JP3433916B2 - ATM virtual path capacity setting method and ATM virtual path capacity setting apparatus - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an asynchronous transfer mode (ATM) network.
The capacity of the ATM virtual path set between the TM exchanges is
The present invention relates to a technology for designing a virtual path so as to satisfy a predetermined cell loss rate target value, and in particular, efficiently sets an ATM virtual path capacity that satisfies a predetermined cell loss rate target value (CLR). The present invention relates to an ATM virtual path capacity setting method and an ATM virtual path capacity setting apparatus suitable for performing the above.

[0002]

2. Description of the Related Art Research and development of ATMs as information transmission techniques for providing multimedia services have been actively pursued. ATM divides various kinds of information into short fixed-length blocks with headers called "cells", and multiplexes the blocks in units of an efficient statistical multiplexing system.

A cell is a concept of a time slot, but does not appear in a time-period manner, but is dynamically assigned in response to a demand for information extraction that fluctuates over time. By changing the number of cells, The communication speed can be set variably. Therefore,
ATM is a transfer mode in which all digital information from voice and data to images can be transmitted in a unified manner. The communication network that performs communication based on this ATM is AT
It is called M network.

[0004] In an ATM network, it is possible to logically set a band usable in two exchange periods as an ATM virtual path. Also, under the limitation of the capacity of the ATM virtual path,
An unspecified number of information sources can logically set a virtual circuit sharing the ATM virtual path.

The configuration of a transmission line network using an ATM virtual path is described in, for example, "Structure of High-speed Burst Multiplexing Transmission System" by Sato, Kaneda and Tokizawa (Information of IEICE Information Network Research Group, IN87) −
84, 1987).

[0006] A functional configuration of a conventional ATM switch related to setting of the capacity of an ATM virtual path will be described below with reference to the drawings. FIG. 11 shows an ATM in a conventional ATM exchange.
FIG. 3 is a block diagram illustrating a configuration example of a virtual path capacity setting mechanism.

First, the physical functional configuration will be described. A
The cell transmission device 100a of the TM exchange is provided with a cell transmission interface (denoted by "IF" in the figure) 101 for each physical transmission medium 200 as a transmission destination, and transmits cells exceeding the capacity of the transmission medium 200. Is to be suppressed.

[0008] The logical functional configuration will be described. ATM virtual paths (described as "virtual paths" in the figure) 201, 202,
... Each shaper (described as “S” in the figure) 111, 1
Are connected to each other.

The shapers 111, 112,...
This suppresses cell transmission exceeding the capacity logically determined for the virtual paths 201, 202,.... Each cell transmission interface 101 and shaper 111,
Between 112,..., Cells are transmitted at timing according to a certain scheduling rule.

The shapers and the ATM virtual paths have a one-to-one correspondence. For example, the shapers 111 and 112 respectively correspond to the ATM virtual paths 201, 202,.

Here, in order to make the explanation easy to understand, the capacity of each ATM virtual path is determined by the shapers 111 and 11.
It is assumed that the cell transmission speed is the same as the cell transmission speed of 2,. Are connected to the shapers 111, 112,..., Respectively, and the cell transmission waiting buffers 121, 122,.
In this example, when a cell exceeding the ATM virtual path capacity flows in, the cells are held up to the size of the cell transmission waiting buffers 121, 122,... And the next cell transmission timing is waited.

Next, along with the flow of cells, cell transmission waiting buffers 121, 122,...
The shaper 111, 112,..., The cell transmission interface 101 will be described. Cells switched according to the transmission medium 200 in the ATM exchange are routed to the cell transmission device 100a. Logically, AT
The cells are divided for each destination of M virtual paths 201, 202,..., And the cell flows 301, 302,
.. Are added to the cell transmission device 100a.

For example, the cell transmission waiting buffer 121 to which the cell stream 301 is added counts the number of cells arriving during the measurement time period T. Then, the cell transmission waiting buffer 12
1 is processed in the order of arrival, and the shaper 111
If is running, it is stored in the cell transmission waiting buffer 121 and waits for the turn of processing. The number of cells waiting for transmission is referred to as Q length.

The order of cell processing in the cell transmission waiting buffer 121 is performed according to a scheduling rule defined for the shaper 111. When the order of the processing reaches the shaper 111, the cells stored in the cell transmission waiting buffer 121 pass through the shaper 111, advance to the cell transmission interface 101, and are transmitted to the transmission medium 200.

Next, a conventional ATM virtual path capacity setting procedure in the above-described ATM exchange will be described as first to fourth prior arts. "First Conventional Technique" First, as a first conventional technique, a technique of modeling a cell arrival process by a Poisson process will be described.

This technique is characterized in that it is possible to easily determine the ATM virtual path capacity by using only one traffic measurement item such as the arrival rate of cells arriving at the ATM virtual path. That is, in FIG.
21 is b and the cell loss rate target value is CLR, the cell flow 3 of the arrival rate λ to the ATM virtual path 201
01 that satisfies the cell loss rate target value CLR
The M virtual path capacity C is calculated by the following equation (Equation 1).

[0017]

(Equation 1)

In this equation, M is a constant for converting the number of cells per unit time into a capacity.

This equation (Equation 1) is derived based on the following points. That is, the Q length distribution S and the Q length are K
If the longer probability is P [S> K], it is known that the following holds according to the principle of large deviation. That is, for η satisfying the following equation (2), the equation (3) is further satisfied.

[0020]

(Equation 2)

(Equation 3)

However, A (t) in “Equation 2” represents the number of cells arriving before time t.

In Equation 3, when the Q length K on the left side matches the buffer size b, the following Equation 4 is considered, the cell arrival process is regarded as Poisson process, and C is obtained from η. Is the equation of the above “Equation 1”.

[0023]

(Equation 4)

FIG. 12 shows a functional configuration of an ATM exchange for realizing the first prior art, and its description will be given.

FIG. 12 is a block diagram showing an example of the functional configuration of an ATM exchange for realizing the first prior art.

In the following description, the same components as those already described with reference to FIG. 11 are denoted by the same reference numerals, and description thereof will be omitted.

In the ATM exchange shown in the figure, soft counters 131, 132,.
, 122,..., When cells are logically allocated to the ATM virtual paths 201, 202,.

The setting device 400a summarizes the counting results of the soft counters 131, 132,... Inputted via the communication line 401a at every predetermined period T, and sets the arrival rate, which is the number of arriving cells per unit time. λ is calculated by the following equation (5).

[0029]

(Equation 5)

However, in the first prior art, there is no guarantee that modeling of the cell flow by the Poisson process is sufficiently valid. That is, the burstiness of traffic is not small enough to be modeled in the Poisson process.

Reports such as a large burst property regarding ATM traffic are reported in, for example, WELeland, MSTaqqu, WW
`` On the self-simil '' by illianger, and DVWilson
er nature of Ethernet traffic (extended version) '' I
EEE / ACM Trans.Networking, vol.2, no.1, pp.1-15, 1994.
Familiar with. Therefore, it is dangerous to set the capacity of the ATM virtual path of the currently operating ATM network according to the first conventional technique.

[Second Prior Art] Next, as a second prior art, a technique of collecting all the arrival times of cells and restoring the arrival process to obtain (set) an ATM virtual path capacity with no excess or deficiency. explain.

FIG. 13 is a block diagram showing an example of the functional configuration of an ATM exchange for realizing the second conventional technique.
In this ATM exchange, the cell transmission interface 101
The flowing cell is copied (captured) or a header part thereof is copied by an external device 500 provided between the communication medium 200 and the transmission medium 200, and the copied data is transferred via a communication line 501. Send to storage device 510.

When a cell arrives, the storage device 510 stores the time stamp of the arrival time and the cell header. The header of this cell includes an identifier for specifying the ATM virtual path. The storage device 510 classifies these time stamps for each ATM virtual path, and determines the cell arrival process for each ATM virtual path. analyse. Examples of this analysis technique include a technique for creating a probability distribution of an arrival process, and a technique for obtaining the average and variance of the arrival intervals, the third moment, and the like.

However, the storage device 510 capable of giving such a time stamp is expensive. The main cause is
This is due to the technical difficulty of time-stamping in sub-microsecond units required to analyze incoming cells. Further, since continuous capture can be performed only while the amount of memory mounted in the storage device 510 permits (capturing for several seconds is performed every few minutes), setting accuracy other than the capturing interval cannot be guaranteed.

Also, depending on the specifications of the external device 500,
Some devices cannot capture only the header information added by the communication provider, and the payload carrying the customer's communication information may be sent to the outside.

Further, communication disconnection when the external device 500 is connected between the cell transmission interface 101 and the transmission medium 200 is inevitable. That is, in this technology, ATM
Due to the setting of the virtual path capacity, the communication of the customer is hindered.

Since the cell flow that can be captured is a cell flow prepared by the shapers 111, 112,..., The burst characteristics will be small, and the cell transmission waiting buffers 121, 122,.・ ・ Cells with cell loss in are not captured. This has a small effect on low-load operation, but has a large effect on accuracy reduction in setting the ATM virtual path capacity during high-load operation. Therefore, setting the ATM virtual path capacity of the ATM network in operation based on the second conventional technique is prohibited and cannot be executed.

"Third Prior Art" Next, the third prior art will be described. Before that, the Q length distribution will be described with reference to FIG.

FIG. 14 is an explanatory diagram showing an example of the Q length distribution. This Q length distribution is created from traffic data actually measured by a technique similar to the second conventional technique.

The Q length distribution refers to the cell transmission waiting buffers 121, 122,
.. Is a probability distribution of the number of cells (Q length) stored in the cell, and one broken line in FIG. 14 indicates a Q length distribution based on arrival time data of about 130,000 cell traffics. Is drawn. The horizontal axis is the number of cells in the buffer (Q length)
k, the vertical axis represents the probability P [S that cells equal to or larger than the cell number are stored in the cell transmission waiting buffers 121, 122,.
> K] in logarithm.

In a range smaller than the reciprocal of the number of digits of the measured cell number, the Q length distribution has no meaning. For example, if traffic data captures 130,000 cells, 1
Probabilities below 0 ^-4 are not significant. Further, in the attenuation rate of the Q length distribution, the tail indicates that the Q length k is relatively long.

In general, it is known that, in a stochastic process that satisfies stationarity, rarity, and independence, the occurrence probability of an event follows an exponential distribution ("the Poisson's law of a few").
This means that the tail of the Q length distribution is approximated by a straight line in the case of the system configuration of FIG. Here, the Q length threshold value and the number of times the Q length threshold value is exceeded will be described.

To grasp the entire distribution of the Q length involves technical difficulties as described in the second prior art. Therefore, in order to partially understand the characteristics of the Q length, a mark (Q length threshold) is put in the middle of the buffers arranged in one line, and the Q length is set to the Q length threshold every time a cell arrives. There is a technique of determining whether or not the number exceeds the threshold value and counting the result (the number of times the Q length threshold value is exceeded). A value obtained by normalizing the number of times the Q-length threshold has been exceeded by the number of arrival cells is the Q-length threshold excess frequency.

