set within the transmission wait allowable delay time. If it is determined that the information transmission of the priority cannot be realized, a reset cell of the level determined not to be realized is generated. Further, with the counter of the priority j reaching the upper limit value, the transmission waiting information of the priority becomes When remaining, a reset cell at level j + 1 is generated to reset the counters from 1 to j.

[Action]

In the present invention, in order to handle information having different priorities, a communication state of (priority + 1) level is provided, and when transmission of high-priority information is prevented by low-priority information, low-priority communication is performed. Shifts to the communication section where the information cannot be transmitted. If the next communication section does not start even after the elapse of a certain time, the processing shifts to a new communication section.

As a result, communication that satisfies the permissible delay time based on the high priority information is performed.

〔Example〕

 Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

FIG. 2 is a configuration diagram of a node in one embodiment of the present invention.

In FIG. 2, 1 is a cell sent from an adjacent node,
A transmission path for transmitting a cell to another adjacent node; 2 a node; 3 a cell receiving unit for receiving the transmitted cell;
4 is a cell transmission unit for transmitting cells, 5 is a management unit for controlling headers for transmission and reception cells, 6 is a reception buffer for temporarily storing received cells, 7 is a transmission buffer for temporarily storing transmission cells, and 8 is a terminal. And a terminal corresponding unit 9 for controlling an interface with the terminal. The transmission buffer 7 has a plurality of transmission queues 71 and 72 for priority, each of which is divided into a connection queue and a connectionless queue. Note that the connection type is for transmitting information after performing a transmission procedure according to a protocol in advance, and the connectionless type is for transmitting information immediately without a prior transmission procedure.

In this embodiment, when a cell in a cell frame arrives via the transmission line 1, the cell receiving unit 3 receives the cell, and when the destination of the cell is its own node address, information in the cell is received by a reception buffer. 6 and temporarily stored. The information stored in the reception buffer 6 in this way is read by the terminal correspondence unit 8 and sent to the corresponding terminal 9 as user information.

Further, the cell receiving unit 3 transfers the received control information (in the cell header) of the cell to the management unit 5.

In this way, the management unit 5 checks the occupancy information of the arriving cell, and stores information to be transmitted from the transmission buffer 7 in the user information field of the cell when the vacancy information is empty, and then closes the occupancy bit to transmit. Set the original to the own node address,
Set the destination to the receiving node address. Further, the cell is transmitted to the transmission line 1 via the cell transmission unit 4.

At this time, both the connection method and the connectionless method are simultaneously supported as the packet exchange type. That is, if the transmission procedure is required before transmission, data of the connection method is transmitted. If the transmission procedure is not required and the transmission can be performed directly, the data of the connectionless method is transmitted.

FIG. 3 is an explanatory diagram showing a structure of a cell on a transmission line in one embodiment of the present invention.

In FIG. 3, OH indicates a transmission overhead (portion for synchronization or the like), and F indicates a portion such as framing for identifying the head of a cell.

In this embodiment, the transmission frame of the layer 1 (125 μs)
Contains information and transmission overhead, and a cell frame corresponds to this information portion.

The cell frame is composed of a portion F such as framing (for example, a flag composed of codes such as 010101...) And a plurality of cells. Further, one cell includes a cell header, user information, and a cell tailor.
Note that the cell tailor may not be provided.

 The cell header includes the following information.

(1) Cell occupancy display (indicating whether valid information is included in the user information) (2) Source address (address of the node transmitting the cell, only when the occupancy bit is occupied) (3) Reception source address (This is a destination address for receiving a cell, and has a meaning only when the idle / occupied bit is occupied.) (4) Access control information area (Used for detecting the end of transmission of another node) (5) Reset bit (indicating that it is a reset cell indicating transition to the next communication section) FIG. 4 is an explanatory diagram of a communication network configuration to which the interrupt reset communication system using a timer according to the present invention is applied. is there.

In FIG. 4, (a) is a ring network in which the node 2 is actively connected to the transmission line 1, and (b) to (e) are bus networks in which the node 2 is passively connected to the transmission line 1. .

Further, (b) shows a multiple head-end multiple access bus network in which a plurality of head ends 10 exist in the network and a plurality of cell receiving units and cell transmitting units exist in each node.

(C) is a multiple head-end single access bus network, in which a plurality of head-ends 10 exist in the network and each node has a single cell receiving unit and a single cell transmitting unit.

