JP2001320374A - Connection reception control method and atm exchange - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a connection reception control method for reducing the load on the connection reception control of an ATM exchange and the occurrence of packet rejection. SOLUTION: In the connection reception control of a nonparametric method, when packet communication service is supplied in AAL, a false appearance is given that monitored cells are decided to be only the head cells of respective packets, subsequent cells arrive at PCR entered at connection time, and the respective packets are constituted of K-pieces of cells.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an ATM (Asynchronous Tran) for performing packet communication.
The present invention relates to a connection admission control method to be performed when a new connection connection request is made in an sfer mode network and an ATM exchange capable of executing the connection admission control method.

[0002]

2. Description of the Related Art ATM stores information in 53-byte "cells".
This is a transfer method that divides the data into units and transfers them at an arbitrary timing.Since the cells are short and fixed in length, the exchange can perform high-speed hardware processing without being aware of changes in the packet length unlike the packet method. And high-speed transfer is possible.

[0003] Further, the terminal declares in advance what communication traffic characteristics the communication to be performed is and what kind of quality is required by an ATM-specific parameter in advance, and is composed of an exchange and a transmission path. ATM
The network also has a feature that the quality during communication is guaranteed by providing an ATM connection according to the network.

FIG. 12 is a schematic diagram showing a configuration of a conventional ATM network. As shown in FIG.
The TM network 100 includes a plurality of ATM switches 110
Consists of In FIG. 12, the ATM terminal 101
First divides information to be sent to the other party into cell units. The cell is VP (Virtual Path) or VC
(Virtual Channel) to the ATM exchange 110 via a virtual communication path. Note that, for the non-ATM terminal 102, the above processing is performed by the ATM.
Terminal adapter (TA) 1 with interface
03 through the ATM network 10
0 can be enjoyed. Hereinafter, the ATM terminal 10
In the case of 1, the non-ATM terminal 102 is also included.

[0005] Normally, VP / VC is multiplexed on one transmission line, and in order to identify VP / VC, a cell is:
In the header, VPI / VCI (V
actual Path Identifier / Vi
true Channel Identifier).

FIG. 13 is a block diagram showing a schematic configuration of the ATM switch 110. As shown in FIG. The ATM exchange 110 includes a line handling unit 120 for accommodating subscribers and connecting to other ATM exchanges, and a switch unit 130 for controlling exchange of cells exchanged in the ATM network 100.
And a control unit 140 that performs various traffic controls necessary for ATM exchange. Also, ATM
A management device 150 for controlling the operation of the exchange 110 is also connected.

The traffic control executed by the control unit 140 includes routing control and connection admission control (CAC: Connection Admission).
onControl), usage parameter control (UP
C: Usage Parameter Contro
l), shaping, cell priority control, and the like, which are implemented by software. Hereinafter, these basic traffic controls will be described.

[0008] The start of communication in the ATM exchange is performed by the ATM.
This is performed by the terminal 101 transferring a call signal by a cell to the ATM exchange 110. A predetermined VCI is given to a cell for transferring a call signal,
It is distinguished from cells for transmitting information (payload). In response to the reception of the cell indicating the call signal, the ATM exchange 110 obtains information such as destination information of a connection destination, a service type, and a cell speed.
Determine which other ATM exchange in 0 to go through.
This process is called routing control.

The ATM exchange 110 transfers control cells one after another to the connected ATM exchange determined by the above routing control, and each ATM exchange can pass the cell under the required service conditions. Check if it is possible. For example, it is determined whether or not the requested traffic does not exceed the transmission capacity using parameters such as a cell delay and a cell loss rate. Then, the ATM terminal 101
To the communication mode, and shifts to the communication mode. This process is called connection admission control (hereinafter, referred to as CAC).

In the ATM, by managing the entire capacity, the network equipment can be effectively used even for bursty traffic. However, when a specific call flows a large number of cells exceeding the contracted amount, the entire network can be used. There is a risk of deteriorating quality. Therefore, in the ATM switch 110,
When the call is accepted, it is monitored whether the declared value of the transmission capacity matches the actual inflow of cells. For example,
For each contract with the subscriber, if the actual traffic exceeds the specified value, the cells that violate the specified rules are discarded. This process is called usage parameter control.

As described above, in the ATM switch 110, a large number of connections are made at the same time, but the cells input to the ATM switch 110 are checked for the usage parameter control so as to satisfy the above-mentioned conditions permitted by the CAC. is recieving. However, in the ATM switch 110, the switch unit 1
Because cells are collected on each route by 30, cells may be concentrated on a particular route, and as a result, cells may have to be delivered at a fast instantaneous rate beyond the capacity of the VP. For this reason, the ATM switch 110 temporarily accumulates each cell in the buffer memory and sets the instantaneous speed of the cell to a constant value (V
The cell is read so as not to exceed (P capacity). This function is called shaping, and the above-described buffer memory is usually provided in the switch unit 130.

Here, excessive transmission of cells transferred from a specific user can be prevented beforehand by the above-mentioned usage parameter control, but excessive traffic is input to the entire ATM switch 110. If A
In the TM exchange, cell loss may occur as described above. For example, if traffic having a very high peak is input suddenly when the input data amount is excessively large, the buffer memory in the switch unit 130 instantaneously overflows.

[0013] Therefore, in the ATM exchange, in preparation for such a situation, which cell is transmitted with priority is controlled, and a cell loss rate of a specified quality or more is secured for a cell to be transmitted with priority. I have. Therefore, the cell contains C in its header.
An LP (Cell Loss Priority) bit is defined so that the user can indicate which cell is desired to be transmitted with priority. The cell priority control refers to a process of determining which cell is to be discarded according to the state of the CLP bit when it becomes necessary to discard the cell.

Among such traffic controls in the ATM switch 110, CAC is an important function in an ATM network and is an interesting research subject. As a specific method of connection admission control, a non-parametric method of performing admission control based on actual traffic measurement results and a parametric method of performing admission control using declared traffic parameters are known. Each method is described, for example, in the document "Basic of Communication Traffic Theory and Multimedia Communication Network, The Institute of Electronics, Information and Communication Engineers, 1995". Hereinafter, an outline of these non-parametric methods and parametric methods will be described.

(Non-Parametric Method) First, the non-parametric method is characterized in that control based on constantly measured items of actual traffic is performed while assuming as little traffic characteristics as possible. In particular, as examples of the CAC according to this non-parametric method, a dynamic CAC and an advance information update CAC are known.
For these CACs, see, for example, the document “B-ISD
Measurement driven traffic technology in NTT, NTT R & D
Vol. 44, 1995 ".

In the dynamic CAC, when the cell loss rate of a VP in an ATM network is evaluated, the number of cells arriving within a predetermined measurement time within the VP (hereinafter referred to as the number of arriving cells) is counted, and a new CAC is calculated. The cell loss rate after accepting the connection request is estimated, and acceptance judgment is made by comparing with a specified value.

