KR100191706B1 - Atm control device for ring type atm node - Google Patents

Abstract

The present invention relates to a communication node that executes ATM communication for multimedia communication. In order to transmit the allocated communication cell, if the window size in the ring is W cell and the number of transmitting cells assigned to the communication node in this W cell is n _a , the transmitting cell is selected for at least two (Wn _a ) cells. When the number obtained by the transmission buffer stored internally, the number of transmission cells counting the number of transmission cells from the node in the W window size, and the number obtained by the number of transmission cell counters correspond to the number assigned to the communication node, transmission cells in the transmission buffer are transmitted to the transmission path. It was set as the structure containing the transmission control means.

By doing so, it is possible to transmit the transmission cell to the transmission path with a delay less than the delay determined for the entire system, and to prevent abandonment of the transmission cell in the communication node.

Description

Access Control Device of Ring ATM Node

1 is a detailed block diagram showing an apparatus for controlling access of a communication node according to the present invention.

2 shows the format of a cell of a user network interface (UN) as defined in ITU-T Recommendation I.361.

3a and 3b are explanatory diagrams of sliding window control.

4 is a diagram showing the worst case of the cell delivery pattern to the communication node of the sliding window control system according to the present invention.

5 is an explanatory diagram showing a case where a transmitting cell is inserted into a communication node when a cell is received in a transmission path in the worst case as shown in FIG.

FIG. 6 is an explanatory diagram showing a transmitting cell standby state in a communication node when a cell is received in a transmission path in the worst case as shown in FIG.

FIG. 7 is a diagram showing, in concrete numerical terms, a standby state of a transmitting cell and a cell on one transmission path in one node. FIG.

8 is an explanatory diagram of a jump window control.

9 is a diagram showing the worst case of transmission cell generation in jump window control.

FIG. 10 is an explanatory diagram showing a worst case of a cell arrival pattern to a communication node of a jump window control system. FIG.

FIG. 11 shows the worst case of the cell delivery pattern to the communication node in FIG.

FIG. 12 is a diagram for describing a time required for cell generation during jump window control. FIG.

FIG. 13 is a view for explaining a time required for cell generation during jump window control. FIG.

FIG. 14 is a diagram for describing a time required for cell generation during jump window control. FIG.

FIG. 15 is a diagram for explaining insertion of a transmission cell into a communication node when a cell is received in a transmission path in the worst case shown in FIG.

FIG. 16 is a diagram showing concrete states of transmission cell standby states between cells and nodes in a transmission path; FIG.

FIG. 17 is a diagram showing a communication sequence in a conventional ATMR. FIG.

18 illustrates the format of a cell used in a conventional ATMR.

19 is a detailed block diagram showing a conventional access control device for a communication node.

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a communication node that executes ATM communication for multimedia communication. In particular, the present invention relates to a communication node having a buffer for preventing cell abandonment and at the same time suppressing a delay time when a cell changes according to the situation of a communication network. It relates to a control device.

In an access control apparatus in a communication network in which several communication nodes are connected to each other through one or several transmission paths and perform packet communication between several terminals provided in these communication nodes, each slot having a fixed length ( A slot ring system for carrying out packet communication using a cell) is known. One example is an ATMR (ATM Ring) that realizes a high-speed LAN (Local Area Network).

In the ATMR protocol, specific cell header information is used as disclosed in Japanese Patent Application No. 50310/1988. A circular communication section (window size) common to all communication nodes is provided, and each communication node has a window size register indicating a limit value for the number of transmitting cells in the communication section (communication band), and every cell or transmission ( If there is a cell to be transmitted, the communication node is sent to the maximum cell indicated by the window size register of each communication section described above. As described above, in the ATMR protocol, a circulation communication section is provided and the number of cells to be transmitted in the communication section is limited for each communication node, so that voice and video communication can be executed within a short allowable delay time. Referring to FIG. 17, a communication sequence according to the ATMR protocol will be described.

The basic cyclic communication section is called a window. Each communication node transmits the contents of a cell to be transmitted within the setting window size time. When all nodes have completed transmission, one of the nodes sets the reset cell and transfers it to the transmission path. Each node is prepared for reset by checking that the reset node is assigned, and the next window recognizes that each node is ready for reset when the reset cell rotates the ring and returns to the initiating node. The cycle is executed and transmission begins.