(D) shows a single head-end multiple access bus network in which a single head-end 10 exists in the network and a plurality of cell receiving units and cell transmitting units exist in each node.

(E) is a single head end single access type bus network, in which a single head end 10 exists in the network and a single cell receiving unit and a single cell transmitting unit exist in each node. Show.

Since these bus networks (b) to (e) are passively connected, a head end 10 for generating a transmission frame is required, and the head end 10 sends a received cell to another transmission path. The cells need to be cycled. The function of the head end 10 may be performed by the nearest node 2.

Also, each node has a different address,
As shown in (b) and (d), when one node has a plurality of cell receiving units and cell transmitting units, different addresses are used.

Further, the information in the form shown in FIG. In that case, a cell frame is used to detect the head of the cell used by each node. That is, since the head of the cell frame is provided with a flag for identifying the head, by detecting this, the head of the cell group used by a certain node can be known. Also, one node can transmit information by using a predetermined number of cells in a communication section.

Also, the communication format of the present invention does not necessarily have to be a ring format. However, for example, in the case of using the format shown in (b), the nodes at both ends separate cells received on one transmission path from each other. It is necessary to make a round by transmitting to a transmission line of the same.

In addition, by configuring an active circuit type as shown in (a), a cell can be terminated at a plurality of nodes.
By configuring a passive circuit type as shown in (b), one node may transmit a cell frame.

The function of the head end 10 is necessary only in the case of a passive circuit. This is because in an active circuit, a node creates a signal and sends out a cell separately from the input signal,
In a passive circuit, an input signal is interrupted as it is, so that a function of first creating a cell and transmitting it to a transmission line is required.

FIG. 5 is an explanatory diagram showing a state transition of a node in one embodiment of the present invention.

In this embodiment, the plurality of nodes A, B, C,... Have different addresses, and are actively or passively connected to the transmission path.

In addition, each node has an upper limit of the number of usable cells determined from the throughput and delay conditions required for communication,
A counter for recording the number of cells already used is provided.

This means that if any node detects that it has used cells up to its upper limit or that it has no more information to send, it resets the counters for all nodes, so that node displays a reset. At the same time as transmitting the reset cell, a transition is made to a state in which the reset display of the next arriving reset cell is reaped (reap wait state). Next, the node changes to the normal state by cutting the reset display.

The timing at which the node that has transmitted the reset cell resets its own counter may be either when transmitting the reset cell or when the reset cell is reaped.
In the case of, the counter is reset at the same time as transitioning to the reaping waiting state, so that information can be transmitted even in the reaping waiting state, whereas in the case of, the counter is not reset in the reaping waiting state, so information is transmitted only in the normal state Becomes possible.

When the node A of the present embodiment determines that all nodes have stopped transmitting information using the control information on the cell, the node A resets the counter, sends a reset cell, and transitions to a pruning wait state. I do. In addition, when the reset cell is transmitted via the transmission line, the other nodes B and C reset the counter and transit to a new normal state. Similarly, after confirming that all the nodes are in the transmission stop state, the node C resets the counter and sends out a reset cell, and the other nodes A and B relay this reset cell. Further, the node C in the pruning waiting state reaps the reset cell that has circulated. Hereinafter, similarly, each node repeats the process of transmitting, relaying, and pruning reset cells.

 Next, the state of the node will be described.

Each node has a state indicating whether or not information can be transmitted, and a state indicating whether to cut off when a reset cell is received. 10 (a) to 10 (c) show the relationship between them. (A) is an example in which the states are tightly coupled. A node that does not generate a reset cell transitions between a transmission state and a transmission stop state, and a node that generates a reset cell goes through a reaping wait state. In this example, transmission of information is prohibited in the reaping waiting state. (B) is an example in which the states are loosely coupled. Nodes that do not generate reset cells transition between the transmission state and the transmission stop state, and remain in the normal state. The node that generates the reset cell also transits between the normal state and the pruning state, and resets the counter simultaneously with the generation of the reset cell, thereby transiting from the transmission stop state to the transmission state. (C) This is an example in which the original state is loosely coupled. The node that generates the reset cell is
By resetting the counter at the same time as the pruning of the reset cell, the state transits from the transmission stop state to the transmission state.
However, the difference from (a) is that there is a possibility that a transition to the transmission state occurs when information to be transmitted arrives even in the reaping waiting state.