In this dynamic CAC, all arriving cells are monitored while performing window control based on the measurement time, and the distribution characteristics are constantly updated from the number of arriving cells for each collected window. Then, at the time of a new connection connection request, it estimates how the above-mentioned constantly updated distribution characteristic changes with the expected number of arrival cells within the measurement time derived from the declared traffic parameters, and The reception decision is made by evaluating the cell loss rate when the request is received.

FIG. 14 is an explanatory diagram for explaining the operation of the above dynamic CAC. First, it is assumed that the number of cells per unit time is counted and distribution data p (k) is obtained as shown in FIG. Then, in the reception determination at the time of a new connection connection request, when this connection request is received, what happens to the distribution data p (k) is predicted.

For example, PCR (Peak Cell)
When it is predicted from the traffic parameters such as the declared value of Rate: maximum cell rate) that a connection request VC is likely to occur in R cells per measurement time, p (k) after accepting this VC is shown in FIG. As shown in FIG.
P (k) shown in (a) is shifted to the right by R. If the new p (k) obtained in this way is applied to a decision formula relating to the cell loss rate, acceptance decision can be made.

The arrival cell is monitored by the ATM switch 11.
This is performed by the line corresponding unit 120 of 0. FIG. 15 is a block diagram illustrating a schematic configuration of the line corresponding unit. In FIG. 15, a line corresponding unit 120 includes an O / E unit 121 and an E / O unit 1 that perform electric / optical conversion that is input / output via a line.
22, a physical layer processing unit 123 that performs transmission convergence sublayer processing, a traffic monitoring mechanism 124 that determines the suitability of received traffic using the above-described UPC and the like, and an arrival cell monitoring unit 1 that monitors arrival cells.
25.

Here, a protocol provided based on the ATM method includes a physical layer for specifying physical media and an A for handling cell transfer common to all services.
It has a hierarchical structure of a TM layer and an ATM adaptation layer (hereinafter referred to as AAL) in which a plurality of protocols are defined by handling functions corresponding to each service. Further, the physical layer includes a physical medium sublayer that provides a function of realizing the specification of the physical medium and electrical / optical conversion, and a transmission convergence sublayer that provides a function of inserting and extracting cells by forming a transmission frame. There are two provisions. That is, in the line corresponding unit 120, the physical layer described above is realized by the O / E unit 121, the E / O unit 122, and the physical layer processing unit 123.

FIG. 16 is a schematic block diagram of the arrival cell monitoring unit. As shown in FIG. 16, the arrival cell monitoring unit 125
A window control 126 for controlling a monitoring cycle, a counter 127 for counting arrival cells, and a control unit 140
Control unit interface 1 that notifies monitoring results to
28. Note that the window control 126 and the counter 127 are allocated for each monitoring unit.

In the arrival cell monitoring unit 125, when a cell arrives,
First, determine which monitoring unit the arrival cell corresponds to,
Perform cell sorting. For example, if the monitoring unit is VC
In the case of (1), sorting is performed according to the value of VPI / VCI in the cell header. Next, the arrival cell monitoring unit 125
Increments the counter corresponding to the VC by one. The window control 126 performs time monitoring. When the monitoring time has been reached, the control unit interface 128 is instructed to read the counter 127, and after reading, initializes the counter 127 and enters the next monitoring cycle. After reading the counter 127, the control unit interface 128 notifies the control unit 140 of the read value.

On the other hand, the prior information update type CAC performs performance estimation by updating the performance (advance information) based on an assumed traffic model with performance measurement values. Includes some statistical assumptions.

In this prior information update CAC, prior information is determined in advance assuming a certain traffic model (which may be different from the actual traffic model), and the true performance (eg, cell loss rate) and the prior information are determined. Is assumed, the Bayesian estimation problem for unknown parameters is replaced, and an estimated value is obtained from the measurement result.
At this time, it is necessary to make statistical assumptions about the error between the estimated value and the true performance and the accuracy of the prior information.

As described above, in the non-parametric method, the dynamic CAC and the advance information updating CAC are used.
However, it is necessary to monitor all the arriving cells in order to measure the cell loss rate and the like.

(Parametric Method) Next, the parametric method will be described. In the CAC based on the parametric method, the reception is determined using only the traffic parameters declared by the user. Then, the declared traffic parameters are applied to a mathematical model, and control is performed by numerical calculation. It should be noted that UBR (Un) which is a traffic class in which a necessary bandwidth is not allocated in advance and the data is transferred if there is a vacancy on the transmission path.
About the specified Bit Rate),
Since there is no traffic parameter to be declared and there is no band to be guaranteed, it is generally excluded from CAC. That is, from the viewpoint of the band, the connection of the UBR class is always accepted.

However, in the actual service, communication is hardly performed depending on the load of the network, and there is a case where the convenience is extremely low for the user. In order to correct this, it is fully conceivable that the network operator gives consideration to give some bandwidth to UBR class connections. UB at this time
MCR (Minimum Cell) in R class
Rate: minimum guaranteed bandwidth).

At the time of a traffic class connection request with MCR, a CA by a parametric method is used.
When C is performed, it is general to guarantee by MCR. FIG. 17 shows C by the parametric method.
FIG. 4 is an explanatory diagram for explaining MCR guarantee in AC. As shown in FIG. 17, the MCR-guaranteed cell arrival model provides guarantees for cells arriving at a certain interval 1 / MCR or more, and cells arriving within that interval are not guaranteed.
That is, this cell arrival model is based on the cell arrival interval = 1 /
CBR (Constant Bit Ra)
te) is considered a class.

By the way, when the above-described cell arrival model is applied to a packet communication application, there is a possibility that packets are frequently discarded at the application level. FIG. 18 is an explanatory diagram for explaining the state of discarding at the packet level when a packet is composed of three cells.

In a traffic class with MCR, cells do not always arrive at a rate lower than MCR.
Therefore, in the cell arrival model shown in FIG.
MCR is guaranteed at the ATM layer level.
As shown in FIG. 8, when a packet is composed of a plurality of cells, some of the cells are discarded, so that packet assembly in the AAL fails and the packet is not transferred between applications. Can be considered.

The CAC in the parametric method
As an example of this algorithm, for example, the document “Proposal of Ultra-light CAC Mechanism in ATM,
5, 1997 ". In this reference, V
It shows modeling for BR (Variable Bit Rate) class traffic and an evaluation method using the model. Traffic, Clump
In a model similar to the ed Cell model, an ON section in which cells arrive by PCR and an OFF section in which cells do not arrive are alternately repeated. FIG.
FIG. 3 is an explanatory diagram for explaining a Cell model.