FIG. 17 is a diagram showing a system configuration, where 1 is a communication node, 2 is a transmission path, and 3 is a terminal that is a communication device connected to the communication node. 18 shows a specific cell format used in this system, in which the header portion 30a is different from the standard specification. 19 is a block diagram showing details of each communication node, in which (11) is a destination address identification unit for identifying a destination address of a received communication cell, and (10) is the busy of the header 30a. The highest address identification / setting unit for checking and rewriting the address, and 12 denotes a cell transmission control unit for controlling the transmission of the transmitted cell or the transmission from the node. 16 denotes a receiving cell buffer, and 20 denotes a reset flag storage for storing therein data indicating that all other communication nodes are currently in the transmission stop state, and this reset flag storage is stored at the beginning of a window cycle. It is reset. In the figure, reference numeral 13a denotes a window controller, 14 denotes a transmission controller which controls transmission cell transmission to a transmission path, 15 denotes a transmission buffer, 17 denotes a number of transmission cells, 18 denotes a window size. The window size register 19, which sets therein the number of transmission cells allocated to each node, is a comparator. The number of transmission cells assigned to each node in the window size is the sum of the number of transmission cells equal to or less than the window size (less than the window size).

This operation will be described in detail.

All communication nodes 1 connected to the ring-shaped transmission path each have a window size register 18 indicating a limit value of the number of transmission cells in the communication section and a transmission cell number counter 17 for counting the number of transmission cells in the communication section. Include. When the communication cell is transmitted from the terminal 3 connected to the communication node 1, the communication node 1 stores the communication cell in the transmission buffer 15, and stores the communication cell by the FIFO device in the transmission path 2 Send it out. The communication cell sent to the transmission path 2 is received by the communication node to which the terminal which is the destination of the communication cell is connected, and is then transmitted to the terminal. As shown in Fig. 18, this communication cell has an address area for setting the address of the communication node to which the communication cell is transmitted, and the first time that the communication node which executes ATM communication sets the highest address for this communication node. And a cell header 30a having an address area and a user data area for setting user data therein. For ATM communication between communication nodes, control is performed as follows according to the highest address value set in the highest address area of the cell header described above. This communication cell is transmitted according to a timing (cell slot) specified between communication nodes.

When a communication cell with the reset data set in the highest address area of the communication cell is received when the communication node is in the transmission stop state, the reset flag 20 and the number of transmission cell counters 17 are reset and the next window cycle starts window control. do.

If the value displayed by the number of transmission cells counter of the communication node is less than the limit of the number of transmission cells set in the window size register and the cell exists in the transmission buffer of the communication node, the highest address of the other communication node is the highest address identification / setting. When described in the address field of the communication node received by the section 10, the address of the communication node is set in the address field of the communication cell to which the downstream neighbor communication node is to be transmitted, and the communication cell is located in the downstream neighbor. Sent to the communication node. Here, downstream means the transmission direction of the cell.

When the transmission of the cell allocated to the node at the window size is completed, the value indicated by the number of transmitting cells counter of the transmitting node is equal to or larger than the limit value of the transmitting cell set in the window size register, or the cell is stored in the transmission buffer of this communication node. In the absence of a situation, when the highest address of another communication node is described in the highest address area of the received communication cell, the value set in the highest address area of the receiving cell is the highest address area of the communication cell to be transmitted to the downstream neighbor communication node. Is set, the cell is transmitted to a downstream neighbor communication node.

In the processing (handling) of the communication cell as described above, if the address described in the address area of the communication cell received by the communication node does not match the address of the communication node, the address area and the user data area of the communication cell are not changed. The received communication cell is transmitted to the downstream neighbor communication node. When the address described in the address area of the communication cell received by the communication node matches the address of the communication node, that is, the receiving cell is a communication cell transmitted to the communication node or the received communication cell is not a cell actually used for communication, If the value indicated by the communication cell counter of the communication node is less than the value in the window size register and the cell exists in the transmission buffer of the communication node, the communication cell existing in the header of the transmission buffer of the communication node replaces the received cell. Is sent to the downstream neighbor communication node.

When the address written in the address area of the communication cell received by the communication node matches the address of the communication node or the received communication cell is not a cell actually used at the time of communication, the value displayed by the number of transmitting cells of the transmitting node is displayed at If the cell is larger than the value in the size register or there is no cell in the transmission buffer of the communication node, the cell not used for actual communication is transmitted to the downstream neighbor communication node instead of the received cell.