Next, a procedure for extending the communication state of each node and the reset display of each cell to the N + 1 level and transitioning to the communication state of level i by the circulation of the reset cell of level i will be described.

FIG. 6 is an explanatory diagram showing a state transition when all nodes are reset by a single reset cell in one embodiment of the present invention. The numbers in parentheses indicate the reset level of the reset cell or the communication level of the node.

In the present embodiment, when all nodes are reset by a single reset cell, the node that has been triggered to transition to a new communication state generates a reset cell indicating the next communication level and generates a new level. Transitions to the state of waiting for reaping.

For example, the node A sends out a reset cell of the communication level (1), cuts the reset cell that has circulated from the reaping waiting state (1), and returns to the normal state (1). Also, the other nodes B and C relay the reset cell and return to the normal state (1). Similarly, each node sends out a reset cell, and the node that relayed the reset cell transitions to the communication level.

FIG. 1 is an explanatory diagram showing a state transition when all nodes are reset by a plurality of reset cells having different levels in one embodiment of the present invention. The numbers in parentheses indicate the reset level of the reset cell or the communication level of the node.

In this embodiment, when reset cells having different levels occur at the same time, the node that has generated the low-level reset cell transitions to a low-level reaping wait state. However, when a high-level reset cell is received, it is reaped. Transition to a high-level normal state.

For example, the node A generates a reset cell of level (3) and transits to the reaping waiting state (3). Upon receiving the reset cell of level (1), the node A immediately transits to the high level normal state (1). I do.

Also, the node that generated the high-level reset cell
When a low level reset cell is received, it is reaped, but remains in a high level pruning wait state. Furthermore, when the same high level reset cell is received, it is reaped and transits to the normal state.

For example, after generating a reset cell of level (1), if the node C receives a reset cell of level (3) in the pruning wait state (1), the node C reaps the reset cell, and the pruning wait state (1) changes. do not do. Thereafter, when the reset cell of the level (1) is received, the reset cell is pruned, and the state transits to the normal state (1).

As for the change of the level relating to the information transmission, as described with reference to FIG. 10, (a) an example in which the level is tightly coupled to the pruning waiting state, (b) an example in which the communication level is changed simultaneously with the occurrence of the reset cell, c) An example in which the communication level is changed simultaneously with the pruning of the reset cells is considered.

Next, with reference to Table (a), the relationship between the communication level of the node and the priority of the information that can be transmitted when the information to be transmitted has the highest priority from the highest priority 1 to the lowest priority N will be described.

In this embodiment, if the communication level j is in the range of 1 to N, the information of the priority levels 1 to j can be transmitted. If the communication level j is N + 1, the information of the priority levels 1 to N can be transmitted. Transmission is possible. In a communication level at which transmission of information of a plurality of priorities is possible, information of a high priority is transmitted first.

Next, with reference to Table (b), the relationship between the reset cell level and the counter and timer to be reset when the information to be transmitted has the highest order of priority from the highest order N to the lowest order N will be described.

In this embodiment, if the reset level is r, r is 2
To r + 1, the counters and timers corresponding to the priorities 1 to r-1 are reset, and when r is 1, none of them is reset.

In addition, the timer corresponding to each priority is reset independently of the reset associated with the reception of the reset cell when the information of the corresponding priority is lost.

 Next, a method of determining the upper limit value of the counter will be described.

In this embodiment, as shown by the following equations (i) and (ii),
The number of cells required to satisfy the throughput required by each communication within the allowable delay time is calculated, and the sum of the required cells at the same priority is set as the upper limit of the priority of the node.

(I) Required number of cells = Requested throughput × Permissible delay time /
Number of cells (ii) Upper limit value of counter = Σnumber of required cells Next, the relationship between the allowable delay time and the upper limit value of the counter will be described.

FIG. 7 is an explanatory diagram showing the relationship between the allowable delay time and the upper limit value of the counter in one embodiment of the present invention.

In FIG. 7, TMX _i is the allowable delay time of priority i,
CMX _i is the upper limit of the counter.

In this embodiment, from the highest priority, so as to transmit the information of CMX _i in time of TMX _i, and controls the priority.

As a result, while each node is performing communication at a throughput sufficiently smaller than the required throughput, a reset cell of level N + 1 is generated at a shorter time interval than TMX _{1 having} the strictest allowable delay time. That is, all cells to be transmitted can be transmitted.

Further, when the throughput of each priority increases, the following processing is performed.