As shown in FIG. 19, Clumped C
The cell model is an SCR (Sustainable C)
The transmission part based on the “cell rate” (cell rate) is clamped to the transmission part based on the PCR, and the end point of the idle state is set as the reproduction point ((a) in the figure). Is the reproduction point ((b) in the figure).

[0034]

However, in the conventional non-parametric CAC, it is necessary to monitor all arriving cells, as is apparent from the above-mentioned document. There was a problem that it would increase.

On the other hand, C by the conventional parametric method
In the AC, a CBR model of cell arrival interval = 1 / MCR is used for a traffic class with MCR guarantee, so that there is a problem that packet discard frequently occurs in packet transfer between applications.

The present invention has been made in order to solve the above-mentioned problems. In the non-parametric method, the load on the ATM exchange is reduced by monitoring only the first cell, and in the non-parametric method, the burst characteristic is reduced. Connection admission control method and ATM capable of reducing the occurrence of packet discarding by employing a cell arrival model and securing a large buffer capacity
The purpose is to obtain an exchange.

[0037]

Means for Solving the Problems To solve the above-mentioned problems,
In order to achieve the object, in a connection admission control method according to the present invention, an ATM for performing packet communication
In an ATM network accommodating terminals, connection admission control (CAC) performed when a new connection connection request is made from a user or a network administrator
When performing the non-parametric method of making an admission decision based on the actual traffic measurement result, set the number of cells constituting one packet to a predetermined value, monitor only the first cell of the packet transferred in the ATM network, The succeeding cell of the packet is treated as having arrived at the declared maximum cell rate (PCR).

According to the present invention, when a packet communication service is provided in the AAL or the like, in the non-parametric CAC, only the head cell of each packet is monitored, and a subsequent cell is declared when a connection is established. Arrived by the PCR and each packet is K
Since it is presumed to be composed of individual cells, it is not necessary to monitor all arriving cells.

In the connection admission control method according to the next invention, in an ATM network accommodating an ATM terminal performing packet communication, a connection admission control executed when a new connection connection request is made from a user or a network administrator is realized. When performing the non-parametric method of making an admission decision based on the traffic measurement result, the number of cells constituting one packet is set to a predetermined value, and only the first cell of the packet transferred in the ATM network is monitored. The succeeding cell of the packet is characterized in that the subsequent cell is treated as continuously arriving within a range satisfying the declared cell delay fluctuation tolerance (CDVT).

According to the present invention, when a packet communication service is provided in the AAL or the like, in the non-parametric method CAC, only the head cell of each packet is monitored, and subsequent cells are declared when a connection is established. If a cell in one packet arrives without exceeding the declared CDVT, it is treated as a predetermined number of cells in the order of arrival from the head cell, and arrives beyond the CDVT. In this case, since the cells are treated as a predetermined number of cells that have arrived in sequence at the timing of continuous arrival by PCR, the arrival information of the succeeding cells can be obtained even by the appearance of such arrival of the succeeding cells. There is no need to monitor incoming cells.

[0041] In the connection admission control method according to the next invention, in the above invention, when the user or the network administrator declares the number of cells of the maximum packet to be transferred on the connection at the time of the connection connection request. The number of cells is set to the predetermined value.

According to the present invention, when a connection connection request is made, as in the case where the traffic class is GFR,
When the number of cells constituting the maximum packet has been declared, the declared number of cells can be used as the number of cells constituting the packet.

In the connection admission control method according to the next invention, in the above invention, if the number of cells of a packet transferred on the connection is not declared by a user or a network administrator at the time of a connection connection request, The predetermined number of cells is set to the predetermined value.

According to the present invention, when the connection connection request is made, as in the case where the traffic class is CBR,
If the number of cells constituting one packet is not declared, the number of cells constituting the packet can be a fixed number of cells previously given.

In the connection admission control method according to the next invention, in the above invention, among the head cells of the packet, only the head cell in which the CLP bit contained in the head cell is 0 is monitored, and the CLP The first cell whose bit is 1 is not monitored.

According to the present invention, only the cell whose CLP bit is 0 among the head cells of the packet is monitored,
Cells with an LP bit of 1, that is, cells with a low priority set are excluded from the monitoring target, so that the number of cells to be monitored can be reduced.

In the connection admission control method according to the next invention, in the above invention, among the head cells of the packet, the head cell in which the CLP bit included in the head cell is 0 and the CLP bit are one. It is characterized in that both head cells are to be monitored.

According to the present invention, the head cells of all packets can be monitored regardless of the contents of the CLP bit.

In the connection admission control method according to the next invention, in an ATM network accommodating an ATM terminal that performs packet communication, a connection admission control performed when a new connection connection request is made from a user or a network administrator is declared. In the case of performing a parametric method of making a reception decision based on the given traffic parameters, if the number of cells constituting one packet is set to a predetermined value and the minimum guaranteed bandwidth (MCR) is included in the traffic parameters, The cells constituting one packet are treated as if they arrived at the declared PCR, and the average cell arrival rate (SCR) is the MCR as a cell arrival model required in the parametric method. Cells not received like Characterized in that it employs a model having between.

According to the present invention, the average of arriving cells is MC
Since the section arriving and the section not arriving by PCR are monitored so as to be R, a model having a strong burst property can be adopted as a cell arrival model for obtaining arrival information of a subsequent cell. Therefore, more cells can be held in the larger buffer amount.

In the connection admission control method according to the next invention, in an ATM network accommodating an ATM terminal that performs packet communication, a connection admission control performed when a new connection connection request is issued from a user or a network administrator is declared. When the parametric method of making an admission decision based on the traffic parameters is performed, the number of cells constituting one packet is set to a predetermined value, and when the MCR is included in the traffic parameters, one packet is formed. Are treated as if they continuously arrived within a range that satisfies the declared CDVT. As a cell arrival model required in the parametric method, a cell having an SCR equal to the MCR is used as a cell arrival model. Mode with reception section Characterized by employing the Le.

According to the present invention, when a packet communication service is provided in the AAL or the like, in the CAC of the parametric method, only the head cell of each packet is monitored and a subsequent cell is declared at the time of connection connection. It is assumed that the cells have arrived by PCR, and if cells in one packet arrive without exceeding the declared CDVT, they are treated as a predetermined number of cells in the order of arrival from the head cell, and arrive after the CDVT. In
It treats as a predetermined number of cells arriving in sequence at the timing of continuous arrival by PCR, and then monitors sections arriving by PCR and sections not arriving so that the average of the arriving cells is MCR. Therefore, it is possible to obtain the arrival information of the succeeding cell even by such a simulated arrival of the succeeding cell, and it is not necessary to monitor all the arriving cells, and a cell arrival model for obtaining the arrival information of the following cell As a result, a model having a strong burst property can be adopted, and more cells can be held in the larger reserved buffer amount.