If the value displayed by the number of transmission cell counters of the communication node is equal to or larger than the limit value of the transmission cell set in the window size register or the highest address equivalent to the communication node is set in the highest address area of the received communication cell, the communication system. The reset data for initializing the number of transmission cell counters of all the communication nodes within is set as a trigger for resetting to the address area of the transmission cell to be transmitted to the downstream neighbor communication node, and the communication cell is transmitted to the downstream neighbor communication node. Data relating to the fact that the communication cell with the reset data set therein has been transmitted is stored in the communication node. In the communication node to which reset data is transmitted, when the highest address of another communication node is described in the highest address area of the received communication cell, the highest address of the communication node (to which the reset data is transmitted) is transmitted to the downstream neighbor communication node. If communication is set in the lowest address area of the communication cell to be communicated and communication is sent to the downstream neighbor communication node, and reset data is described in the address address area of the received communication cell, the communication node remembers that the node has transmitted reset data. Cancels and sets the highest address of the communication node in the highest address area of the communication cell to be transmitted to the downstream neighbor communication node, and transmits the communication cell to the downstream neighbor communication cell.

Next, when the communication cell with the reset data set therein rotates once, the next window cycle is executed and transmission starts.

In the above communication system, the time from when the communication cell from the terminal is stored in the transmission buffer of the communication node until the communication cell reaches the target terminal through the transmission path is the time of the communication cell to be transmitted through the transmission path or It is divided into a fixed delay time such as the time required for processing in the communication node and a delay time that varies depending on the situation of the communication network, such as the time required for access to the transmission path or the time required for waiting in the communication node. The window size of each communication node connected to the communication network and the communication period set in common among all communication nodes are closely related to the delay time which changes according to the situation of the communication network.

The method of securing a delay time until a communication cell transmitted from a communication node passes through a transmission path and several communication nodes reach a communication node transmitting a communication cell again in a communication network connected to each other by a ring-shaped transmission path is provided in Japan. Patent Publication No. 329045/1992. In the above-described delay time securing method, all communication nodes of the communication network recognize the communication section according to the communication cell used in the ATMR protocol, and the maximum delay time of each communication node is suppressed to be twice as long as the communication section. The window size is determined.

In the above-described example of the prior art, all nodes first determine a common window period and execute control so that the transmitting cell assigned to each transmitting node is transmitted within this determined window period, so that each node checks the common reset in the window. It is necessary to provide a reset flag setting circuit.

In addition, a maximum address having a unique cell format different from the standard cell specified in Recommendation ITU-T I. 361 is specified, and a common window while all communication nodes check and communicate the contents of the cell header in each communication node. It is necessary to recognize the specific address setting / identification circuit.

In addition, depending on the delay time that needs to be measured for the delay of the communication cells classified at least in the high priority class, the cells having the high priority may be delayed or abandoned at times.

An object of the present invention is a ring-type ATM in which communication cells assigned to each communication node are transmitted within a window period, which is a communication section having a standard cell format, without requiring a reset recognition circuit or a specific maximum address setting / identification circuit in each node. It is to provide an access control device of a node.

The access control apparatus of the ring-type ATM node according to the present invention determines whether the number of transmission cells waiting for transmission in the communication node within the window size specified in the system is within the number of transmission cells allocated to the communication node, and makes empty cells in the communication node. If it is determined that there is a cell or a cell target and there are transmitting cells within the number of transmitting cells allocated to the communication node, the cell is transmitted to the transmission path. If the number of cells waiting to be transmitted is equal to or larger than the number of cells allocated to the communication node, transmission from the communication node is suppressed until the next window size period starts. The transmitting cell is held in the transmission buffer for at least two (Wn _a ) cell times. Here, the window size is equal to the W cell time, and n _a represents the number of cells that the communication node can transmit to the window W.

Further, up to 2n _a of the transmitting cell are held in the transmission buffer in the communication node, and transmitted to the transmission path without fail during the next window size period.

The access control apparatus of the ring-type ATM node according to the present invention determines whether or not the number of transmission cells waiting for transmission in a communication node within a window size periodically refreshed according to a period specified in the system is within the number of transmission cells assigned to the communication node. In case it is determined that there are empty cells or cell targets in the communication node and there are transmitting cells within the number of cells allocated to the communication node, the transmitting cell is transmitted to the transmission path. When the transmitting cell waiting for transmission is equal to or larger than the number of cells allocated to the communication node, transmission cell transmission from the communication node is suppressed until the next window size period starts. Next, the transmitting cell is held in the transmission buffer for at least (N + 3) W− (N + 3) n _a cell time. Here, N represents the total number of communication nodes in the communication network.