The third prior art is characterized in that, in the above-described technique, two Q length thresholds are used, and the Q length distribution is determined by a straight line determined from two Q length threshold excess frequencies obtained therefrom. Approximate the tail of the attenuation. The solid line shown in FIG. 14 indicates that the Q length distribution is approximated by a straight line passing through the Q length threshold value at two points and the corresponding Q length threshold value exceeding frequency.

Hereinafter, a case where two Q length threshold values are set in each cell transmission waiting buffer 121, 122,... Of the cell transmission device 100b will be described with reference to the functional configuration diagram of FIG.

FIG. 15 is a block diagram showing an example of the functional configuration of the ATM exchange when two Q length thresholds are set. Each of the cell transmission waiting buffers 121, 122,... In the cell transmission device 100b has a first Q length threshold 151A, a second Q length threshold 151B, and a first Q length threshold 152A, respectively. And a second Q length threshold 152B,
Are set respectively, and the cell flows 301 and 30 are set.
It is determined whether the Q length exceeds the first and second Q length thresholds when 2 arrives.

Of the determination results, the results in which the Q length exceeds the first Q length threshold value and the results in which the Q length exceeds the second Q length threshold value are the same as those for counting the number of arriving cells. Counting is performed every cycle T. Then, the result of the counting is the number of times the first and second Q length thresholds have been exceeded. FIG. 16 shows a table in which only the data of the portion related to the ATM virtual path 201 among the counting results in the cell transmission device 100b is summarized.

FIG. 16 is an explanatory diagram showing an example of the result of counting cells by the cell transmitting apparatus in FIG. The data counted in the cell transmission device 100b in FIG. 15 is transmitted from the cell transmission device 100b to the virtual path capacity setting device 600 via the communication line 601. The data is counted as the table 1100 shown in FIG. Is accumulated in

In this table 1100, 111
Reference numerals 0, 1120, 1130, and 1140 denote records each indicating a column of a counting cycle number, the number of arrival cells, the number of times the first Q-length threshold is exceeded, and the number of times the second Q-length threshold is exceeded.
.. Indicate records of counting cycle numbers, and 1121, 1122,.
, 1113,..., And 1131, 1132,.
, The records of the number of times exceeding the first Q length threshold value corresponding to 1,1112,..., And the second Q lengths corresponding to the counting cycle numbers 1111, 1112,. This is a record of the number of times the threshold has been exceeded.

The virtual path capacity setting device 600 shown in FIG.
The data in the table 1100 in FIG.
The set capacity for the virtual path 201 is calculated. In addition, Shioda, Toyoizumi, Yokoi, Tsuchiya, Saito,
“Self-sizingnetwork: a new network concept based o
n autonomous VP bandwidth adjustment, "Proc. of ITC
15, pp. 997-1006, using the calculation formula described in 1997,
This is performed as follows.

First, a record sequence 1120 of the number of arrival cells
Using {di}, the ATM virtual path usage rate {ρi} in period i is calculated by the following equation (6).

[0053]

(Equation 6)

Here, M is a constant for converting the number of cells per unit time into a capacity, and C is a capacity set in the ATM virtual path 201. Further, using the record sequence 1130 {q _1i } of the first Q-length threshold excess count and the record sequence 1140 {q _2i } of the second Q-length threshold excess count, the following “Equation 7” is obtained. According to the formula 8, the first threshold excess frequency {p _1i } and the second threshold excess frequency {p
_2i ｝ is calculated.

[0055]

(Equation 7)

(Equation 8)

Further, the Q-length distribution attenuation rate {δ _i } in the period i is calculated by the following equation (9). here,
“S ₁ ” is the first Q length threshold 151A, and “s ₂ ”
Is the second Q-length threshold 151B.

[0057]

(Equation 9)

[0058] Further, the intercept constant beta _i in period i is calculated by the following "number 10" type.

[0059]

(Equation 10)

Next, the setting capacity C _i in period i is calculated by the following "number 11" type.

[0061]

(Equation 11)

" _{M 1i} " and "m _2i " in this equation (Equation 11) are as shown in the following equations (Equation 12) and (Equation 13), respectively, and T is the length of the counting cycle, b Is a buffer size, and CLR is a target value of a cell loss rate.

[0063]

(Equation 12)

(Equation 13)

Using the above calculation results, the set value C _d of the ATM virtual path capacity to be obtained is calculated by the following “Equation 14” and “Equation 15”.
It is calculated by the formula.

[0065]

[Equation 14]

(Equation 15)

Here, α is a constant satisfying 0 ≦ α ≦ 1. However, in the third prior art, a special ATM switch which can set the Q length threshold value must be used. Further, the Q length threshold must be set in hardware. Therefore, changing the threshold is
Very difficult. Also, it is not easy to appropriately select a plurality of Q length thresholds.

For example, if a relatively large value is set as the Q length threshold value, the Q length does not extend to that point, and the Q length threshold value exceeding frequency may not be collected. If a relatively small value is set as the Q-length threshold value, the portion of the Q-length distribution that suddenly drops unlike the tail may be measured, and an approximate straight line on the dangerous side may be obtained. The approximation accuracy of a straight line increases as the interval between the two Q length thresholds increases, but
Due to the above-mentioned problem, it is difficult to find an appropriate interval.

[Fourth Prior Art] Next, as a fourth prior art, the number of cells arriving in a unit time in a cell transmission waiting buffer associated with an ATM virtual path is counted, and the cell arrival process is turned on / off. The ATM virtual path capacity that satisfies the target cell loss rate is set by modeling through the process and estimating the model parameters from the average and variance of the number of arriving cells and the attenuation rate of the complementary distribution of the number of cells arriving per unit time. The technology will be described.

FIG. 17 is a block diagram showing an example of the functional configuration of an ATM switch for realizing the fourth prior art.
The ATM exchange in FIG. 17 has a cell transmission device 100 similarly to each of the ATM exchanges shown in FIGS. 11 to 13 and 15, and the cell transmission device 100 is provided for each physical transmission medium 200 as a transmission destination. Cell transmission interface (in the figure,
101) is provided, and the transmission medium 2
Cell transmission exceeding the capacity of 00 is suppressed.

Each cell transmission waiting buffer 121, 12
The soft counters 131, 132,.
Are provided along with the soft counters 131 and 13.
ATM logical paths 201,
When the cell flows 301 and 302 are sorted into 02,..., The number of arriving cells for each ATM virtual path is counted. The count results of the soft counters 131, 132,...
Sent to 0.

This ATM virtual path capacity setting device 1500
Has a cell counter 151, a table 152, a threshold register 153, an excess number counter 154, a table 155, a frequency calculator 156, and a set capacity calculator 157. The soft counters 131, 132,. Based on the counting result,
The ATM virtual path capacity that satisfies the target cell loss rate is set for the ATM virtual path terminating at the ATM exchange.

That is, the soft counters 131 and 13
Based on the counting results of 2, 2,...
, Are counted in the table 152, and based on the registered contents of the table 152, the excess number counting section 154 causes the threshold registration section 15 to register.
The number of times (exceeding number) exceeding the product of the number of cells arriving per unit time (arriving cell number threshold value) and the cycle T (excess number of times) set in advance in 3 is counted for each cycle, and the counting result is stored in a table 155. And the frequency calculation unit 156 determines the frequency of the number of times of excess based on the registered contents of the table 155.

Then, based on the frequency calculated by the frequency calculation unit 156 and the number of arriving cells calculated by the cell counting unit 151, the set capacity calculation unit 157 performs processing described below in detail, and the cell transmission waiting buffer 121 , 122,... Are efficiently obtained, and the set capacity of the ATM virtual path that satisfies the predetermined cell loss rate target value is obtained with high speed and high accuracy.

FIG. 18 is an explanatory diagram showing a configuration example of a table for registering the counting result of the cell counting unit in FIG. 17, and FIG. 19 is a configuration of a table for registering the counting result of the excess number counting unit in FIG. It is explanatory drawing which shows an example. The table 152 shown in FIG. 18 is a table in which only the data related to the ATM virtual path 201 among the count results of the cell transmission device 100 of the ATM exchange in FIG.

In this table 152, 1210,
Reference numerals 1220 and 1212 denote records each indicating a column of a counting cycle (counting cycle number) and a column of the number of arrival cells.
Are records indicating the number of the counting cycle, 122
Are counting cycle numbers 1211, 121.
It is a record of the number of arrival cells corresponding to 2,. Here, the counting cycle number 1211 indicates the immediately preceding counting cycle,
The counting cycle number 1212 indicates the counting cycle immediately before the counting cycle number 1211.

In the table 155 shown in FIG.
310 and 1320 are 1 in the table 152 of FIG.
Similarly to 210 and 1220, each is a record showing a column of the counting cycle number (counting cycle number) and the number of arrival cells, respectively.
Reference numeral 330 denotes a record of the number of times that the number of arriving cells exceeds the threshold value in each cycle number (the number of times the arrival threshold has been exceeded). Reference numeral 1340 denotes the number of arriving cells in units of a cycle kT which is k (integer) times the cycle T. The record 1350 is a record of the number of times the arrival threshold has been exceeded in units of kT.

In the ATM virtual path capacity setting device 1500 shown in FIG. 17, the data of the table 152 is processed by the excess number counting section 154 as follows.
55, and the set capacity for the ATM virtual path 201 is calculated by the frequency calculation section 156 and the set capacity calculation section 157. For the sake of simplicity, it is assumed that there is no missing data in the following description.

A predetermined past data use cycle T ′
(For example, one hour), a predetermined counting cycle T (for example, 20
If the measurement result in (seconds) is obtained, the total count number n in this case is 180.

[0079] number of times of excess counting unit 154, a product of the previously registered arrival cell count threshold to the threshold register 153 and the measurement period T, the record sequence number arriving cells in the table 152 1220 {d _i} (I = 1,..., N), it is determined whether the number of arrival cells exceeds the threshold, and the number of times of exceeding is determined by the record sequence 133 in the table 155 of FIG.
Register in 0 {q _i }. Here, q _i takes 1 or 0.

Based on the registered contents of this table 155,
The frequency calculation unit 156 of FIG. 17 calculates the frequency P of exceeding the threshold value of the number of arrival cells in the cycle T based on the following “Equation 16”.

[0081]

(Equation 16)

The excess number counting section 154 obtains the number of arrival cells {d _ki } in a cycle kT that is an integer k times the counting cycle T based on the following equation (17), and obtains the table 155 in FIG.
Is registered in the record column 1340 of the number of arrival cells in.

[0083]

[Equation 17]

Further, the excess number counting section 154 compares the product of the threshold value of the number of arriving cells and the period kT with the record sequence {d _ki } of the number of arriving cells to determine the number of arriving cells in the period kT. The threshold excess number {q _ki } is obtained, and the record column 1350 of the arrival threshold excess number in the table 155 in FIG.
Register with.