[0053] In the connection admission control method according to the next invention, in the above invention, when a user or a network administrator declares the number of cells of the maximum packet to be transferred on the connection at the time of the connection connection request. The number of cells is set to the predetermined value.

According to the present invention, in the CAC of the parametric method, a packet is formed when the number of cells constituting the maximum packet is declared, such as when the traffic class is GFR at the time of a connection connection request. The declared cell number can be used as the cell number.

[0055] In the connection admission control method according to the next invention, in the above invention, when the user or the network administrator does not declare the number of cells of the packet transferred on the connection at the time of the connection connection request, The predetermined number of cells is set to the predetermined value.

According to the present invention, in the CAC of the parametric method, at the time of a connection connection request, as in the case where the traffic class is UBR including MBR,
If the number of cells constituting one packet is not declared, the number of cells constituting the packet can be a fixed number of cells previously given.

In order to execute the connection admission control method according to the present invention, the ATM exchange according to the next invention is provided with a head cell recognition unit for recognizing a head cell and scrutinizing a CLP bit. Features.

According to the present invention, instead of the conventional arrival cell monitoring unit in the line corresponding unit, the recognition of the head cell and the C
The provision of the head cell recognition unit for examining the LP bit enables the provision of an ATM exchange capable of executing the connection admission control method according to the present invention.

[0059]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of a connection admission control method and an ATM exchange according to the present invention will be described below in detail with reference to the drawings. The present invention is not limited by the embodiment.

Embodiment 1 First, a connection admission control method according to the first embodiment will be described. The connection admission control method according to the first embodiment is characterized in that when a packet communication service is provided in the AAL, only the first cell of each packet is monitored.

FIG. 1 is an explanatory diagram for explaining a cell monitoring method of the connection admission control method according to the first embodiment. As shown in FIG. 1, the cell to be monitored is only the first cell of each packet. However, when the non-parametric CAC is executed, information on all the arriving cells is required as described above. The handling of subsequent cells is problematic. That is, in order to apply the connection admission control method according to the first embodiment to the CAC algorithm based on the non-parametric method,
It is necessary to obtain the arrival information of the succeeding cell by some means.

FIG. 2 is an explanatory diagram for explaining the handling of cells other than the head cell. As shown in FIG. 2, the following cell of each packet arrives at the PCR declared at the time of connection connection, and it appears that each packet is composed of K cells. In this case, assuming that the monitoring time of the head cell (first cell) of a certain packet is t (0), the subsequent n-th (2 ≦ n ≦
The arrival time t (n) of the cell in K) can be expressed as t (n) = t (0) + (n-1) / PCR. That is, by monitoring only the first cell, the arrival information of the succeeding cell can be easily obtained.

The first cell of a packet is identified by examining the contents of the cell. However, the content to be inspected depends on the AAL type. For example, in AAL type 5 generally used in data communication, one packet is one AAL protocol data unit (PDU: Protocol Data Uni).
The cell is transferred at t), and whether or not it is the first cell can be determined according to the contents of the PTI field in the cell header, as described later.

FIG. 3 is an explanatory diagram for explaining the contents of the PTI field in AAL type 5. In AAL type 5, when a PDU of user data is divided into cells, as for the last cell, as shown in FIG. 3, a PTI (Payloa) consisting of 3 bits in a cell header is used.
d Type Indicator: a field indicating how the cell payload is used) 1 is set to the least significant bit of the field, and 0 is set to the other cells.

Note that the most significant bit of the PTI is used to identify whether the cell is a user cell or a control cell.
Indicates that the cell is a control cell. Therefore, the least significant bit of the PTI field is the user cell (PT
This is used to identify whether the most significant bit of the I field = 0 and the last cell.

As described above, in the connection using the AAL type 5, the function of the PTI field is used.
When it is determined that the cell of the packet is a cell following the last cell of the previous packet, it can be identified as the first cell. Note that the first cell after the connection is set is the first cell.

FIG. 4 is an explanatory diagram for explaining a method of identifying the head cell in AAL type 5. First, a head cell indication flag indicating that the head cell has been received is provided for each connection, and the cell received when this flag is ON is regarded as the head cell. Note that the initial value of the head cell indication flag when the connection is set is ON. It is also turned on when receiving the last cell, and turned off when receiving a cell regarded as the first cell. Since the initial value at the time of connection setting is ON, the first cell after connection setting can be recognized as the first cell.

Thereafter, the PTI field of each cell is closely examined, and when a cell with the most significant bit = 0 and the least significant bit = 1, that is, a user cell which is the last cell is received, the leading cell indication flag is turned ON. When a further cell is received in this state (this is the first cell), the flag is turned off before the next cell is received. That is, it is possible to recognize that the head cell of each packet has been received when the head cell indication flag falls.

In the AAL type 3/4, one packet is transferred by one AAL-PDU. However, in AAL type 3/4, when a PDU is divided into cells, the SDUs located at the first two bits in the payload of each cell are
In the T (Segment Type) field, information as to whether or not the cell is the first cell is defined.

FIG. 5 is an explanatory diagram for explaining the contents of the ST field in AAL type 3/4. As shown in FIG. 5, in the ST field, the bit state is
"10" indicates a head cell, "00" indicates a continuation cell, "01" indicates a tail cell, and "11" indicates a single cell.

As described above, in the connection using AAL type 3/4, the cell with ST = "10" or "11" can be regarded as the head cell by utilizing the function of the ST field. Specifically, the ST field of each cell is closely examined, and the cell in which the upper bit = 1 is regarded as the first cell.

In each connection, which AA
Whether the L type is used can be determined from information declared by the user (for example, “AAL parameter” that is a signaling information element).

As described above, according to the connection admission control method according to the first embodiment, even in the non-parametric CAC, only the head cell of each packet is monitored, and the subsequent cells are connected to the connection connection. Since it arrives at the PCR declared at the time and each packet appears to be composed of K cells, there is no need to monitor all the arriving cells,
The load on the ATM switch can be reduced.

Embodiment 2 Next, a connection admission control method according to the second embodiment will be described. In the first embodiment, in the cell monitoring method in CAC by the non-parametric method, in order to monitor only the first cell, the arrival information of the succeeding cell is obtained as a packet consisting of the arrival by PCR and K cells. The connection admission control method according to the second embodiment is different in a method of obtaining arrival information of a subsequent cell, that is, a cell arrival model.

FIG. 6 is an explanatory diagram for explaining a cell arrival model in the connection admission control method according to the second embodiment. In the connection admission control method according to the second embodiment, a cell arrival model as shown in FIG.
Pretend to be an actual cell arrival. That is, the plurality of cells constituting each packet are K cells, and the declared P
The time difference from the case of continuous arrival at CR is the CDVT (Cell Delay V) declared at the time of connection connection.
arbitration tolerance (cell delay fluctuation permissible value), within the range not exceeding,
It is assumed that the packets arrive continuously at the link speed in order from the first cell.