Further, up to (N + 2) n _a of the transmitting cell is held in the transmission buffer of the communication node, and the transmitting cell is transmitted to the transmission path within the next window period.

In addition, the transmission cells belonging to the upper two high priority groups are transmitted from the transmission buffer in the communication node to the transmission path.

Also, a transmitting cell having a high priority and held in the transmission buffer in the communication node is transmitted to the transmission path before the cell is abandoned.

The above objects and novel features of the present invention will become apparent from the description of the present specification and the accompanying drawings.

The following description of the first embodiment of the present invention is a basic concept of the present invention, assuming that the window size is determined for each node without combining with other nodes, the window is controlled according to the so-called sliding window control system It is becoming.

Suppose a cell is transmitted between several communication nodes interconnected with each other through a transmission path in a relay-preferred transfer system in which a cell received in a transmission path is transmitted first. The maximum value of the delay time change according to the situation in the communication network and the cell transmitted from the communication node during the time from the time stored in the communication buffer of the communication node until the cell reaches the communication node that transmitted the cell. Describes the number of buffers of the communication node (required) required to avoid giving up. In the following description, window control is executed in accordance with the sliding window control in which the windows are sequentially updated once per cell time.

FIG. 1 is a detailed block diagram showing an access control apparatus for a communication node according to the present embodiment, in which (13) shows a window control section for controlling window periods for each node. The other part is the same as that of the prior art shown in FIG. As is clear from the figure, this embodiment does not require the busy address identification / setting section and the reset flag storage section. That is, the cell header may be based on a standard configuration as shown in FIG. 3 is an explanatory diagram for the operation of the sliding window control. FIG. 3A shows the relationship between the actual time, the window period and the number of transmission cells, and FIG. 3B is an example of the elapsed data in the cell transmission stored in the window control unit 13.

The operation of the node having the above configuration will be described below. When the system is initialized, each node 1 sets the window size determined by the system in the window size register 18 and the number of transmit cells allocated to the node in the transmit cell number counter 17. The window control unit 13 checks the number of transmitted cells counter 17 by itself for each cell slot, compares the checked number with the number of transmitted cells assigned to the node in the comparator 19, and the number is less than or equal to a predetermined value. In this case, the transmission control unit 14 transmits the transmission cell of the node from the transmission buffer 15 to the transmission path 2.

An example of the delay change according to the situation of the communication network in the relay priority transmission system is an access delay from the time when the cell transmitted from the communication node 1 is stored in the transmission buffer 15 in the communication node until the cell is actually transmitted. to be.

The transmitting cell generated in the communication node 1 can actually transmit from the transmission buffer 15 in the communication node to the transmission path 2 only when a cell slot or a cell transmitted to the communication node is received. . The transmitting cell may be inserted directly into a cell slot not yet used, and the cell sent to the communication node is fetched into the communication node 1 and the transmitting cell is inserted at the same time. Therefore, in the relay-first transmission system, the delay time according to the situation in the communication network from the time when the cell is stored in the transmission buffer 15 in the communication node 1 until the cell reaches the communication node that transmitted the cell. In order to obtain the maximum value of the change, it is only necessary to obtain the maximum access delay time at the communication node. In the relay-first transmission system, the access delay is maximized when the cell pattern reaching the communication node from the transmission path in the communication node is a cell slot that is already used or a pattern in which the maximum number of cell slots that are not sent to the communication node continues. .

Next, the maximum access delay will be described with reference to FIGS. 4 and 5. FIG.

Hereinafter, a window having a start position in a cell slot having a transmitting cell inserted in the communication node a (where the window size is equal to the W cell time) will be described. In this window, the maximum number of cells that can be created in a communication node other than communication node a is n _i -n _a ). here, n _i is the total number of cells that can actually be transmitted to the transmission path through the transmission path in the window W in all communication nodes connected to each other. N _a indicates the number of cells that the communication node a can transmit to the window W.