Based on the registered contents of this table 155,
The frequency calculation unit 156 of FIG. 17 calculates the threshold cell excess frequency P _{k of the} number of arrival cells of the period kT based on the following “Equation 18”.

[0086]

(Equation 18)

P and P _k are calculated for each counting cycle. P is calculated from data for the past n (= T ′ / T) cycles, and P _k is calculated for the past (n−
k + 1) It is calculated from the data for the cycle. The average cell rate λ within the time T in the past data use cycle T ′
Is calculated by the following equation (19).

[0088]

[Equation 19]

The variance V of the number of arriving cells is expressed by the following “Equation 2”.
0 ”formula.

[0090]

(Equation 20)

Next, the approximation to the ON-OFF process will be described. Now, let ψ be the peak cell rate of the past data use period T ′ in the ON-OFF process. If the reciprocal of the peak cell rate is called a time slot, the transition probability from the OFF state to the ON state between time slots is a, and the transition probability from the ON state to the OFF state is d. Here, the ON state refers to a state in which a cell exists in a time slot, and the OFF state refers to a state in which a cell does not exist in a time slot.

The transition probabilities a and d and the peak cell rate ψ are obtained by the following equations (Equations 21 to 23).

[0093]

(Equation 21)

(Equation 22)

(Equation 23)

In the equation (23), γ is the threshold value of the number of arriving cells, and μ (θ) satisfies the following equation (24).

[0095]

(Equation 24)

Incidentally, the calculation formulas of “Equation 23” and “Equation 24” are based on “Exponential bounds for queue” by NGDuffield.
ues with Markovian arrivals ", Queueing Systems17, p
pp. 413-430 (1994).

Under the approximation of the ON-OFF process, the target cell loss rate CLR is approximated by the following equation (25).

[0098]

(Equation 25)

Here, x satisfies the following equation (26).

[0100]

(Equation 26)

The value y that satisfies the expressions 25 and 26 is obtained, and the ATM virtual path capacity C _d that satisfies the target cell loss rate CLR is calculated from the following expression 27. You.

[0102]

[Equation 27]

Here, M is a constant for converting the number of cells per unit time into a capacity. The estimated cell loss ratio CLR _e is
For a, d, and 算出 calculated from the above equations 21 to 23, x and y are calculated from equation 26 and the following equation 28. It is calculated by substituting into the right side of “Equation 25”.

[0104]

[Equation 28]

Note that γ ′ × M (Equation 27) is the current ATM virtual path capacity. However, in such a fourth conventional technique, the accuracy of the arrival cell number threshold value exceeding frequency P is at most accurate.
{1 / (T ′ / T)}. The same can be said for the arrival cell number threshold excess frequency P _k of the period kT. Further, even if a different value is set as the threshold value, the arrival cell number threshold value exceeding frequency P may not change.

That is, the accuracy of the frequency of exceeding the threshold value of the number of arrival cells and the frequency of exceeding the threshold value of the number of arrival cells at period kT greatly depend on the threshold value. It is extremely difficult to set in advance a threshold value that can obtain the frequency of exceeding the threshold value of the number of arrival cells and the frequency of exceeding the threshold value of the number of arrival cells with period kT with high accuracy.

The features of the above-described conventional technology relating to the setting of the ATM virtual path capacity are summarized as follows.

According to the first prior art, the ATM exchange can determine the parameters of the ATM virtual path capacity calculation formula with simple traffic measurement items, but the ATM cell traffic is expected to have only the fluctuation of the Poisson process. However, there is a problem that the burstiness of actual cell arrival cannot be considered.

According to the second prior art, it is possible to accurately calculate the ATM virtual path capacity by capturing cell traffic data by an external device, but it requires an expensive external device and requires a customer to disconnect the communication, which is complicated. There is a problem that the traffic must be captured by operating an external device and the data must be analyzed, and the analysis result can be applied only for a short period of time during which the capture can be performed.

According to the third prior art, it is possible to obtain a straight line that approximates the tail of the attenuation rate of the Q length distribution relatively accurately from the two Q length threshold excess frequencies, and the ATM is thereby used.
The calculation of the virtual path capacity is possible. However, this technique cannot be realized unless a special ATM switch in which a Q length threshold can be set. Further, it is difficult to select the Q length threshold value, and since the Q length threshold value is set in hardware, it is very difficult to change the Q length threshold value when the setting is inappropriate. There is a problem.

In the fourth prior art, since the necessary count data is only the number of arriving cells, no special device is required.
Further, since the cell arrival process is modeled by a very bursty model called an ON-OFF process, the model can sufficiently express the burstiness of the cell arrival process. However, the accuracy in calculating the frequency of exceeding the threshold value of the number of arrival cells required for estimating the parameters in the ON-OFF process and the frequency of exceeding the threshold value of the number of arrival cells in the period kT is not sufficient. Unless is set, there is a problem that the accuracy of the frequency of exceeding the threshold value of the number of arrival cells and the frequency of exceeding the threshold value of the number of arrival cells of period kT is further deteriorated.

[0112]

The problem to be solved is that, in the prior art, only data obtained by simple traffic measurement in an ATM switch is used without using a special measurement form such as an external device. Therefore, it is impossible to accurately measure the measurement items required for setting the ATM virtual path capacity.

An object of the present invention is to solve the problems of the prior art and to provide an ATM virtual path capacity setting method and an ATM virtual path capacity setting method capable of accurately setting an ATM virtual path capacity of an ATM network in actual operation. An object of the present invention is to provide an ATM virtual path capacity setting device.

[0114]

To achieve the above object, an ATM virtual path capacity setting method and an ATM according to the present invention are provided.
The virtual path capacity setting device first counts the number of cells (first arriving cells) arriving at the cell transmission waiting buffer associated with the ATM virtual path at every predetermined cycle T, and based on the counting result, A value (second arriving cell number) is obtained by dividing the total number of cells for each period kT (k: an integer of 2 or more) by k.
Next, at a predetermined period T '(T'> T), the period T
On the basis of the counting result for each of the first arrival cells, the number of the first arrival cells that are equal to or more than the number of the cells are totaled for each first arrival cell, and the total value is calculated for all of the first arrival cells in the period T ′. Dividing by the sum of the first arrival cell number, and calculating the value (frequency of exceeding the first arrival cell number threshold value, frequency of less than the first arrival cell number threshold value);
The logarithm of each of the first arrival cell number threshold value excess frequency and the first arrival cell number threshold value less frequency is calculated, and further approximated by a predetermined m-th order expression. Similarly, an approximation of each logarithm of the second arrival cell number threshold excess frequency and the second arrival cell number threshold lower frequency with respect to the second arrival cell number by m-order expression is performed, and Attenuation rate of the complementary distribution of the number of cells arriving per unit time, based on the first and second estimated logarithmic threshold exceeding frequencies obtained as a result of the approximation and the first and second estimated logarithmic threshold below frequencies. And the rate of rise of the distribution. Then, an average (average cell rate) of the number of arriving cells in the cycle T ′ is obtained based on the first number of arriving cells counted in each cycle T, and the average cell rate and the first number of arriving cells counted in the cycle T are calculated. The variance of the number of arriving cells in the period T ′ is calculated based on the average cell rate, the variance, the decay rate of the complementary distribution, and the rate of increase of the distribution. And the probabilities a, d for transitioning from ON to OFF and from OFF to ON for each reciprocal time (time slot) of the peak cell rate, and based on these peak cell rates, the probabilities a, d, and the target cell loss rate. , An ATM virtual path capacity that satisfies the target cell loss rate is obtained. Before obtaining the first logarithmic threshold excess frequency and the second logarithmic threshold excess frequency, the numbers of first and second arrival cells in each cycle T are sorted in descending order.

[0115]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a flowchart showing an example of the processing operation of the ATM virtual path capacity setting method of the present invention, and FIG. 2 is a configuration according to the present invention of an ATM virtual path capacity setting apparatus of the present invention and an ATM exchange provided with the same. FIG. 3 is a block diagram showing one embodiment of the present invention.

The ATM exchange shown in FIG. 2 has a cell transmission device 100 like the ATM exchanges shown in FIGS. 11 to 13, 15, and 17 in the prior art. Cell transmission interface (described as “IF” in the figure) for each physical transmission medium 200
01 is provided to suppress cell transmission exceeding the capacity of the transmission medium 200.

As a logical functional configuration, the cell transmission interface 101 has an ATM virtual path 201,
Each of the shapers 202,... (Described as “S” in the figure) 1
Are connected respectively. Are configured to suppress cell transmission exceeding the capacity logically defined for the ATM virtual path, and perform timing control with the cell transmission interface 101 according to a certain scheduling rule. The cell is transmitted. Also, each of the shapers 111, 112,.
.. Are ATM virtual paths 201, 202,.
・ One-to-one correspondence.

Further, each of the shapers 111, 112,.
Are connected to cell transmission waiting buffers 121, 122,..., Respectively.
, 122,..., When a cell exceeding the ATM virtual path capacity flows in, the cell transmission waiting buffers 121, 12
.., And waits for the next cell transmission timing.

Each cell transmission waiting buffer 121, 12
The soft counters 131, 132,.
Are provided along with the soft counters 131 and 13.
ATM logical paths 201,
When the cell flows 301 and 302 are sorted into 02,..., The number of arriving cells for each ATM virtual path is counted. The count results of the soft counters 131, 132,... Are sent to the ATM virtual path capacity setting device 400 via the signal line 401.

The ATM virtual path capacity setting device 400 of this embodiment
Is a cell counting unit 1, a table 2, an average / variance calculating unit 3,
Tables 4 to 6, average counting section 4a, rearranging sections 5a and 6
a, excess frequency calculation units 7 and 8, tables 9 and 10, lower frequency calculation units 11 and 12, tables 13 and 14, increase rate of arrival cell number distribution / attenuation rate calculation unit of complementary distribution, peak rate calculation unit 16 , An ATM virtual path capacity calculation unit 17,
Based on the count results of the soft counters 131, 132,..., The ATM virtual path capacity that satisfies the target cell loss rate is set for the ATM virtual path terminating in the ATM exchange.

That is, the soft counters 131 and 13
Based on the counting results of 2,...,
Cell transmission waiting buffer 12 for each preset unit period T
The number of cells arriving at 1, 122,... Is counted and registered in Table 2. On the basis of the registered contents of this table 2, the average / variance calculating unit 3 calculates the average number of arrival cells and the variance in the cycle T ′ (T ′> T).

The average counting unit 4a calculates the number of cells arriving at the period kT (k: an integer of 2 or more) / k (the second number of arriving cells) based on the registered contents of the table 2. Register with. Then, by the rearranging unit 5a,
In the cycle T ′, the data is rearranged in the descending order of the value of the number of arrival cells and registered in the table 5. Further, the rearranging unit 6a
As a result, based on the registered contents of Table 4, in a cycle T ′,
The data is rearranged in the descending order of the value of the second arrival cell number and registered in the table 6.