When CDVT is exceeded, PC
It is assumed that the vehicle arrives at the timing of continuous arrival at R. Here, the link speed actually indicates the physical link speed or the speed of the VP accommodating the connection, and indicates in what unit the opposite device (user, TA, switch) shapes the cell transmission rate. It depends.

Consider the arrival time of the n-th cell in this case. First, the number of cells that continuously arrive at the link speed is m
And This m is the link speed L, PCR and CDV
With respect to T, it can be obtained from the relationship that (m-1) × (1 / PCR-1 / L) ≦ CDVT and m × (1 / PCR-1 / L)> CDVT hold.

As a result, the first cell (1
Assuming that the monitoring time of the (th cell) is t (0), the subsequent n
The arrival times t (n) of the (th) (2 ≦ n ≦ K) cells are sequentially determined at intervals of m, and when 1 ≦ n ≦ m, t (n) = t (0) + (n−1)
× 1 / L, when m + 1 ≦ n ≦ 2m, t (n) = t (0) + mx
1 / PCR + (nm−1) × 1 / L, where 2m + 1 ≦ n ≦ 3m, t (n) = t (0) +2
m × 1 / PCR + (n−2m−1) × 1 / L.

That is, (i-1) × m + 1 ≦ n ≦ i ×
In the case of m, t (n) = t (0) + (i−1) × mx
1 / PCR + {n− (i−1) × m−1} × 1 / L (where i is a positive integer).

This is because only the head cell is monitored,
This means that the arrival information of the succeeding cell can be easily obtained. The method of identifying the first cell of the packet is the same as the method described in the first embodiment, and a description thereof will not be repeated.

As described above, according to the connection admission control method according to the second embodiment, the first embodiment
By adopting the same cell monitoring method and head cell identification method as described above, the same effect as in Embodiment 1 can be enjoyed, and the cells in one packet arrive without exceeding the declared CDVT. In this case, the cells are treated as K cells in the order of arrival from the top cell, and when the cells arrive beyond the CDVT, they are treated as the K cells sequentially arrived at the timing of continuous arrival by PCR. By the appearance of the arrival of a subsequent cell,
The arrival information of the succeeding cell can be obtained.

Embodiment 3 Next, a connection admission control method according to the third embodiment will be described. In the connection admission control method according to the third embodiment, when the number of cells constituting the maximum packet is declared at the time of a connection connection request, the number of cells is set as the above-described cell number K in the first or second embodiment. The cell arrival model is obtained by the procedure shown in FIG.

For example, GFR (Guaranteed)
In the case of a Frame Rate (Connection) connection, the parameter MFS (Maximum Frame Size)
Since the number of cells constituting the maximum packet is declared by e), this MFS value may be regarded as K. here,
GFR is defined as a frame, which is a collection of cells constituting an information unit on the upper layer of the ATM layer, and the ATM network discards a plurality of cells in units of frames when congestion occurs, thereby causing invalid cell transfer and accompanying cells. A traffic class that prevents loss.

As described above, according to the connection admission control method according to the third embodiment, the first embodiment
By adopting the same cell monitoring method, head cell identification method and cell arrival model as those of the second or the second embodiment, it is possible to enjoy the same effects as the first or the second embodiment, and as in the case where the traffic class is GFR. When the number of cells constituting the maximum packet is declared, the reported number of cells can be used as the number of cells constituting the packet.

Embodiment 4 Next, a connection admission control method according to the fourth embodiment will be described. In the connection admission control method according to the fourth embodiment, when the number of cells constituting one packet is not declared at the time of a connection connection request, the first or second embodiment uses the fixed number of cells previously determined. The cell arrival model is obtained by the procedure shown in FIG.

The procedure is as follows. First, the number of cells constituting one packet is fixedly given, and for a connection in which the number of cells constituting one packet is not declared, this fixed value is replaced by the number of cells K. Shall be applied as A unit for giving the number of cells is uniquely assigned to a connection, such as one for an ATM switch, one for a line, or one for a VP. For example, in the case of a CBR connection, since the number of cells is not declared, a fixed value is regarded as the number of cells K.

As described above, according to the connection admission control method according to the fourth embodiment, the first embodiment
By adopting the same cell monitoring method, head cell identification method, and cell arrival model as in the second or second embodiment, it is possible to enjoy the same effects as in the first or second embodiment, as well as in the case where the traffic class is CBR. In the case where the number of cells constituting the packet is not reported, the number of cells constituting the packet can be a fixed number given in advance.

Embodiment 5 Next, a connection admission control method according to the fifth embodiment will be described. In the connection admission control method according to the fifth embodiment, of the head cells of the packet, only the cell having the CLP bit of 0 is to be monitored, and the cell having the CLP bit of 1, that is, the cell whose priority is set to be low is to be monitored. It is characterized by being removed from.

FIG. 7 is an explanatory diagram for explaining a method of selecting a monitoring target cell by the connection admission control method according to the fifth embodiment. As shown in FIG.
PC (Network Parameter Cont)
roll), the traffic monitoring mechanism 124
Among the cells determined to be conforming cells, CLP =
The packet transmitted in step 1 and the packet in which the cell whose CLP = 1 due to the tagging function in the traffic monitoring mechanism 124 is the first cell are not considered in the CAC processing.

That is, when CLP = 0, the cell arrival model shown in the first or second embodiment is adopted,
The arrival information of the succeeding cell is obtained, and if CLP = 1, no processing is performed and the arrival of the next head cell is waited.

As described above, according to the connection admission control method according to the fifth embodiment, the first embodiment
By adopting the same cell monitoring method, head cell identification method, and cell arrival model as those in Embodiment 2 or 2, it is possible to enjoy the same effects as in Embodiment 1 or 2, and in the head cell identification, the CLP bit is used. By monitoring, cells set with low priority are excluded from the monitoring target, so that the cell monitoring load on the ATM exchange can be further reduced.

Embodiment 6 FIG. Next, a connection admission control method according to the sixth embodiment will be described. In the connection admission control method according to the sixth embodiment, the CL
It is characterized in that the head cell of every packet is monitored, regardless of the contents of the P bit. That is,
All packets whose head cell is a cell determined to be a conforming cell by the traffic monitoring mechanism 124 are CA
It is to be considered in the C processing.

Specifically, the logic for inspecting the required CLP bit in the fifth embodiment is not required,
Simply, when the first cell is received, the arrival information of the succeeding cell is obtained by the method described in the first or second embodiment.