Total number of cells that all communication nodes can send to Window W n _i is at most W. For this reason, it appears continuously in the window W with the maximum (Wn _a ) of the cell slots already used or set in the communication node as the cell pattern arriving at the transmission path in the communication node a. These cells may extend over several windows. In addition, assuming that a cell slot having a transmission cell inserted in the communication node a is extended to a section (band) 2W consisting of two windows, a maximum of 2 (Wn _a ) cell slots that have already been used or are not transmitted to the communication node a are It may appear continuously. When the above-mentioned arrival pattern is generated, in this section 2W, 2 (Wn _a ) of the cell slots that have already been used or not transmitted to the communication node are not used before or after being continuously transmitted to the communication node a or not. N _a in the cell slot is always present.

The number of cells that can be transmitted in this section 2W in the communication node a is 2n _a cells. In this case, the maximum access delay in this case is generated at the head of _a series of 2 (Wn _a ) cell slots in which the transmitting cell from communication node a cannot be used and enters the transmission buffer. It is a period of up to.

Fifth Fig, first transmission of the 2n _a for the generated transmission cell cell to n _a is either already in use or communication node after a series of 2 (Wn _a) a non-transmission cell slot as not yet used or a communication node is transmitted with n _a of the cell slot sent to a, the remaining n _a of the transmission cell is sent is not already in use or with the communication node a a n _a cell slot sent out to present the head of the next interval 2W. As such, it can be seen that the maximum access delay time of the transmitting cell transmitted from the communication node a is 2 (Wn _a ) cell times.

In all communication nodes connected to each other through a transmission path, cell transmission is performed in accordance with a relay-first transmission system in which a fixed-length cell reaching the transmission path is preferentially transmitted while executing sliding window control on a cell transmitted from the communication node. Is executed, the maximum access delay time from the time when the cell transmitted from communication node i is stored in the transmission buffer in communication node i until the cell reaches the communication node i transmitting the cell is

From this, it can be seen that the maximum access delay time is within twice the window size W cell time.

Next, in order to ensure that the cell transmitted at the communication node is not abandoned at the communication node, the minimum number of buffers required (required) for the communication node is considered.

6 is an explanatory diagram for the remaining transmission cells in the transmission buffer. As described with reference to FIG. 4, when the maximum number of cell slots that have already been used or not transmitted to the communication node appear continuously, the maximum number of transmitting cells is stored in the transmission buffer in the communication node. Up to two (Wn _i ) of the cell slots that have already been used or not sent to the communication node appear consecutively at the communication node i. The transmitting cells transmitted from the communication node i during this time are stored in the transmission buffer, and the maximum number of transmitting cells generated from the communication node i in the 2 (Wn _i ) cell slot is 2n _i . For this reason, the minimum number of transmission buffers necessary for ensuring that the cell transmitted at the communication node i is not abandoned at the communication node i is 2n _i .

7 shows the worst case of the arrival pattern from the transmission path and the buffer required in this case when the number of communication nodes is 4, the number of cells that can be transmitted by each communication node within W cell time and the window size is 8 cell time. It shows the number. In other words, the number of buffers required in this case is 4, which is 2n _a .

In the above description of Embodiment 1, it is assumed that the same cell is used for all types (types) of communication without considering the use of each individual cell for other purposes. In this case, communication in which the delay time is necessary and communication in which the delay time is not necessarily processed are processed under the same conditions. For this reason, by using a structure in which a cell that is preferentially processed according to the delay time is applied to a communication that requires a delay time, and a cell that is not preferentially processed according to the delay time is used for a communication that does not necessarily require a delay time. In addition, by providing a different buffer in each communication node in the communication network and storing the preferentially processed cells according to the delay time in the buffer for storing the preferential cells, the preferred cells are always drawn out from the head of the buffer and transmitted. By using this type of structure, it is possible to prevent unnecessary delays caused by performing a communication in which a delay time is not necessarily required for communication in which a delay time is absolutely necessary.

In addition, several priority levels are introduced to realize buffers each containing several priority bands.

When the above described processing is executed in each communication node in the communication network, it becomes clear that the above-described maximum access delay time of the preferred cell and the minimum number of preferred cells for each communication node are the same as those in the first embodiment.

In the following description of this embodiment, it is assumed that window control is executed in accordance with a so-called jump window control system.

Hereinafter, the maximum delay time from the time when the transmitting cell is stored in the buffer to the time when the cell reaches the communication node transmitting the transmitting cell according to the relay priority transmission system in the ring-type ATM system including a plurality of communication nodes. The number of buffers required to prevent the transmitting cell to be transmitted from being abandoned at the communication node will be described.