Then, based on the registered contents of Table 5,
The excess frequency calculation unit 7 sets each arrival cell number as a threshold value, sums up each arrival cell number equal to or more than this threshold value, and divides this total value by the total number of all cells arriving in the period T ′. The first frequency of exceeding the threshold value of the number of arrival cells is calculated, and the first frequency of exceeding the threshold value of the number of arrival cells is calculated to calculate the first frequency of exceeding the threshold value of the number of arrival cells. Register with.

Similarly, based on the registered contents of Table 5, the less-frequency calculating section 11 sets each arrival cell number as a threshold value and sums the respective arrival cell numbers less than the threshold value.
This total value is divided by the sum of the total number of cells arriving in the period T ′ to obtain a frequency less than the first threshold value of the number of arriving cells.
The first logarithm of the number of cells less than the threshold is calculated by taking the logarithm of the frequency below the threshold of the number of arrival cells in Table 13.
Register with.

Further, based on the registered contents of the table 6, the excess frequency calculation unit 8 sets each second arrival cell number as a threshold value and sums up each second arrival cell number equal to or more than this threshold value. Is divided by the total value of all the second arrival cell counts during the period T ′ to obtain a second arrival cell count threshold value exceeding frequency. The logarithm of the frequency is taken to calculate the second logarithm-arriving-cell-number-threshold-exceeding frequency and registered in the table 10.

Similarly, based on the registered contents of Table 6, the less-frequency calculating section 12 sets each second arriving cell number as a threshold value and sums each second arriving cell number less than the threshold value. Then, the total value is divided by the total value of all the second arrival cell numbers during the period T ′ to obtain a frequency lower than the second arrival cell number threshold value. By taking the logarithm of less than the frequency, the frequency below the second logarithmic arrival cell number threshold value is calculated and registered in the table 14.

Then, based on the registered contents of table 9, first, the first logarithmic arrival cell number threshold value exceeding frequency is determined by the rising rate / complementary distribution attenuation rate calculation section 15 of the arrival cell number distribution. The coefficient is calculated by approximation with an m-order function.
Based on the registered contents of Table 10, the second logarithm arrival cell number threshold excess frequency is approximated by an m-th order function of the threshold to calculate its coefficient, and from these two sets of coefficients, the arrival cell number complementary distribution is calculated. Is calculated.

Next, based on the registered contents of the table 13, the rate of increase of the arrival cell number distribution and the attenuation rate of the complementary distribution determine the frequency less than the first logarithmic arrival cell number threshold value as the threshold value m. Calculates the coefficient by approximating by the following function and calculates the coefficient by approximating the frequency less than the second logarithmic arrival cell number threshold by the m-th order function of the threshold based on the registered contents of the table 14. Then, the rate of increase of the distribution of the number of arrival cells is calculated from these two sets of coefficients.

In this manner, the rate of increase of the distribution of the number of arrival cells
The number of arriving cells calculated by the compensating distribution attenuation rate calculator 15, the increase rate of the complement distribution attenuation rate and the number of arrival cells, and the average cell rate and variance of the number of arriving cells calculated by the average and variance calculator 3. From the peak rate calculation unit 16
The peak cell rate when the process of arriving the cell at the M virtual path is approximated as the ON-OFF process is calculated by performing processing described later in detail.

Further, the peak rate calculation section 16 sets the state in which a cell is present in the time slot to ON, sets the state in which no cell is present in the time slot to OFF, and sets the state in which the cell is not present to OFF in the time discretized by the reciprocal of the peak cell rate (time slot). The transition probability to OFF and the transition probability from OFF to ON are calculated.

Thus, the peak rate calculating section 16
And the peak cell rate obtained in
The ATM virtual path capacity calculation unit 17 satisfies the target cell loss rate that satisfies the predetermined target cell loss rate based on the three values of the transition probability from the ATM virtual path and the transition probability from the OFF state to the ON state.
Find the TM virtual path capacity at high speed and with high accuracy.

FIG. 3 is an explanatory diagram showing a configuration example of the table (2) in FIG. 2, FIG. 4 is an explanatory diagram showing a configuration example of the table (4) in FIG. 2, and FIG. FIG. 6 is an explanatory diagram showing a configuration example of the table (5), FIG. 6 is an explanatory diagram showing a configuration example of the table (6) in FIG. 2, and FIG.
FIG. 8 is an explanatory diagram showing a configuration example of the table (9) in FIG. 2, FIG. 8 is an explanatory diagram showing a configuration example of the table (10) in FIG. 2, and FIG. 9 shows a configuration example of the table (13) in FIG. FIG. 10 shows the table (1) in FIG.
It is explanatory drawing which shows the example of a structure of 4).

Table 2 shown in FIG. 3 corresponds to A in FIG.
Among the counting results of the cell transmission device 100 of the TM exchange, AT
This is a table in which only data of a portion related to the M virtual path 201 is summarized.

In this table 2, reference numerals 21 and 22 denote records each indicating a column of the counting cycle (counting cycle number) and the number of arrival cells, and 211, 212,... , 221, 222,... Are records of the number of arrival cells (d ₁ , d ₂ ,..., D _n ) corresponding to the counting cycle numbers 211, 212,. Here, the count cycle number 211 indicates the immediately preceding count cycle, and the count cycle number 212 indicates the count cycle immediately before the count cycle number 211.

In table 4 shown in FIG.
2 is the number of the counting cycle (counting cycle number) and the cycle kT, respectively.
Is a record showing the column of the average number of cells reached in
, 412,... Are records indicating numbers around the counting, and 421, 422 are counting cycle numbers 411, 412,.
, The average number of cells arriving at a period kT (g ₁ , g ₂ ,
, G _{n-k + 1} ), and the average number of arrival cells (g ₁ , g ₂ ,..., G _{n-k + 1} ) of this period kT is calculated by the following “Equation 29”. It is obtained by the formula.

[0136]

(Equation 29)

Here, the counting cycle number 411 indicates the immediately preceding counting cycle, and the counting cycle number 412 indicates the counting cycle number 41.
This shows a counting cycle one before one.

In Table 5 in FIG. 5, the result obtained by rearranging the numbers of arriving cells in Table 2 in FIG. 2 in descending order of the values is registered. Reference numeral 51 denotes a column of the number of arriving cells arranged in descending order of the number of arriving cells. 511,
.. Are records of the number of arrival cells rearranged in descending order of the arrival cell numerical value, and are represented by the following equation (30).

[0139]

[Equation 30]

[0140] Each value and its formula in "Equation 30" are expressed as "dA ₁ > dA ₂ >> ... dA _n " in the text.

In Table 6 in FIG. 6, the result obtained by rearranging the number of second arrival cells in Table 4 in FIG. 2 in descending order of value is registered, and 61 is arranged in descending order of the number of second arrival cells. Are records indicating the second column of the number of arrival cells, 611, 612,... Are records of the second number of arrival cells rearranged in descending order of the number of arrival cells. It becomes as shown by the formula.

[0142]

[Equation 31]

The values and their expressions in “Equation 31” are described in the text as “gA ₁ > gA ₂ >...> G
A _{n−k + 1} ”.

The tables 5 and 6 shown in FIGS. 5 and 6 correspond to the cell transmission device 1 of the ATM exchange shown in FIG.
This is a table in which only the data of the portion related to the ATM virtual path 201 out of the count results of 00 is summarized.

Table 9 in FIG. 7 is for registering the calculation result of the excess frequency calculation unit 7 in FIG. 2. In this table 9, reference numeral 91 denotes the same number of arrival cells as in Table 5 shown in FIG. 911, 912,... Are records showing columns of the number of arrival cells arranged in descending order.
It is a record of the number of arrival cells rearranged in descending order of the arrival cell numerical value. Reference numeral 92 denotes a record indicating a column of the frequency of exceeding the threshold value of the arrival cell number, and 921, 922,...
Are records of the arrival cell number threshold excess frequency when each of the records 911, 912,... Is regarded as a threshold, and is expressed by the following equation (32).

[0146]

(Equation 32)

The value of the frequency exceeding the threshold value of each arrival cell number in “Equation 32” is represented by “dF ₁ , dF ₂ , d”
F ₃ ,..., DF _n .

Are the logarithms of the respective records 921, 922,..., And 941, 942,.
,... Are obtained by the following equation (33).

[0149]

[Equation 33]

951, 952,..., 96
1,962, ... and 971,972, ...
Are records obtained by the following equations (Equation 34) to (Equation 36) for the respective records 911, 912,....

[0151]

(Equation 34)

(Equation 35)

[Equation 36]

In the example of FIG. 7, "m = 5 (fifth order)" is set to simplify the description.

The table 10 in FIG. 8 registers the calculation result of the excess frequency calculation unit 8 in FIG.
In this table 10, reference numeral 1010 denotes a record indicating a column of the number of arrival cells arranged in descending order of the number of second arrival cells, which is the same as the table 6 shown in FIG.
012,... Are records of the number of arrival cells rearranged in descending order of the arrival cell numerical value.

Reference numeral 1020 denotes a record indicating a column of the arrival cell number threshold value exceeding frequency.
.. Are respectively 1011, 1012,.
Is a record of the frequency exceeding the threshold value of the arrival cell number when each record is regarded as the threshold value, and is obtained by the following expression (Expression 37).

[0155]

(37)

Note that the value of the frequency exceeding the threshold value of each arrival cell number in Expression 37 is represented by “gF ₁ , gF ₂ , g”
F _3, ···, expressed as gF _{n-k + 1} ".

Are logarithms of the records 1021, 1022,..., Respectively. 1041, 1042,.
.. and 1051, 1052,..., 1061, 10
, And 1071, 1072,... Are records obtained by the following equations 38 to 41 with respect to the records 1011, 1012,. It is.

[0158]

(38)

[Equation 39]

(Equation 40)

[Equation 41]

The table 13 in FIG. 9 is for registering the calculation result of the less-frequent frequency calculating unit 11 in FIG. 2. In this table 13, 810 is the same as the table 5 shown in FIG. This is a record showing a column of the number of arrival cells arranged in descending order,
.. Are records of the arrival cell number rearranged in descending order of the arrival cell number.

Reference numeral 820 denotes a record indicating a column having a frequency lower than the threshold value of the number of arrival cells, and 821, 822,...
.. Are records having a frequency less than the threshold value of the number of arrival cells when each of the records 811, 812,... Is regarded as a threshold, and is represented by the following expression (42).

[0161]

(Equation 42)

The value of the frequency less than the threshold value of each arrival cell number in “Equation 42” is referred to as “dM ₁ , dM ₂ ,.
.., DM _j ,..., DM _n .

Are the logarithms of the records 821, 822,..., Respectively, and 841, 842,.
,... Are obtained by the following equation (43).

[0164]

[Equation 43]

Also, 851, 852,..., 86
1,862, ... and 871,872, ...
Are records obtained by the following equations (Equation 44) to (Equation 46) for the records 811, 812,....

[0166]

[Equation 44]

[Equation 45]

[Equation 46]

In the example of FIG. 9, "m = 5 (fifth order)" is set for simplification of the description.