That is, when an information transmission request of priority i occurs in a state where the counter of priority i (CTR _i ) = the upper limit value of CTR _i (CMX _i ), a reset cell of level i + 1 is generated,
Reset the counter. This is, CT within TMX _i hours
R _i is not reset, in order to deal with the situation can not information transmission of the next cycle.

Also, the possibility of completing the information transmission of CMX _i within TMX _i time is determined. This is to cope with a situation in which information transmission corresponding to CMX _i is not completed within TMX _i hours.

Next, a method for judging the occurrence of these situations will be described with reference to FIGS.

8 and 9 show one embodiment of the present invention.
FIG. 9 is an explanatory diagram illustrating processing when the information transmission amount has reached an upper limit value of a counter.

In this embodiment, when each node does not know the communication status of the other node, it is determined that a situation has occurred by the following equation (iii).

_{(Iii) (TMX i -TMR i} ) / TMX i <Min (CM X i -CTR i, QUE i) / (CMX i + B i) where, Min (A, B) is the argument A, in the B, small Is a variable that returns the value of

 The meaning of the equation (iii) is shown in FIG.

In FIG. 8, QUE _i is the queue length of priority i (varies by node), CTR _i is the counter of priority i (varies by node), and CMX _i is the upper limit of the counter (CTR _i ) (varies by node) ), TMR _i is the timer of priority i (the difference between nodes is less than one round of the transmission line), and TMX _i is the allowable delay time of priority i (common to all nodes).

In this embodiment, the queue length of the priority order i, the counter,
Although the upper limit value of the counter differs from node to node, the allowable delay time is common to all nodes, and the timer is restarted by the circulation of the reset cell.

Here, all nodes have the information transmitted at the same pace, from node Assuming cell CMX _i in time of TMX _i are transmitted at equal intervals, the remaining time and the transmission future in TMX _i The amount of information that may be transmitted is compared, and the possibility of transmission completion within TMX _i time is determined.

For example, at node j, the remaining time in TMX _i obtained from TMR _i is smaller than QUE _i , so it is determined that the possibility of completion of transmission within the time is low, and a level 1 reset cell is transmitted. .

In consideration of the case where the above assumption is lost, it estimates further less amount of information side, by providing a safety factor B _i, can be further enhanced reliability.

Further, in the present embodiment, when each node has grasped the communication state of another node, it is determined that a situation has occurred by the following equation (iv).

_{(Iv) TMX i -TMR i <} min (CMX i -CTR i, QUE i) + (E_CMX i -E_CTR i) + Σ (A _CMX j -A_CTR j) + α where, E_CMX _i, that during TMX _i Time E_CTR _i is the maximum value of the number of cells of priority i passing through the node due to transmission of the node including the node, and E_CTR _i is the priority value that has already passed through the node due to transmission of the node including the node during TMR _i time. number of cells in order i, A_CMX _j is the transmission of the node, including the node during TMX _i times, the maximum number of cells priority j passing through that node, A_CTR _j, the in TMR _i time The number of cells of the priority order j that have already passed through the node including the node, Σ indicates the sum of 1 to i−1 regarding the subscript j, and α indicates the round trip time of the transmission path in terms of the number of cells.

 The meaning of the equation (iv) is shown in FIG.

In this embodiment, the two counters E_CTR _i and A_CTR _j (j = 1 to i−1) are reset at the same time as CRT _i is reset.

Therefore, it is possible to determine from the equation (iv) whether or not the information transmission of the priority order i or higher is completed within the remaining time of TMX _i .

〔The invention's effect〕

According to the present invention, in the reset type communication system, a communication state of (priority level + 1) level is provided in order to handle information having different priorities, so that information of all priorities can be normally transmitted, and low-priority information can be transmitted. When transmission of high-priority information is hindered by the information, it is possible to shift to a communication section in which low-priority information cannot be transmitted and transmit high-priority information. If the next communication section does not start even after the lapse of a certain time, the information is transferred to a new communication section, so that information with a high priority can be transmitted.

Furthermore, in a state in which information of different priorities can be transmitted, it is possible for each node to independently determine whether or not a previously allocated high-priority information amount can be transmitted within an allowable delay time, and it is determined that transmission cannot be performed. In such a case, by transiting to a communication state in which low-priority information is not transmitted commonly to all nodes, high-priority information can be transmitted.