8 is an explanatory diagram of jump window control in which a window is updated once every W cell time. Here, W represents a value of the window size normalized to the cell time.

The detailed configuration of the communication node 1 is the same as that shown in FIG. 1, and the window control unit 13 monitors the next W cell slot time every time the W cell slot time shown in FIG. 8 passes and controls necessary. Run

Next, the delay at the time of access according to the above-described operation will be described.

When a cell slot or a cell transmitted to a communication node that has not been used is received, transmission from the transmission buffer 15 in the communication node 1 to the transmission path 2 is actually executed. Cells not yet used are fetched into cells used for transmission or sent to the communication node, and then the transmitting cells are inserted. Therefore, in order to obtain the maximum value of the delay time from the time when the transmission cell in the communication node 1 is stored in the transmission buffer 15 to the time when the transmission cell reaches the communication node that transmitted the transmission cell, Similarly, it is only necessary to obtain the maximum access delay time at the communication node. The access delay time is maximized when the maximum number of cell slots already used or not sent to the communication node is continued as the cell pattern reaching the communication node 1 in the transmission path 2.

In jump window control, the transmitting cell may be generated in any cell slot in the window. For this reason, as shown in FIG. 9, cells which can be transmitted in two consecutive windows can be generated continuously. Assuming that a cell is reaching communication node a in a transmission path, if a cell as shown in FIG. 9 occurs in the same cell slot in all communication nodes, a cell slot that has already been used or has not been sent to communication node a is It is generated continuously. Briefly, assuming that the number n _i transmittable in the window is the same for all communication nodes, the window size W can be obtained by the following equation.

Here, N represents the total number of communication nodes in the communication network, and W represents a value obtained by converting the window size into the number of cell slots.

Next, with reference to FIG. 10, the worst condition of the cell delivery pattern to a given communication node a which can be known from the above equation in the transmission path will be described. Here, it is important that the window size W is determined by the total number of cells that can be transmitted to the window in all communication nodes. In other words,

to be.

here, to be. For this reason, the following formula is established.

Hereinafter, it is assumed that the window band 1 of the cell pattern reaching the communication node in the transmission path can be used by the transmitting cell. The number of transmitting cells that can be included in the window band ① is expressed by the following equation.

It becomes clear that all cell slots in this window band 1 can be used by the transmitting cell. Next, assume a condition that a band ② consisting of a window band ① and a sequential window band is used by a transmitting cell. The number of transmitting cells included in station ② is represented by the following equation:

Under the condition that all cell slots in this band ② are used by the transmitting cell

The following formula is obtained by this.

As can be seen from Equation (1), the above equation shows a condition that is satisfied when the number of communication nodes N is two or more, but all cell slots in this band ② assuming that there are several communication nodes in the communication network. Is used by the communication cell.

Similarly, assume a condition that N in the continuous window band ③ having the window band? At the head is used by the transmitting cell. The number of transmission cells included in the band ③ is represented by the following equation.

The above formula can be modified from the formula (1) to the following formula.

The conditions under which all cell slots in this band ③ can be used by the transmitting cell are:

The following formula is obtained by this.

It can be seen from Equation (1) that the equivalent sign of the above equation becomes valid, indicating that all cell slots in this band ③ can be used by the communication cell. In this case, the length of the band in which the cell slots which are already used or not transmitted to the communication node a continuously appear includes a cell transmitted from a communication node other than the communication node a in the window ⑤ following the band ④ (N + 2). W- (N + 2) n _a cell slot.

Under the condition that the window band④ including (N + 1) of the continuous window band including the window band① at its head is used by the transmitting cell

It is clear from equation (1) that the above conditions are satisfied.

As described above and shown in FIG. 11, cell slots that have already been used or have not been sent to communication node a may appear continuously at maximum (N + 2) W− (N + 2) n _a .

Hereinafter, the access delay time in the transmitting cell when the cell from the transmission path is received by the communication node a in the above-described pattern is considered. The worst case is the first cell slot of a series of (N + 2) W- (N + 2) n _a cell slots that are already received or have not been sent to communication node a and received from the communication node. This is the case where the transmitting cells are generated in the transmission pattern as shown in FIG. 9, and the maximum number of transmitting cells is generated in the sequential window.