The table 14 in FIG. 10 is for registering the calculation result of the lower frequency calculation unit 12 in FIG. 2. In this table 14, 1410 is the same as the table 6 shown in FIG. Is a record showing a column of the number of arrival cells arranged in descending order of the number of arrival cells.
411, 1412,... Are records of the number of arrival cells rearranged in descending order of the arrival cell numerical value.

Reference numeral 1420 denotes a record indicating a column having a frequency lower than the threshold value of the number of arrival cells.
.. Are respectively 1411, 1412,.
Is a record with a frequency less than the threshold value of the number of arrival cells when each record is regarded as a threshold value, and is obtained by the following expression (47).

[0170]

[Equation 47]

Note that the value of the frequency exceeding the threshold value of each arrival cell number in “Equation 47” is represented by “gM ₁ , gM ₂ ,.
.., GM _j ,..., GM _{n−k + 1} .

Are the logarithms of the records 1421, 1422,..., Respectively. Also, 1441, 1442,.
.. and 1451, 1452, ..., 1461, 14
, And 1471, 1472,... Are records obtained by the following equations (48) to (51) with respect to 1411, 1412,. It is.

[0173]

[Equation 48]

[Equation 49]

[Equation 50]

(Equation 51)

As described above, in FIG. 2, the ATM virtual path capacity setting device 400 rearranges the data of the tables 2 and 4 in the descending order of the value of the number of arriving cells to create the tables 5 and 6. The excess frequency calculation unit 7 and the excess frequency calculation unit 8 process the respective data of the tables 5 and 6 as described below to create the tables 9 and 10, and furthermore, the tables 9 and 10 Based on each data of the table 10, the rising rate of the arrival cell number distribution and the attenuation rate calculating section 15 of the complementary distribution,
Calculate the attenuation rate of the distribution complementary to the number of arriving cells.

The ATM virtual path capacity setting device 400
Is calculated by the less-frequency calculator 11 and the less-frequency calculator 12.
The data of Table 5 and Table 6 are processed as described below to create Tables 13 and 14, and further, Table 13 and Table 1
Based on the respective data of No. 4, the increase rate of the arrival cell number distribution is calculated by the increase rate / complementary distribution attenuation rate calculation unit 15 of the arrival cell number distribution.

Then, based on the rate of increase and the rate of decay, the peak rate calculating section 16 changes the peak cell rate when the arrival process of the cell arriving at the ATM virtual path is regarded as the ON-OFF process, and changes the ON state from the OFF state to the ON state. The transition probability to the state, the transition probability from the ON state to the OFF state is calculated, and based on the peak cell rate, the transition probability from the OFF state to the ON state, and the transition probability from the ON state to the OFF state, A
The ATM virtual path 2 is calculated by the TM virtual path capacity calculation unit 17.
The ATM virtual path capacity that satisfies the target cell loss rate of 01 is calculated.

For the sake of simplicity, in the following description, it is assumed that there is no missing data. Assuming that a measurement result in a predetermined count cycle T (for example, 1 second) in a predetermined past data use cycle T ′ (for example, 30 minutes) is obtained, the total count number n in this case is “1800 (= 30 × 6
0) ".

The cell counting unit 1 reaches the cycle T every cycle T.
The number of arrived cells is counted and registered in Table 2. Also,
The average counting unit 4a outputs the averaged value for each cycle kT as a text.
Register in Table 4. In addition, the rearranging unit 5a
In the rule 2, the n data in the cycle T 'are
And d_i, i = 1, 2,..., n
Then, this is registered in Table 5. Arrange in this way
The replaced data is dA _i, i = 1,2, ..., n
And is represented by the following “Equation 52”.

[0179]

(Equation 52)

The rearranging section 6a stores
And “n−k + 1” data in the cycle T ′
And g_i, i = 1, 2,..., n−k + 1 in descending order
And this is registered in the table 6. Where
The replaced data is gA _i, i = 1, 2,..., n
−k + 1 and is represented by the following “Equation 53”.

[0181]

(Equation 53)

On the basis of the data in Table 2, the average / variance calculation unit 3 calculates the average cell rate λ in the past data use period T ′ by the following expression (Expression 54), and calculates the variance V of the number of arriving cells by “Equation 54”. Equation 55 ”is calculated.

[0183]

(Equation 54)

[Equation 55]

Next, in the excess frequency calculating section 7, each of the rearranged data (dA _i , i = 1, 2,.
., N), using the respective thresholds, the first arrival cell number threshold value exceeding frequency (dF _i , i = 1, 2,..., N) ), And its logarithm (logdF _i , i = 1, 2,..., N)
And {(dA _i / dA _n ) −1} ² , {(dA _i / dA)
_n ) -1｝ ³ , {(dA _i / dA _n ) -1} ⁴ , ｛(dA _i /
dA _n ) -1｝ ⁵ , i = 1, 2,..., n,
Each value indicated by “Equation 57” is calculated.

[0185]

[Equation 56]

[Equation 57]

Similarly, in the excess frequency calculating section 8, each of the rearranged data (gA _i , i = 1, 2, 2,
, N−k + 1), and using the respective threshold values, the second threshold value exceeding frequency (gF _i , i = 1, 2,.・, N−k +
1), and its logarithm (logF _i , i = 1,
2,..., N−k + 1), and ｛(gA _i / gA)
_{n + k-1) -1}} 2, {(gA i / gA n + k-1) -1} 3,
_{{(GA i / gA n +} k-1) -1} 4, {(gA i / gA n
_{+ k−1} ) −1｝ ⁵ , i = 1, 2,..., n + k−1, that is, each value indicated by “Equation 59” is calculated.

[0187]

[Equation 58]

[Equation 59]

On the basis of these results, the rising rate of the arrival cell number distribution and the attenuation rate of the complementary distribution are calculated by the equation (56).
(LogdF _i , i = 1, 2,...,
n) is approximated by the following equation (60), and coefficients a ₂ , a ₃ , a ₄ and a ₅ are obtained by regression analysis.

[0189]

[Equation 60]

Similarly, the logarithm (logF _i , i = 1, 2,..., N−k + 1) of the data of “Equation 58” is calculated in the rise rate of the arrival cell number distribution and the attenuation rate of the complementary distribution. Is approximated by the following equation (61), and coefficients b ₂ , b ₃ , b ₄ , and b ₅ are obtained by regression analysis.

[0191]

[Equation 61]

[0192] Further, in the rise rate of the arrival cell number distribution and the attenuation rate of the complementary distribution, the respective coefficients (a ₂ ,
a ₃ , a ₄ , a _{5 ,} b ₂ , b ₃ , b ₄ , b ₅ ), the attenuation rate of the distribution complementary to the number of arriving cells is approximately obtained as in the following equation (62).

[0193]

(Equation 62)

Here, σ is a value that satisfies the following equation (63).

[0195]

[Equation 63]

Further, the rate of rise of the distribution of the number of arrival cells and the attenuation rate of the complementary distribution are calculated by the calculation unit 15 as follows. First, in the less-than-frequency calculating unit 11, each of the sorted data (dA _i , i
= 1, 2,..., N), using the respective thresholds, the following frequency of the first arrival cell number threshold (dM _i , i = 1, 2) ,..., N) and its logarithm (logdM _i , i = 1, 2,.
···, n) and {1- (dA _i / dA ₁ )} ² ,
{1- (dA _i / dA ₁ )} ³ , {1- (dA _i / d)
A ₁ )｝ ⁴ , {1- (dA _i / dA ₁ )} ⁵ , i = 1,2,2
.., N, that is, the respective values represented by “Equation 65” are calculated.

[0197]

[Equation 64]

[Equation 65]

Similarly, in the less-frequency calculating section 12, each of the rearranged data (gA _i , i = 1,
2,..., N−k + 1), with the respective threshold values, the frequency (gM _i , i = 1, 2, ..., n-
k + 1), and its logarithm (logM _i , i =
1, 2,..., N−k + 1), and ｛1− (gA _i
/ GA ₁ )｝ ² , ｛1- (gA _i / gA ₁ )｝ ³ , ｛1-
(GA _i / gA ₁ )｝ ⁴ , {1- (gA _i / gA ₁ )} ⁵ , i
= 1, 2,..., N + k−1, that is, “Equation 67”
Each value shown by is calculated.

[0199]

[Equation 66]

[Equation 67]

On the basis of these results, the logarithm (logdM _i , i = 1, 2,.
···, n) is approximated by the following equation (Expression 68), and “Expression 69” is obtained by regression analysis. The expression 69 is denoted by aM ₂ , aM ₃ , aM ₄ and aM ₅ in the text.

[0201]

[Equation 68]

[Equation 69]

Similarly, the logarithm (logM _i , i = 1, 2,..., N−k) calculated by the less-frequent frequency calculator 12 is calculated by the rise rate / complementary distribution attenuation rate calculator 15 of the arrival cell number distribution.
+1) is approximated by the following equation 70, and equation 71 is obtained by regression analysis. Incidentally, the "number 71" is in the text shown in _{bM 2, bM 3, bM 4} , bM 5.

[0203]

[Equation 70]

[Equation 71]

Further, the rise rate of the arrival cell number distribution and the attenuation rate of the complementary distribution calculation section 15 calculate each of these coefficients (a
_{M 2, aM 3, aM 4} , aM 5, bM 2, bM 3, bM 4, bM
₅ ) From the rise rate of arrival cell number distribution (however, γ ≦ λ)
Is approximately obtained by the following equation (72).

[0205]

[Equation 72]

Here, the following “Equation 7” in “Equation 72” is used.
“3” (“σM” in the text) is a value that satisfies the expression of “Equation 74”.

[0207]

[Equation 73]

[Equation 74]

Next, the approximation to the ON-OFF process will be described. Now, let ψ be the peak cell rate of the past data use cycle T ′ in the ON-OFF process. If the reciprocal of the peak cell rate ψ is called a time slot, a transition probability a from the OFF state to the ON state between time slots and a transition probability from the ON state to the OFF state are d. Here, the ON state refers to a state in which a cell exists in a time slot, and the OFF state refers to a state in which a cell does not exist in a time slot.

The transition probabilities a and d and the peak cell rate ψ are calculated by the following equations (“Equation 75” and “Equation 7”).
6 "," Equation 77 ").

[0210]

[Equation 75]

[Equation 76]

[Equation 77]

Here, in the equation of “Equation 77”, γ corresponds to the threshold value of the number of arriving cells, and μ (θ) is expressed by the following equation (“Equation 7”).
8)).

[0212]

[Equation 78]

[0213] The calculation formulas of "Equation 77" and "Equation 78" are as follows.
"Exponential bounds for queue" by NG Duffield
s with Markovian arrivals ”, Queuing System, 17, pp.
413-430 (1994).

The peak rate calculation section 16 calculates the average ("Equation 54"), the variance V ("Equation 55") obtained by the average and variance calculation section 3, and the rate of increase of the arrival cell number distribution.
Using the decay rate of the complementary distribution of the number of arriving cells ("Equation 62") and the increase rate of the distribution of the number of arriving cells ("Equation 72") obtained by the decay rate calculation unit 15 of the complementary distribution, "Equation 75", "Equation 75" Equation 76 ",
The transition probabilities a and d and the peak cell rate ψ are calculated from each equation of “Equation 77”.