As described above, according to the connection admission control method according to the sixth embodiment, the first embodiment
By adopting the same cell monitoring method, head cell identification method, and cell arrival model as those in Embodiment 2 or 2, it is possible to enjoy the same effects as in Embodiment 1 or 2, and in the head cell identification, the CLP bit is Irrespective of the indicated priority, the arrival information of the succeeding cell can be obtained.

Embodiment 7 FIG. Next, an ATM exchange according to a seventh embodiment will be described. In the ATM exchange according to the seventh embodiment, in order to execute the connection admission control methods described in the first to sixth embodiments, the first cell of the first cell shown in the first embodiment is replaced with the conventional arrival cell monitoring unit 125. It is characterized by including a head cell recognizing unit for performing recognition and scrutinizing CLP bits described in the fifth embodiment.

FIG. 8 is a block diagram showing a schematic configuration of the head cell recognizing unit of the ATM exchange according to the seventh embodiment. The arrival cell monitoring unit shown in FIG.
CLP decoding section 151 for extracting bits and ST for realizing head cell identification in AAL type 3/4
A decoding unit 152, a PTI decoding unit 153 for realizing head cell identification in AAL type 5, and an AN
D gates 161 to 163 and control unit interface 12
8 is provided.

In FIG. 8, first, ST decoding section 15
2 or the PTI decoding unit 153 identifies the head cell shown in the first embodiment for the arriving cell.
The P decoding unit 151 determines that C
A detection signal of LP = 0 is output.

For ALL type 5, the first embodiment
As described above, the arrival cells are sorted according to the end cell identification signal output from the PTI decoding unit 153, and the head cell indication flag 154 is turned on. Then, the state value of the head cell instruction flag 154 and the signal output from the PTI decoding unit 153 indicating a cell other than the tail cell are input to the AND gate 163, and the output result is input to one of the input terminals of the AND gate 162. input. The AND gate 162 is connected to the other of its input terminals.
The CLP detection signal output from the CLP decoding unit 151 is input, and the output result is input to the control unit interface 128 as a head cell identification result.

That is, when a cell arrives, the head cell recognizing unit first looks at the head cell indication flag 154 and turns it on.
If so, the control unit interface 128 is notified of the start cell reception. If only the cell with CLP bit = 0 is to be monitored, the CLP bit is examined at the same time, and only when CLP = 0, the first cell is notified of reception.

Although the head cell indication flag 154 exists for each connection, it is 1-bit information, and the other one exists for each line. In comparison, the hardware volume may be very small.

On the other hand, for ALL type 3/4, S
The head cell identification signal output from the T decoding unit 152 and the CLP detection signal output from the CLP decoding unit 151 are input to an AND gate 161, and the output result is input to the control unit interface 128 as a head cell identification result. If only the cell with the CLP bit = 0 is to be monitored, the CLP bit is inspected at the same time, and C
Only when LP = 0, the reception of the head cell is notified.

As described above, according to the ATM exchange according to the seventh embodiment, the conventional arrival cell monitoring unit 12
In place of the fifth embodiment, the connection reception described in the first to sixth embodiments is provided by including the first cell recognition unit that performs the recognition of the first cell described in the first embodiment and the scrutiny of the CLP bit described in the fifth embodiment. It is possible to realize a control method.

Embodiment 8 FIG. Next, a connection admission control method according to the eighth embodiment will be described. In the connection admission control method according to the eighth embodiment, in the first embodiment, a method for obtaining arrival information of a succeeding cell is described in the first embodiment.
In other words, the cell arrival model is a model having a strong burst property, and the average obtained by monitoring the section where cells arrive by PCR and the section where cells do not arrive is the MCR.

FIG. 9 is an explanatory diagram for explaining a cell arrival model employed in the connection admission control method according to the eighth embodiment. As shown in FIG. 9, K cells constituting a packet arrive by PCR (when the cell is a PC
The section arriving at R is called the ON section, and its section length is x
[Cell time]), followed by a section where no cell arrives (this is called an OFF section, and the section length is indicated by y [cell time]). This is repeated continuously so that the cell arrives at MCR on average.

Here, the relational expression between x and y is as follows: x = K / PCR, MCR = PCR × x / (x + y) From these two equations, y = K × (1 / MCR-1 / PCR) is derived.

When the cell arrival model shown in FIG. 9 is compared with the conventional cell arrival model shown in FIG. 17, both cells arrive at the MCR on average, but the cell arrival model in the eighth embodiment is Since the model has a strong burst property, the amount of reserved buffers differs greatly. That is, the model of the eighth embodiment having a strong burst property requires a large amount of buffer, and as a result, it is possible to suppress the possibility of discarding at the packet level, which is a concern in the conventional model.

In fact, if the arrival timing of the first cell of a packet does not exceed the cell arrival timing in the MCR, all packets are transferred even if a plurality of subsequent cells constituting each packet arrive by PCR. be able to. In the example shown in FIG. 9, the head cell of packet # 2 is
Since the packet arrives earlier than the cell arrival timing in the MCR, the packet # 2 is discarded as a result, but in the packet # 1, the leading cell does not exceed the cell arrival timing in the MCR.

Although the following cells also exceed the MCR,
Since the packet arrives at an interval not exceeding the PCR, the packet is normally transferred. Note that packet # 3 is normally transferred even though the head cell has arrived earlier than the cell arrival timing in the MCR because a part of the cells constituting the previous packet # 2 has been discarded. Will be.

Note that the cell arrival model described in the eighth embodiment is based on the above-mentioned document “Ultralight CAC in ATM”.
Proposed mechanism, Clumped Cell model that can be handled by SSE97-35, 1997 "(Fig. 1
9), the content of this document can be applied to the CAC algorithm itself.

As described above, according to the connection admission control method according to the eighth embodiment, the first embodiment
By adopting the same cell monitoring method and head cell identification method as described above, the same effect as in Embodiment 1 can be enjoyed, and at the same time, the cell arrival model for obtaining the arrival information of the succeeding cell has a burst characteristic. In consideration of a strong model, the section arriving by PCR and the section not arriving are monitored so that the average of the arriving cells becomes the MCR, so that more cells are held in a larger reserved buffer amount. This makes it possible to reduce discard at the packet level.

Embodiment 9 FIG. Next, a connection acceptance control method according to the ninth embodiment will be described. In the connection admission control method according to the ninth embodiment, as in the second embodiment, the plurality of cells constituting each packet are K cells, and the time difference from the case of continuous arrival by the declared PCR is: C declared at connection connection
If the DVT is not exceeded, it shall arrive continuously at the link speed from the first cell within the range not exceeding the DVT.
If it exceeds VT, it will arrive at the timing of continuous arrival by PCR. Embodiment 9
Is characterized in that the cell arrival model described in Embodiment 8 is further applied to this cell arrival model.