In a band of (N + 2) W- (N + 2) n _a cell slots in which cell slots that have already been used or not transmitted to communication node a are continuously received, the transmitting cell transmitted from communication node a is sent to the transmission buffer. Stored. Hereinafter, the maximum number of transmitting cells generated by the communication node a in the band consisting of (N + 2) W− (N + 2) n _a cell slots is considered. The minimum time required for generation of (N + 1) n _a of the transmitting cell from communication node a is (N-2) W + 2 cell time as shown in FIG. Therefore, the following formula

(N + 2) W− (N +) in which the number of (N + 1) n _a or more of the transmitting cells transmitted from the communication node a is already used or the cell slots that are not sent to the communication node a are continuously received. 2) generated in the n _a cell slot band, only the following conditions are required,

This indicates that the number of communication nodes is two or more. Assuming that there are several communication nodes in the network, the number of (N + 1) n _a or more transmitted from communication node a has been consecutively received (N + 2) W− which has already been used or has not been sent to communication node a. It can be seen that the condition of generation in a band consisting of (N + 2) n _a cell slots is satisfied.

The minimum time required for a transmission cell of (N + 2) n _{a to} be generated in communication node a is NW + 2n _a cell time as shown in FIG. That is, the following formula

Is provided, and the transmission cell of (N + 2) n _a or more transmitted from communication node a is condition N. When 4 is satisfied, i.e., the number of communication nodes is 4 or more, it is generated in a band consisting of consecutively received (N + 2) W- (N + 2) n _a cell slots that have already been used or not transmitted to the communication node.

Similarly, the minimum time required for a transmission cell of (N + 3) n _{a to} be generated in communication node a is (N + 1) W + 2n _a cell time as shown in FIG. That is, the following formula

(N + 2) W- (N + 2) n which is received continuously from the communication node, where more than (N + 3) n _a transmitting cells are not already used or are not sent to communication node a. _a not generated in the band consisting of cell slots.

In the above description, consecutively received (N + 2) W- (N + 2), in which _a transmitting cell of (N + 2) n _a transmitted from communication node a has not already been used or has not been sent to communication node a. It will be apparent from the following description that the maximum number of transmitting cells is generated in the case of being generated in a band consisting of n _a cell slots and in the case of an access delay time in the transmitting cell.

15 is an explanatory diagram of an access delay.

In the arrival pattern shown in FIG. 11, it is impossible to insert the transmission cell from the communication node a into the first N window. Since at least n _a of the cell slots not yet used or transmitted to the communication node a exists in the next window, the previously created transmission cell of n _a is inserted therein. That is, the access delay time in the transmitting cell is (N + 2) W− (N + 2) n _a cell time. Transmitting _a cell in the n generates the second is further delayed by a first window (W-hour cell) from the cell slot with the transmission of the cells n _a first insert is produced in its interior. In other words, the access delay time in the transmitting cell is (N + 3) W− (N + 3) n _a cell time. In the transmission cells sequentially generated, the access delay time is (N + 3) W- (N + 3) n _a cell time.

Relay priority transmission system in which the arriving cell preferentially transmits by providing jump window control for the cell transmitted from the communication node under the condition that all transmitting nodes have the same number of cells that can be transmitted in a normal window having the same size. In this case, the maximum access delay time from the time when the transmitting cell is stored in the transmission buffer in the communication node i to the time when the transmitting cell arrives at the communication node i transmitting the transmitting cell is not only the number N of communication nodes in the communication network. Depending on the window size W, this means that the delay time is made sure to have (N + 3) W− (N + 3) n _i cell time.

In order to ensure that the cell to be transmitted at the communication node is not abandoned in the communication node, the minimum number of buffers in the communication node is considered.

As described with reference to FIG. 11, when the maximum number of cell slots that have not been used or transmitted to the communication node appear continuously, the maximum number of transmitting cells is stored in the transmission buffer in the communication node. The cell slots as described above appear consecutively at maximum (N + 2) W− (N + 2) n _i . The transmitting cell from the communication node i that can be generated within this cell time is stored in the transmission buffer, and can generate a cell of (N + 2) n _i as described above. For this reason, the minimum number of buffers required to ensure that a cell to be transmitted at communication node i is not abandoned in communication node i depends on the number of cells that can be transmitted into the window at communication node i as well as the number of communication nodes in the communication network. The number is (N + 2) n _i .

For reference, FIG. 16 shows an example of the number of buffers required when the worst case of the arrival pattern from the transmission path and the number of communication nodes are four. The number of cells that can be transmitted within W cell time in each communication node is shown in FIG. 2 and the window size is 8 cell hours.