From the expressions of “Expression 75” and “Expression 76”, a and d are expressed as functions of ψ, and they are substituted into the expression of “Expression 72” and “Expression 77” to obtain dA _i , i = 1, With respect to 2,..., N, assuming that a set of i that can obtain ψ (peak cell rate) in the range of dA ₁ <ψ is {i ′}, ψ satisfying the following expression (79) Ask.

[0216]

[Expression 79]

Under the approximation of the ON-OFF process, the target value of the cell loss rate is approximated as in the following “Expression 80”.

[0218]

[Equation 80]

Here, x satisfies the following “Equation 81”.

[0220]

[Equation 81]

Y that satisfies Equation 80 and Equation 81 is calculated, and the ATM virtual path capacity C _d that satisfies the target cell loss rate CLR is calculated from the following Equation 82. .

[0222]

(Equation 82)

Here, M is a constant for converting the number of cells per unit time into capacity.

The ATM virtual path capacity calculation unit 17 calculates the transition probabilities a and d obtained by the peak rate calculation unit 16, the peak cell rate ψ, and the target cell loss rate CLR.
The ATM virtual path capacity C _d that satisfies the target cell loss rate is calculated using “Equation 80”, “Equation 81”, and “Equation 82”.

Next, the estimated cell loss rate CLR _e is obtained by calculating x, y from the following “Equation 83” and “Equation 84” for a, d, ψ obtained by the peak rate calculation unit 16. ,
It is calculated by substituting into the right side of “Equation 80”.

[0226]

[Equation 83]

[Equation 84]

Hereinafter, an example of the processing operation of such an ATM virtual path capacity setting device will be described with reference to FIG. First, the measurement period (cycle) T 'number arrival cells every predetermined period T in (d _i) counts (step S1), and based on the number of the arriving cells (d _i), the period T' average between The cell rate λ and the variance V of the number of arriving cells are obtained (step S2), and the average number of arriving cells (g _i ) for each period kT in the period T ′ is calculated (step S3).

Sorted in descending order within the period T ′ based on the number of arriving cells (d _i ) (dA _i , dA ₁ > dA ₂ > ·)
..> DA _n ; see “Equation 52” (Step S)
4). Similarly, those sorted in the descending order within the period T ′ based on the average number of cells (g _i ) (gA _i , gA ₁ > g)
_{A 2>···> gA n-} k + 1; "number 53" see) seek (step S5).

Further, the result obtained in step 4 (d
Based on A _i ), the value (dA _i ) of this result is regarded as a threshold, and the number of arrival cells exceeds the threshold frequency (dF _i ;
(See "Equation 56") (step S6). Similarly,
Based on the result (gA _i ) obtained in step 5, a threshold value exceeding frequency (gF _i ; see Equation 58) of the number of arrival cells when the value (gA _i ) of the result is regarded as a threshold value is obtained. (Step S7).

The result obtained in step 4 (dA _i )
The basis, the resulting value (dA _i) the arriving cell count threshold than frequency when considered as a threshold (dM _i; "6
4 ") (step S8). Similarly, based on the result (gA _i ) obtained in step 5, when the value of the result (gA _i ) is regarded as the threshold, the frequency of arrival cell count less than the threshold (gM _i ; see Equation 66) ) Is obtained (step S9).

The result (dA _i , dF _i ) obtained in step 6
Then, the coefficients a ₂ , a ₃ , a ₄ , and a ₅ when the logarithm (logdF _i ) is expressed by m (here, 5) of “dA _i ” are obtained (step S10). Similarly, from the result (gA _i , gF _i ) obtained in step 7, the logarithm (logF _i )
A coefficient in the case of representation in the 5-order equation of "gA _i" b _2,
b ₃ , b ₄ , b ₅ are obtained (step S11).

Also, the result (dA _i ,
dM _i ), the logarithm (logdM _i ) is calculated as m of “dA _i ”.
(Here, 5) Coefficient aM ₂ , a when expressed by the following equation
M ₃ , aM ₄ , and aM ₅ are obtained (step S12). Similarly, from the result (gA _i , gM _i ) obtained in step 9, the coefficients bM ₂ , bM ₃ , bM ₄ , bM ₅ when the logarithm (logM _i ) is expressed by the quintic of “gA _i ” Is obtained (step S13).

The coefficients (a) obtained in steps S10 and S11
₂ , a ₃ , a ₄ , a _{5 ,} b ₂ , b ₃ , b ₄ , b ₅ ) and the average cell rate (λ) are used to determine (σ) such that “I (λ) = 0”. Attenuation rate (I (γ) of the distribution complementary to the number of arrival cells,
γ ≧ λ) is calculated (step S14). Similarly, the coefficients (aM ₂ , aM ₃ , aM) determined in steps S12 and S13
_{4, aM 5, bM 2,} bM 3, bM 4, bM 5) and based on the average cell rate (lambda), such that "I (λ) = 0,""σ
M ”, and the rate of increase of the distribution of the number of arrival cells (I (γ), γ ≦
λ) is calculated (step S15).

From the average cell rate (λ), the variance (V) of the number of arriving cells, and the attenuation rate (I (γ)) of the complementary distribution of the number of arriving cells, the peak cell rate (ψ) in the range of “dA ₁ <ψ”. ) And the transition probabilities (a) and (d) are obtained (step S
16). Then, the target cell loss rate CLR, the obtained peak cell rate (ψ), transition probabilities (a), (d)
Then, the ATM virtual path capacity (C _d ) is calculated (step S17).

As described above with reference to FIG. 1 to FIG. 10, the ATM virtual path capacity setting method and the AT
The M virtual path capacity setting device counts the number of cells arriving at the cell transmission waiting buffer associated with the ATM virtual path, and calculates the average number of arriving cells, the variance of the number of arriving cells, the decay rate of the complementary distribution of the number of arriving cells, By calculating the rate of increase in the distribution of the number of arrival cells, the cell loss in the cell transmission waiting buffer associated with the ATM virtual path can be efficiently obtained, and the ATM virtual path capacity can be set according to the actual traffic demand. It is possible to set the capacity according to the actual traffic demand at high speed and with high accuracy based on the lightweight measurement.

That is, in this example, the extremely lightweight traffic measurement of the number of cells arriving at the cell transmission waiting buffer makes it possible to quickly estimate the Q-length complementary distribution function, and approximates the cell arrival process by the ON-OFF process. As such, bursty traffic is also considered.

Further, it is not necessary to set the arrival cell number threshold value in advance, and the arrival cell number threshold frequency is determined not by the number of times the arrival cell number exceeds the threshold but by the data of the arrival cell number. Since it is obtained using the threshold value, the frequency of exceeding the threshold value of the number of arrival cells can be calculated with high accuracy.

In order to calculate I (γ), not only the decay rate (I (γ), γ ≧ λ) of the complementary distribution of the number of arriving cells, but also
Using the rate of rise of the distribution of the number of arriving cells (I (γ), γ ≦ λ),
Furthermore, since the approximation is made so as to satisfy the property I (γ) = 0 of I (γ), highly accurate estimation is possible. As a result, it is possible to accurately estimate the rising rate of the distribution of the number of arriving cells and the attenuation rate of the complementary distribution, and it is possible to accurately estimate the Q-length complementary distribution function.

From these, particularly, UBR (Unspecifie)
d Bit Rate) It is possible to design (set) the capacity of the ATM virtual path accommodating the best-effort virtual circuit including the bearer class so as to satisfy a certain cell loss rate target value (not a specified value). Will be possible.

Furthermore, it is possible to check and record daily that the resources of the ATM network are properly arranged in the ATM virtual path and that they are being used effectively, and that the traffic flows evenly throughout the network. Confirmation becomes possible. In addition, it is possible to provide information for performing appropriate reconfiguration in a reasonable operation and period in response to a change in network configuration such as new addition or relocation of a virtual circuit.

The present invention is not limited to the embodiment described with reference to FIGS. 1 to 10 and can be variously modified without departing from the gist thereof. For example, in this example, the part that performs the processing according to the present invention is referred to as a cell counting unit 1.
And average / variance calculation unit 3, excess frequency calculation units 7 and 8, lower frequency calculation units 11 and 12, increase rate of arrival cell number distribution / decay rate calculation unit of complementary distribution, peak rate calculation unit 16 and A
Although the configuration is divided into the TM virtual path capacity calculation unit 17 and each function, the integration and division of each function may be performed in any module unit according to the actual design. Further, the functions of the respective units may be realized using a computer including a central processing unit (CPU) and a storage device.

[0242]

According to the present invention, an AT having a special function is provided.
Without using a special measurement form such as an M-exchange or an external device, the data obtained by simple traffic measurement in the ATM exchange can be used for the A-operation of the ATM network in operation.
The TM virtual path capacity can be accurately set so as to satisfy a predetermined cell loss rate target value (CLR), and efficient operation of network resources can be achieved.

[Brief description of the drawings]

FIG. 1 is a flowchart illustrating an example of a processing operation of an ATM virtual path capacity setting method according to the present invention.

FIG. 2 is a block diagram showing one embodiment of a configuration according to the present invention of an ATM virtual path capacity setting device of the present invention and an ATM exchange provided with the same;

FIG. 3 is an explanatory diagram showing a configuration example of a table (2) in FIG. 2;

FIG. 4 is an explanatory diagram showing a configuration example of a table (4) in FIG. 2;

FIG. 5 is an explanatory diagram showing a configuration example of a table (5) in FIG. 2;

FIG. 6 is an explanatory diagram showing a configuration example of a table (6) in FIG. 2;

FIG. 7 is an explanatory diagram showing a configuration example of a table (9) in FIG. 2;

FIG. 8 is an explanatory diagram showing a configuration example of a table (10) in FIG. 2;

FIG. 9 is an explanatory diagram showing a configuration example of a table (13) in FIG. 2;

FIG. 10 is an explanatory diagram showing a configuration example of a table (14) in FIG. 2;

FIG. 11 is a block diagram showing a configuration example of an ATM virtual path capacity setting mechanism in a conventional ATM exchange.

FIG. 12 is a block diagram showing an example of a functional configuration of an ATM exchange for realizing the first related art.

FIG. 13 is a block diagram showing a functional configuration example of an ATM exchange for realizing the second conventional technique.

FIG. 14 is an explanatory diagram showing an example of a Q length distribution.

FIG. 15 shows an ATM when two Q length thresholds are set.
FIG. 2 is a block diagram illustrating an example of a functional configuration of the exchange.

16 is an explanatory diagram showing an example of a result of counting cells by the cell transmission device in FIG. 15;

FIG. 17 is a block diagram showing an example of a functional configuration of an ATM exchange for realizing a fourth conventional technique.