FIG. 10 is an explanatory diagram for explaining a cell arrival model in the connection admission control method according to the ninth embodiment. FIG. 11 is an explanatory diagram for explaining a cell arrival model approximated in the connection admission control method according to the ninth embodiment. First, FIG.
In the cell arrival model shown in FIG.
Consider the section length of section F.

Here, it is assumed that the number of cells that continuously arrive at the link speed is m. Note that the relationship between m and the link speed L, PCR, and CDVT is obtained from the equation shown in the second embodiment. Thus, the cell arrival model shown in FIG. 10 can be approximated as shown in FIG. That is, in the ON section, bursts of the number m of cells arriving at the link speed are consecutive K / at burst intervals CDVT.
Arrival m + 1. On the other hand, the OFF section is K × (1 / M
CR-1 / PCR) cell time.

As described above, according to the connection admission control method of the ninth embodiment, the second embodiment
In the case where the cell arrival model shown in the eighth embodiment is applied to the cell arrival model shown in FIG. 11, it can be approximated to a model having strong burst characteristics as shown in FIG. 11, and the same effect as in the eighth embodiment can be obtained. You can enjoy.

Embodiment 10 FIG. Next, Embodiment 10
Will be described.
In the connection admission control method according to the tenth embodiment, similarly to the third embodiment, when the number of cells constituting the maximum packet is declared at the time of a connection connection request, the number of cells is set to the above-described cell number K. The cell arrival model is obtained by the procedure described in the eighth or ninth embodiment.

For example, in the case of a GFR connection, the number of cells constituting the maximum packet is declared in the parameter MFS, and this MFS value may be regarded as K.

As described above, according to the connection admission control method according to the tenth embodiment, by adopting the same cell monitoring method, head cell identification method and cell arrival model as those of the first or second embodiment, The same effects as in Embodiment 1 or 2 can be enjoyed,
As shown in the eighth or ninth embodiment, even when the traffic class is GFR, the number of cells constituting the maximum packet is declared even for a model having strong burst characteristics as a cell arrival model. In addition, as the number of cells constituting a packet, the declared number of cells can be used, and the same effect as in the eighth or ninth embodiment can be obtained.

Embodiment 11 FIG. Next, Embodiment 11
Will be described.
In the connection admission control method according to the eleventh embodiment, similarly to the fourth embodiment, when the number of cells constituting one packet is not declared at the time of a connection connection request, the procedure shown in the eighth or ninth embodiment is used. , And a cell arrival model.

Note that the unit for giving the number of cells is one for the ATM exchange, one for the line, or
It is assumed that one connection number is assigned to P and is uniquely given to the connection. For example, UBR including MBR
In the case of a connection, the number of cells is not declared,
The fixed value is regarded as the cell number K.

As described above, according to the connection admission control method according to the eleventh embodiment, by adopting the same cell monitoring method, head cell identification method, and cell arrival model as those of the first or second embodiment, The same effects as in Embodiment 1 or 2 can be enjoyed,
As shown in the eighth or ninth embodiment, the number of cells constituting a packet is not declared even for a model having a strong burst property as a cell arrival model, as in the case where the traffic class is UBR including MBR. In this case, as the number of cells constituting the packet, a fixed number of cells can be used in advance, and the same effect as in the eighth or ninth embodiment can be obtained.

[0121]

As described above, according to the present invention, when a packet communication service is provided in the AAL or the like, in the non-parametric CAC, only the head cell of each packet is monitored, and
Since the succeeding cell arrives at the PCR declared at the time of connection connection and each packet appears to be composed of K cells, there is no need to monitor all the arriving cells, and the load imposed on the ATM exchange is reduced. This has the effect that it is possible to reduce the

According to the next invention, when a packet communication service is provided in the AAL or the like, in the non-parametric method CAC, only the head cell of each packet is monitored, and the subsequent cells are declared when the connection is established. If the cells arrive in the same PCR and the cells in one packet arrive without exceeding the declared CDVT, they are treated as a predetermined number of cells in the order of arrival from the first cell, and arrive in excess of the CDVT. In this case, since the cells are treated as a predetermined number of cells arriving in sequence at the timing of continuous arrival by the PCR, it is possible to obtain the arrival information of the subsequent cells even by such a simulated arrival of the subsequent cells. It is no longer necessary to monitor incoming cells, and the load on the ATM switch can be reduced. The effect say.

According to the next invention, when the number of cells constituting the maximum packet is declared, such as when the traffic class is GFR at the time of the connection connection request, the number of cells constituting the packet is set as the number of cells constituting the packet. There is an effect that the declared number of cells can be used.

According to the next invention, when the number of cells constituting one packet is not declared, such as when the traffic class is CBR at the time of connection connection request, the number of cells constituting the packet is fixed in advance. There is an effect that the number of cells given in a specific manner can be used.

According to the next invention, only the cell whose CLP bit is 0 among the head cells of the packet is to be monitored,
Since cells with a CLP bit of 1, that is, cells with a low priority set are excluded from monitoring targets, the number of cells to be monitored can be reduced, and the load on the ATM switch can be further reduced. To play.

According to the next invention, there is an effect that the head cell of all packets can be monitored regardless of the contents of the CLP bit.

According to the next invention, the average of the arrival cells is M
Since the section arriving and the section not arriving by PCR are monitored so as to be CR, a model having a strong burst property can be adopted as a cell arrival model for obtaining arrival information of a subsequent cell. Thus, more cells can be held in the larger buffer amount, and this has the effect of reducing discard at the packet level.

According to the next invention, when a packet communication service is provided in the AAL or the like, in the parametric method CAC, the cell to be monitored is limited to only the first cell of each packet, and the subsequent cell is declared at the time of connection connection. If a cell in one packet arrives without exceeding the declared CDVT, it is treated as a predetermined number of cells in the order of arrival from the head cell, and arrives beyond the CDVT. In this case, the cell is treated as a predetermined number of cells arriving in sequence at the timing of continuous arrival by PCR, and thereafter, the section arriving by PCR and the section not arriving are monitored so that the average of the arriving cells becomes MCR. Therefore, it is possible to obtain the arrival information of the succeeding cell even by the appearance of the arrival of the succeeding cell. This eliminates the need to monitor all arriving cells, and allows the use of a model with strong burst characteristics as a cell arrival model for obtaining the arrival information of subsequent cells. A large number of cells can be held, and this has the effect of enabling discard at the packet level to be reduced.

According to the next invention, in the CAC of the parametric method, when the number of cells constituting the maximum packet is declared, such as when the traffic class is GFR at the time of connection connection request, There is an effect that the declared cell number can be used as the cell number.

According to the next invention, in the CAC of the parametric method, when the number of cells constituting one packet is not declared, such as when the traffic class is UBR including MBR at the time of connection connection request, There is an effect that the number of cells fixedly provided in advance can be used as the number of cells constituting the packet.