In Example 3 mentioned above, the case where each communication is used for various purposes in communication is not considered, and the case where the same cell is used for all types (types) of communication is demonstrated. In this case, communication in which the delay time is not necessary and communication in which the delay time is necessarily processed under the same conditions. Therefore, if the preferential cell in delay time is applied to communication that requires delay time, and the non-priority cell is applied to communication that does not necessarily require delay time, provide another buffer in all communication nodes in the communication network By storing the preferential cell in the cell that stores the preferential cell, the preferred cell is always drawn from the head of the buffer and transmitted. In this configuration, it is possible to prevent unnecessary delay time caused by performing a communication in which a delay time is not necessarily required for communication in which a delay time is absolutely necessary. The priority levels may be classified into several.

The maximum access delay time and the minimum number of buffers of the preferred cell required by the communication node are the same as those of the third embodiment on the condition that the above-described processing is executed on each communication node in the communication network.

As described above, in the present invention, since each communication node is provided with a transmission buffer having a capacity larger than a predetermined value, the number of transmission cell counters, and the transmission control of the window, each communication node is provided with a delay less than the delay determined in the entire system. The transmitting cell can be transmitted to the transmission path.

In addition, since a larger number of transmission buffers are provided in each node than a predetermined value, it is possible to prevent the transmission cell from being abandoned.

In addition, since each node has transmission buffers having a capacity larger than a predetermined value, transmission cell counters, and transmission control means for monitoring and controlling the transmission of windows, the transmission cells can be transmitted to the transmission paths with a delay less than that determined in the entire system. Can be.

In addition, since each node has a larger number of transmission buffers than the predetermined value for jump window control, it is possible to prevent abandonment of the transmission cell in the communication node.

In addition, since a transmission buffer having two or more types of priorities is provided in each node, it is possible to prevent a high priority transmission cell from being abandoned because a transmission cell having a high priority is preferentially transmitted.

In addition, since more than one predetermined number of priority cells transmit buffers are provided in each node, it is possible to prevent abandonment of a transmission cell having a high priority.

It is a matter of course that the present invention is not limited to the above embodiments and can be modified in various forms without departing from the spirit and scope of the invention.

Claims (8)

If the window size in the ring is W cell and the number of transmitting cells assigned to the communication node in the W cell is n _a , a transmission buffer for storing the transmitting cell therein for a time corresponding to at least 2 (Wn _a ) cells. A transmission control means for transmitting the transmission cells in the transmission buffer to the transmission path if the number of transmission cells counting the number of transmission cells from the node within the window size and the number obtained by the transmission cell number counter correspond to the number assigned to the communication node; Access control device for a ring-type ATM node comprising. The access control apparatus for a ring ATM node according to claim 1, wherein the number of transmitting cells assigned to the communication node is n _a , and at least 2n _a transmission buffer is provided. The transmission cell according to claim 1, wherein the priority of the transmitting cells in the node is classified into two or more levels, and transmission buffers corresponding to the priority levels are provided, respectively. An access control apparatus for a ring-type ATM node sent before a transmitting cell of the terminal. 4. The access control apparatus of claim 3, wherein the minimum number of transmission buffers is at least 2n _a . If each node is N in the ring, the window size is W cell, and the number of transmitting cells assigned to the communication node in the W cell is n _a , at least (N + 3) W- (N + 3) n _a cells A transmission buffer for storing a transmission cell therein for a corresponding time; a window control circuit for counting the W cell time and resetting the window to set a new window size each time the W cell time elapses; communication of the window size A ring type comprising a transmission cell number counter for counting the number of transmission cells from the node and transmission control means for transmitting the transmission cells in the transmission buffer to the transmission path when the number of cells indicated by the transmission cell number counter matches the number allocated to the communication node; Access control device of ATM node. 6. The access control apparatus for a ring ATM node according to claim 5, wherein the number of transmitting cells assigned to the communication node is n _a and at least (N + 2) n _a transmission buffer is provided. The transmission cell according to claim 5, wherein the priority of the transmitting cells in the node is classified into two or more levels, and a transmission buffer corresponding to each of the priority levels is provided. An access control apparatus for a ring-type ATM node sent before a transmitting cell from the same. 8. The access control apparatus of claim 7, wherein the minimum number of transmission buffers is at least (N + 2) n _a .