18 is an explanatory diagram showing a configuration example of a table for registering a counting result of a cell counting unit in FIG. 17;

19 is an explanatory diagram showing a configuration example of a table for registering a counting result of an excess number counting unit in FIG. 17;

[Explanation of symbols]

1: cell counting unit, 2,4 to 6,9,10,13,14:
Table 3: Average / variance calculation unit, 7, 8: Excess frequency calculation unit, 11, 12: Less frequency calculation unit, 15: Increase rate of arrival cell number distribution, attenuation rate calculation unit of complementary distribution, 16: Peak rate Calculator, 17: ATM virtual path capacity calculator, 10
0: Cell sending device, 101: Cell sending interface ("IF"), 111, 112, ...: Shaper ("S"), 121, 122, ...: Cell sending waiting buffer, 131, 132, ... ..: Soft counter, 20
0: Transmission medium, 201, 202, ...: ATM virtual path ("virtual path"), 301, 302, ...: Cell flow, 400: ATM virtual path capacity setting device, 401: Communication line, 21: Record row (counting cycle), 22: record row (number of arrival cells), 211, 212: record (counting cycle), 221, 222: record (number of arrival cells), 4
1: Record row (counting cycle), 42: Record row (number of arrival cells), 411, 412: Record (counting cycle), 4
21, 422: record (number of arrival cells), 51: record indicating a column of arrival cell numbers arranged in descending order of arrival cell number, 511, 512, ...: record of arrival cell numbers rearranged in descending order of value , 61: a record indicating a column of the number of second arrival cells arranged in descending order of the number of second arrival cells,
611, 612,...: Record of the second arrival cell number rearranged in descending order of value, 810: record indicating a column of arrival cell number arranged in descending order of arrival cell number, 811,
812,...: Arrival cell number record rearranged in descending order of arrival cell numerical value, 820: Record indicating a column having a frequency lower than the arrival cell number threshold, 821, 822,.
.. When the records 811, 812,... Are regarded as the thresholds, records having a frequency less than the threshold value of the arrival cell number (“number 42”), 831, 832,.
Logarithm of 1,822,..., 841,842,.
"Equation 43" for each record 811, 812, ...
, 851, 852,...:
For each record 811, 812,...
, 861, 862,...:
"Equation 45" for each record 811, 812, ...
, 871, 872,...:
"Equation 46" for each record 811, 812, ...
91: a record indicating a column of the number of arrival cells arranged in descending order of the number of arrival cells, 911, 9
12,...: Records of the number of arrival cells rearranged in descending order of the arrival cell numerical value, 92: Records indicating columns of the frequency of exceeding the threshold value of the number of arrival cells, 921, 922,. ,..., The record of the frequency of exceeding the threshold value of the number of arrival cells when the
923,...: Logarithm of each record 921, 922,.
912,..., Records obtained by the formula of Equation 33 for each record 911, 912,.
.., 952,..., Each record 911, 912,.
962,...: Records obtained by the formula of “Expression 35” for each record 911, 912,.
.., 972,...: Records obtained by the formula of “Expression 36” for each record 911, 912,.
0: Record indicating the column of the number of arrival cells arranged in descending order of the second arrival cell number, 1011, 1012,...: Record of the arrival cell number rearranged in descending order of the arrival cell number, 1020: Arrival cell number Records indicating columns of the threshold excess frequency, 1021, 1022,...: Each record 1
Record of the number of arrival cells exceeding the threshold when 011, 1012,...
032,...: Each record 1021, 1022,.
· Logarithms of 1041, 1042, ...:
For each record 1011, 1012,...
8, 1051, 1052
..: Records obtained by the formula of “Expression 39” for each of the records 1011, 1012,.
062, ...: Each record 1011, 1012, ...
Records obtained by the formula of “Formula 40” for 10
71, 1072,...: Each record 1011, 101
, 100a, 100b: cell transmission device, 151: cell counting unit, 152: table, 153: threshold registration unit,
154: excess number counting section, 155: table, 156:
Frequency calculation unit, 157: set capacity calculation unit, 151A, 15
2A: first Q length threshold, 151B, 152B: second
Q length threshold value, 400a: setting device, 401a, 50
1,601: communication line, 500: external device, 510: storage device, 600: virtual path capacity setting device, 1100: table, 1110: record sequence (counting cycle), 1111
1112: record (counting cycle), 1120: record string (number of arrival cells), 1121, 1122: record (number of arrival cells), 1130: record string (number of times the first Q length threshold has been exceeded), 1131, 1132: Record (first Q length threshold value excess count), 1140: record sequence (second
, 1141, 1142: record (second Q-length threshold exceeded), 1210: record sequence (counting cycle), 1220: record sequence (number of arrival cells), 1211, 1212 : Record (counting cycle), 1
221, 1222: record (number of arrival cells), 131
0: record sequence (counting cycle), 1320: record sequence (number of arrival cells), 1311, 1312: record (counting period), 1321, 1322: record (number of arrival cells), 1330: record sequence (exceeding arrival threshold) 1331, 1332: record (number of arrival threshold excess), 1340: record sequence (number of arrival cells), 13
41, 1342: Record (number of arrival cells), 1350:
Record sequence (number of arrival threshold excess), 1351, 13
52: record (number of arrival threshold excess), 1410:
.., Records indicating arrival cell number columns arranged in descending order of arrival cell number, 1411, 1412,...
420: Record indicating a column having a frequency lower than the threshold value of the number of arrival cells, 1421, 1422,...
, 1431, 1432,..., Each record 1
Logarithms of 421, 1422, ..., 1441, 144
,...: Records obtained by the formula of “Equation 48” for each record 1411, 1412,.
1, 1452,...: Each record 1411, 141
.., 1461, 1462,...: Each record 141
, 1471, 1472,..., 1471, 1472,...: Records obtained by the equation 51 for each record 1411, 1412,. Record 1500: ATM virtual path capacity setting device, 1501: communication line.

Continuation of front page (56) References JP-A-8-191309 (JP, A) JP-A-11-243398 (JP, A) JP-A-2001-251304 (JP, A) JP-A-11-122267 (JP, A A) JP-A-11-155492 (JP, A) JP-A-6-209329 (JP, A) JP-A-7-250088 (JP, A) JP-A-2000-295231 (JP, A) JP-A 8- 237261 (JP, A) (58) Field surveyed (Int. Cl. ⁷ , DB name) H04L 12/28

Claims (8)