According to the next invention, for example, instead of the conventional arriving cell monitoring unit in the line corresponding unit, a head cell recognition unit for recognizing the head cell and scrutinizing the CLP bit is provided. There is an effect that it is possible to provide an ATM exchange that can execute such a connection admission control method.

[Brief description of the drawings]

FIG. 1 is a diagram for explaining a cell monitoring method of a connection admission control method according to a first embodiment.

FIG. 2 is a diagram for explaining handling of cells other than a head cell in the connection admission control method according to the first embodiment;

FIG. 3 is a diagram for explaining the contents of a PTI field in AAL type 5 in the connection admission control method according to the first embodiment;

FIG. 4 is a diagram for explaining a method of identifying a head cell in AAL type 5 in the connection admission control method according to the first embodiment;

FIG. 5 is a diagram for explaining the contents of an ST field in AAL type 3/4 in the connection admission control method according to the first embodiment;

FIG. 6 is a diagram for explaining a cell arrival model in the connection admission control method according to the second embodiment.

FIG. 7 is a diagram for explaining a method of selecting a monitoring target cell by a connection admission control method according to a fifth embodiment.

FIG. 8 is a block diagram illustrating a schematic configuration of a head cell recognizing unit of an ATM exchange according to a seventh embodiment;

FIG. 9 is a diagram for explaining a cell arrival model employed in the connection admission control method according to the eighth embodiment.

FIG. 10 is a diagram for explaining a cell arrival model in a connection admission control method according to a ninth embodiment.

FIG. 11 is a diagram for explaining a cell arrival model approximated in the connection admission control method according to the ninth embodiment.

FIG. 12 is a schematic diagram showing a configuration of a conventional ATM network.

FIG. 13 is a block diagram showing a schematic configuration of a conventional ATM exchange.

FIG. 14 is a diagram illustrating an operation of a dynamic CAC.

FIG. 15 is a block diagram showing a schematic configuration of a conventional line handling unit.

FIG. 16 is a schematic configuration diagram of a conventional arrival cell monitoring unit.

FIG. 17: MC in CAC by parametric method
It is a figure for explaining R guarantee.

FIG. 18 is a diagram for explaining a state of discarding at a packet level when a packet is composed of three cells.

FIG. 19 is a diagram for explaining a Clumped Cell model.

[Explanation of symbols]

100 ATM network, 101 ATM terminal, 1
02 Non-ATM terminal, 103 Terminal adapter, 1
10 ATM switch, 120 line corresponding unit, 121 O
/ E part, 122 E / O part, 123 physical layer processing part, 124 traffic monitoring mechanism, 125 arrival cell monitoring part, 126 window control, 127 counter, 12
8 control section interface, 130 switch section, 14
0 control unit, 150 management device, 151 CLP decoding unit, 152 ST decoding unit, 153 PTI decoding unit, 154 head cell indication flag, 161-163
AND gate.

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Inventor Riki Ichihashi 2-3-2 Marunouchi, Chiyoda-ku, Tokyo Mitsubishi Electric Corporation F-term (reference) 5K030 HA10 HB29 LC06 MB06 MB09 MB12

Claims (11)

[Claims] 1. An ATM (Asy) for performing packet communication.
In an ATM network accommodating nchronous transfer mode terminals, connection admission control (CA) is performed when a new connection connection request is issued from a user or a network administrator.
C: Connection Admission Co
ntrol), the number of cells constituting one packet is set to a predetermined value, and only the first cell of the packet transferred through the ATM network is set. And for a subsequent cell of the packet,
Declared maximum cell rate (PCR: Peak Cell)
(1 Rate), the connection admission control method. 2. In an ATM network accommodating an ATM terminal performing packet communication, a connection admission control to be performed when a new connection connection request is issued from a user or a network administrator, and a non-parametric method for making an admission judgment based on an actual traffic measurement result. When performing the method, the number of cells constituting one packet is set to a predetermined value, and only the first cell of the packet transferred in the ATM network is monitored.
Declared cell delay fluctuation tolerance (CDVT: Cell)
Delay Variation Toleranc
e) a connection admission control method characterized in that it is handled as continuously arriving within a range satisfying e). 3. The method according to claim 1, wherein, when a connection connection request is issued, when the number of cells of the maximum packet transferred on the connection is declared by a user or a network administrator, the number of cells is set to the predetermined value. The connection acceptance control method according to claim 1. 4. The method according to claim 1, wherein a predetermined number of cells is set to the predetermined value when a user or a network administrator does not declare the number of cells of a packet transferred on the connection at the time of a connection connection request. The connection acceptance control method according to claim 1. 5. A CLP (Cell Loss Prior) included in a head cell of a packet among the head cells of the packet.
5. The connection admission control method according to claim 3, wherein only the first cell whose bit is 0 is monitored, and the first cell whose CLP bit is 1 is not monitored. 6. A head cell in which the CLP bit included in the head cell of the packet is 0,
5. The connection admission control method according to claim 3, wherein both of the head cells whose LP bits are 1 are to be monitored. 7. In an ATM network accommodating an ATM terminal that performs packet communication, a connection admission control performed when a new connection connection request is issued from a user or a network administrator, based on a declared traffic parameter, a parametric judgment. In this case, the number of cells constituting one packet is set to a predetermined value, and a minimum guaranteed bandwidth (MCR: M
In the case where the minimum cell rate is included, the cells constituting one packet are treated as if they arrived at the declared PCR, and the average cell arrival model required in the parametric method is used as the cell arrival model. Cell arrival rate (SCR: Sust
A connection admission control method, which employs a model having a cell non-receiving section in which an enable cell rate is the MCR. 8. In a ATM network accommodating an ATM terminal for performing packet communication, a connection admission control to be performed when a new connection connection request is issued from a user or a network administrator, based on a declared traffic parameter, a parametric judgment. In the case where the method is performed, the number of cells constituting one packet is set to a predetermined value, and when the traffic parameter includes MCR, each of the cells constituting one packet is declared. It is treated as continuously arriving within a range satisfying the CDVT, and a model having a cell non-receiving section such that an SCR becomes the MCR is adopted as a cell arrival model required in the parametric method. Connection admission control Law. 9. The method according to claim 1, wherein, when a connection connection request is issued, when the number of cells of the maximum packet transferred on the connection is declared by a user or a network administrator, the number of cells is set to the predetermined value. Item 7
Or the connection admission control method according to 8. 10. The method according to claim 1, wherein a predetermined number of cells is set to the predetermined value when a user or a network administrator does not declare the number of cells of a packet transferred on the connection at the time of a connection connection request. The connection admission control method according to claim 7. 11. A head cell recognizing unit for recognizing a head cell and scrutinizing a CLP bit to execute the connection admission control method according to at least one of claims 1 to 6. ATM switch.