Therefore, it is possible to guarantee the communication satisfying the allowable delay time from the information of the higher priority, and to improve the use efficiency of the transmission path.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram showing a state transition when all nodes are reset by a plurality of reset cells having different levels in one embodiment of the present invention, FIG. 2 is a configuration diagram of a node in one embodiment of the present invention, FIG. 3 is an explanatory diagram showing a structure of a cell on a transmission line in one embodiment of the present invention, FIG. 4 is an explanatory diagram of a communication network form to which an interrupt reset type communication system using a timer of the present invention is applied, and FIG. FIG. 6 is an explanatory diagram showing a state transition of a node according to an embodiment of the present invention. FIG. 6 is an explanatory diagram showing a state transition when all nodes are reset by a single reset cell in the embodiment of the present invention. FIGS. 8 and 9 are explanatory diagrams showing the relationship between the allowable delay time and the upper limit value of the counter in one embodiment of the present invention. FIGS. 8 and 9 show the information transmission amount reaching the upper limit value of the counter in one embodiment of the present invention. If you do To illustration, FIG. 10 is a diagram showing the relationship between transmission and transmission halt state and the normal-mowing wait state of a node in an embodiment of the present invention. 1: transmission line, 2: node, 3: cell receiving unit, 4: cell transmitting unit, 5: managing unit, 6: receiving buffer, 7: transmitting buffer, 8: terminal corresponding unit, 9: terminal, 10: head end , 71,72: Priority transmission queue, O
H: transmission overhead, F: framing to identify the beginning of the cell, etc., CMX _{1 to} CMX ₃ (CMX _i ): upper limit of the counter,
TMX _{1 to} TMX ₃ (TMX _i ): allowable delay time of priority i, QUE ₁
(QUE _i ): queue length of priority i, CTR ₁ (CTR _i ): counter of priority i, TMR ₁ (TMR _i ): timer of priority i, E_CMX _i : pass through node during TMX _i time E_CTR _i : the number of cells of priority i that passed through the node during TMR _i time, A_CMX _j : the maximum number of cells of priority j that passed through the node during time TMX _i Value, A_CTR _j : number of cells of priority j that passed through the node during TMR _i time, Σ: sum of 1 to i-1 related to subscript j, α: round trip time of transmission path in cell number conversion.

Claims (2)

(57) [Claims] A node connected to a transmission line includes a counter for recording an upper limit of the number of usable cells and the number of already used cells, and a cell configured from control information and user information on the transmission line. When information is transferred between arbitrary nodes by circulating, a node that detects that all nodes are in a state of using a cell up to the upper limit or in a state where there is no information to be transmitted, At the same time as transmitting the reset cell indicating the reset to the transmission line, the state transits to a pruning wait state for pruning the reset display of the arriving reset cell, transits to the normal state by pruning the reset cell, and receives the reset cell. The resetting node resets the counter of its own node, and the information to be transmitted has the highest rank 1 to the lowest rank N (N ≧ N).
When there is a priority order based on the transmission delay allowable delay time up to 2), the node has N counters and upper limit values corresponding to the priority order, a timer and an upper limit value of the timer, and the corresponding priority order. N + 1 level communication status is set, and the cell is provided with an N + 1 level reset display. If the level i of the communication status is within the range of 1 ≦ i ≦ N, priority 1 to priority i If i = N + 1, it is possible to transmit information from priority 1 to priority N. If reset cells of different levels occur simultaneously, a node that has transmitted a low-level reset cell When a high-level reset cell is received, the node transitions to a high-level normal state without cutting the high-level reset cell, and transmits a high-level reset cell. When a low-level reset cell is received, the low-level reset cell is pruned, and only when the high-level reset cell is received, the transition to the high-level normal state is performed. Reset type communication method. 2. The node does not reset the counter and the timer when transmitting a reset cell of level 1, and transmits a reset cell of level j (2 ≦ j ≦ N + 1) from 1 to j. The counter and timer up to -1 are reset, and the state transits to the communication state of level j. At the time of information transmission, the transmission delay allowable delay time is compared with the upper limit value of the counter, and within the transmission wait allowable delay time, When it is determined that information transmission for the upper limit cannot be realized, a reset cell of a level determined to be impossible is generated, and
In a state where the counter of the priority j reaches the upper limit value and the transmission waiting information of the priority remains, the level j +
2. The communication system as claimed in claim 1, wherein one reset cell is generated to reset a counter from 1 to j.