(57) [Claims] 1. The capacity of an ATM virtual path set between ATM exchanges in an asynchronous transfer mode network (ATM network) is
An ATM virtual path capacity setting method of an ATM virtual path capacity setting apparatus for calculating a cell loss target rate, wherein the number of cells arriving at a cell transmission waiting buffer associated with the ATM virtual path (first arriving cell) Number) for each predetermined period T and store it in the first storage means, and read out the storage contents of the first storage means and read out the content of the cycle kT (k: an integer of 2 or more) ) Is the total number of cells
A second step of obtaining a value (a second number of arriving cells) divided by the following formula, and storing the value in the second storage means, and the first storage means in a predetermined cycle T ′ (T ′> T). Is read out, and for each first arrival cell number, a value obtained by summing the respective first arrival cell numbers equal to or greater than the cell number is obtained by summing all the first arrival cell numbers in the cycle T ′. A divided value (first arrival cell number threshold excess frequency) is obtained, and a first logarithmic threshold excess frequency obtained by taking a logarithm of the first arrival cell number threshold excess frequency is stored in a third storage. Reading the stored contents of the first storage means in the third step of storing in the means, and in the cycle T ', and for each first arrival cell number, for each first arrival cell number,
Is calculated by dividing the sum of the numbers of cells arriving at the first time by the sum of the numbers of all first cells arriving in the period T '(frequency less than the threshold value of the first number of cells arriving). A fourth step of storing, in a fourth storage means, a first frequency lower than a logarithmic threshold obtained by taking a logarithm of a frequency lower than the cell number threshold, and storing the second frequency in the second storage means in the cycle T ′ The content is read out, and for each second arrival cell number, a value obtained by summing the number of second arrival cells equal to or greater than the number of cells is divided by the total number of all second arrival cells in the period T ′. Value (second arrival cell number threshold value exceeding frequency) is obtained, and the second logarithmic threshold value exceeding frequency obtained by taking the logarithm of the second arrival cell number threshold value exceeding frequency is stored in the fifth storage means. A fifth step of storing, and the period T ′
Then, the contents stored in the second storage means are read out,
, The value obtained by dividing the sum of the number of second arrival cells less than the number of cells by the sum of the number of all second arrival cells within the period T ′ (the second arrival cell) A sixth step of obtaining a frequency of the second less than the threshold value obtained by taking a logarithm of the second less than the threshold value of the number of arrival cells in a sixth storage means; Reading the first logarithmic threshold excess frequency from the third storage means, obtaining a first estimated logarithmic threshold excess frequency approximated by a predetermined m-th order equation, The second logarithmic threshold value exceeding frequency is read out from the above to obtain a second estimated logarithmic threshold value exceeding frequency approximated by the m-th order expression, and the second estimated logarithmic threshold value exceeding frequency and the first Calculates the decay rate of the complementary distribution of the number of cells arriving per unit time based on the estimated log threshold frequency. And reading a frequency less than the first logarithmic threshold value from the fourth storage means, obtaining a frequency less than a first estimated logarithmic threshold value approximated by the m-th order expression, The second logarithmic frequency less than the second logarithmic threshold value is read from the storage unit of No. 6,
Is calculated based on the second estimated logarithmic threshold frequency and the first estimated logarithmic threshold frequency, the rate of increase of the distribution of the number of cells arriving per unit time. And a cell arriving per unit time obtained in the seventh step, the rate of increase in the distribution of the number of cells arriving per unit time obtained in the eighth step. An ATM virtual path capacity setting method, wherein an ATM virtual path capacity that satisfies the cell loss rate is obtained using an attenuation rate of a complementary distribution of numbers. 2. The ATM virtual path capacity setting method according to claim 1, wherein the storage contents of said first storage means are read out, and based on the number of arrival cells counted for each cycle T,
A ninth step of calculating an average (average cell rate) of the number of arriving cells in the cycle T ′, and a variance of the number of arriving cells in the cycle T ′ based on the average cell rate and the number of arriving cells counted in the cycle T. A tenth step to be obtained, an attenuation rate of the complementary distribution obtained in the seventh step, an increase rate of the distribution obtained in the eighth step, and the above obtained in the ninth and tenth steps. Based on the average cell rate and the variance of the number of arriving cells, the peak cell rate when the cell arrival process is regarded as an ON-OFF process, and from ON to OFF and OFF every time (time slot) which is the reciprocal of the peak cell rate. An eleventh step of obtaining the probabilities a and d of transition from ON to ON,
An ATM virtual path capacity setting method, wherein an ATM virtual path capacity that satisfies the target cell loss rate is obtained using the peak cell rate and the transition probabilities a and d. 3. The capacity of an ATM virtual path set between ATM exchanges in an asynchronous transfer mode network (ATM network) is
An ATM virtual path capacity setting method of an ATM virtual path capacity setting apparatus for calculating a cell loss target rate, wherein the number of cells arriving at a cell transmission waiting buffer associated with the ATM virtual path (first arriving cell) ) Is counted for each predetermined period T and stored in the first storage means, and the storage contents of the first storage means are read out and read for each period kT (k: an integer of 2 or more). Calculating a value obtained by dividing the total number of cells by k (second number of arriving cells) and storing the obtained value in the second storage means; and a predetermined period T ′ (T ′>
In T), the storage contents of the first storage means are read out, and for each first arrival cell number, a value obtained by summing up the respective first arrival cell numbers equal to or greater than the cell number is used in all of the periods T ′. Obtaining a value (first arrival cell number threshold value exceeding frequency) divided by the total of the first arrival cell number of the first and second cells, and a first logarithm of the first arrival cell number threshold exceeding frequency. Storing the logarithmic threshold value exceeding frequency in the third storage means, and reading out the storage contents of the first storage means in the cycle T ', and for each first arrival cell number, Calculating a value obtained by dividing a value obtained by summing the number of first arriving cells less than the total number of all first arriving cells in the period T ′ (frequency less than a threshold value of the first arriving cell number) And storing the first frequency less than the logarithmic threshold value obtained by taking the logarithm of the frequency less than the first arrival cell number threshold value as fourth storage means. Storing the contents of the second storage means in the cycle T ′, and for each second arrival cell number, sum the value of the total number of second arrival cells equal to or greater than the cell number. Calculating a value (frequency of exceeding the threshold value of the second arrival cell number) divided by the total number of all the second arrival cells in the period T ′; Storing the second logarithmic threshold value exceeding frequency obtained by taking the logarithm in the fifth storage means, and reading out the storage contents of the second storage means in the cycle T ′, For each number, a value obtained by dividing the sum of the numbers of the second arrival cells less than the number of cells by the sum of the numbers of all the second arrival cells in the period T ′ (the second arrival cell number threshold) And a second logarithmic threshold having a logarithm of a frequency lower than the second arrival cell count threshold. Storing a value less than the frequency in the sixth storage means, said third
Reading the first logarithmic threshold excess frequency from the storage means, and obtaining a first estimated logarithmic threshold excess frequency approximated by a predetermined m-th order equation; and Reading out a second logarithmic threshold excess frequency and obtaining a second estimated logarithmic threshold excess frequency approximated by the m-th order equation; Obtaining a decay rate of a complementary distribution of the number of cells arriving per unit time based on the estimated logarithmic threshold exceeding frequency;
Reading the frequency less than the logarithmic threshold value and calculating a first estimated logarithmic value less than the threshold value approximated by the m-th order expression; and reading the frequency less than the second logarithmic threshold value from the sixth storage means. Reading and calculating a second estimated logarithmic threshold frequency lower than the m-th order expression, and based on the second estimated logarithmic threshold frequency and the first estimated logarithmic threshold frequency, Calculating the rate of increase in the distribution of the number of cells arriving per unit time; reading the contents stored in the first storage means and arriving in the cycle T ′ based on the number of cells arriving counted in the cycle T; Obtaining an average of the number of cells (average cell rate); obtaining a variance of the number of cells arriving in the cycle T ′ based on the average cell rate and the number of arriving cells counted in each cycle T; A peak cell rate when the cell arrival process is regarded as an ON-OFF process based on the decay rate of the distribution, the rate of increase of the distribution, and the variance of the average cell rate and the number of arrival cells, and the time of the reciprocal of the peak cell rate. From ON to OFF and OFF for each (time slot)
Determining the probabilities a and d for transitioning from ON to ON, and determining the ATM virtual path capacity that satisfies the target cell loss rate based on the peak cell rate, the transition probabilities a and d, and the target cell loss rate. An ATM virtual path capacity setting method, comprising: 4. The ATM virtual path capacity setting method according to claim 1, wherein the storage contents of said first storage means are read out, and said first arriving cell for each cycle T is read out. Rearranging the numbers in a descending order and storing the same in a seventh storage unit; reading out the storage contents of the second storage unit and rearranging the second arrival cell number in each cycle T in a descending order, 8 in the storage means, and when the frequency of exceeding the first arrival cell number threshold value and the frequency of being less than the first arrival cell number threshold value are determined, the first and second storage means are replaced with the first storage means. When the content stored in the seventh storage means is read out and the frequency of exceeding the second threshold value of the number of arrival cells and the frequency of being less than the second threshold value of the number of arrival cells are calculated, To read out the contents stored in the eighth storage means An ATM virtual path capacity setting method, characterized in that: 5. The capacity of an ATM virtual path set between ATM exchanges in an asynchronous transfer mode network (ATM network),
An ATM virtual path capacity setting device for calculating a cell loss target rate, wherein a number of cells arriving at a cell transmission waiting buffer associated with the ATM virtual path (first arriving cell number) is predetermined. Means for counting and storing in the first storage means for each cycle T, and reading the storage contents of the first storage means, and dividing the total number of cells for each cycle kT (k: an integer of 2 or more) by k Means for obtaining the calculated value (second number of arriving cells) and storing it in the second storage means, and reading out the storage contents of the first storage means at a predetermined cycle T '(T'> T). For each first arrival cell number, a value obtained by dividing the total value of the first arrival cell numbers equal to or greater than the cell number by the total number of all the first arrival cells in the period T ′ (the first 1 arrival cell number threshold value exceeding frequency), and the logarithm of the first arrival cell number threshold value exceeding frequency is calculated as Means for storing the first logarithmic threshold value exceeding frequency in a third storage means;
Then, the contents stored in the first storage means are read out,
, The value obtained by dividing the sum of the numbers of the first arrival cells smaller than the number of cells by the sum of the numbers of all the first arrival cells in the period T ′ (the first arrival cell) Means for obtaining a frequency lower than the first threshold value obtained by calculating a logarithm of the frequency lower than the first arrival cell number threshold value in a fourth storage means; At T ′, the storage contents of the second storage means are read out, and for each second arrival cell number, a value obtained by summing the number of second arrival cells equal to or greater than the number of cells is calculated for all of the periods T ′. Is obtained by dividing by the total number of second arrival cells (the frequency of exceeding the second arrival cell number threshold),
Means for storing in a fifth storage means a second logarithmic threshold value exceeding frequency obtained by taking a logarithm of the second arrival cell number threshold value exceeding frequency; Is read out, and for each second arriving cell number, a value obtained by summing the number of the second arriving cells less than the cell number is calculated as the sum of all the second arriving cell numbers in the cycle T ′. A divided value (frequency below the second arrival cell number threshold) is obtained, and the second frequency below the logarithmic threshold obtained by taking the logarithm of the frequency below the second arrival cell number threshold is stored in the sixth storage. Means for storing in the means;
Means for reading the first logarithmic threshold excess frequency from the storage means, and obtaining a first estimated logarithmic threshold excess frequency approximated by a predetermined m-th order equation; and Means for reading out the second logarithmic threshold excess frequency and obtaining a second estimated logarithmic threshold excess frequency approximated by the m-th order expression, the second estimated logarithmic threshold excess frequency and the first Means for obtaining a decay rate of a complementary distribution of the number of cells arriving per unit time based on the estimated logarithmic threshold exceeding frequency, and reading the frequency lower than the first logarithmic threshold from the fourth storage means, means for obtaining a frequency below a first estimated logarithmic threshold value approximated by an m-th order expression; and a second method for reading out the frequency less than the second logarithmic threshold value from the sixth storage means and approximating the frequency by a m-th order expression. Means for determining the frequency below the estimated logarithmic threshold of Means for obtaining a rate of increase in the distribution of the number of cells arriving per unit time based on the second estimated logarithmic threshold frequency and the first estimated logarithmic threshold frequency. An ATM virtual path capacity that satisfies the cell loss rate, using a rise rate of a distribution of the number of cells arriving at the unit and an attenuation rate of a complementary distribution of the number of cells arriving per unit time. Virtual path capacity setting device. 6. The ATM virtual path capacity setting device according to claim 5, wherein the storage contents of said first storage means are read out, and based on the number of arrival cells counted for each cycle T,
Means for calculating an average (average cell rate) of the number of arriving cells in the cycle T ′;
Means for calculating the variance of the number of arriving cells in the cycle T ′ based on the number of arriving cells counted for each, and the decay rate of the complementary distribution, the increase rate of the distribution, and the variance of the average cell rate and the number of arriving cells. Based on the cell arrival process
The peak cell rate when it is regarded as the OFF process, and the reciprocal of the peak cell rate for each time (time slot)
Probability of transition from N to OFF and from OFF to ON a,
means for obtaining an ATM virtual path capacity that satisfies the target cell loss rate using the peak cell rate and the transition probabilities a and d.
TM virtual path capacity setting device. 7. The capacity of an ATM virtual path set between ATM exchanges in an asynchronous transfer mode network (ATM network),
An ATM virtual path capacity setting device for calculating a cell loss target rate, wherein a number of cells arriving at a cell transmission waiting buffer associated with the ATM virtual path (first arriving cell number) is predetermined. A cell counting unit that counts each period T and stores it in the first storage unit; and reads out the storage content of the first storage unit, and reads out the period kT (k: an integer of 2 or more)
An average counting means for obtaining a value obtained by dividing the total number of cells for each by k (second number of arriving cells) and storing the obtained value in the second storage means; and a predetermined period T ′ (T ′> T) The storage contents of the first storage means are read out, and for each first arrival cell number, a value obtained by summing the respective first arrival cell numbers equal to or greater than the cell number is calculated for all the first arrival cells in the period T ′. A value obtained by dividing by the total number of arriving cells (first arriving cell number threshold excess frequency) is obtained.
A first excess frequency calculating means for storing a first logarithmic threshold excess frequency obtained by taking a logarithm of the arrival cell number threshold excess frequency in a third storage means; The storage contents of the storage means are read out, and for each first arrival cell number,
A value obtained by dividing the sum of the numbers of the first arrival cells less than the number of cells by the sum of the numbers of all the first arrival cells in the period T ′ (frequency below the first arrival cell number threshold) And storing the first frequency less than the logarithmic threshold value obtained by taking the logarithm of the frequency less than the first arrival cell number threshold value in the fourth storage means.
And the storage contents of the second storage means are read out in the period T 'and the number of second arrival cells equal to or greater than the number of cells are summed up for each second number of arrival cells. A value (frequency of exceeding the threshold value of the second arrival cell number) obtained by dividing the value by the sum of all the second arrival cell numbers in the period T ′ is obtained,
A second excess frequency calculating means for storing, in a fifth storage means, a second logarithmic threshold value excess frequency obtained by taking a logarithm of the second arrival cell number threshold value excess frequency; The second
Is read out, and for each second arriving cell number, a value obtained by summing the number of second arriving cells less than the cell number is used as the number of all second arriving cells in the period T ′. (Frequency below the second threshold for the number of arrival cells) is obtained by dividing the frequency below the second logarithmic threshold by taking the logarithm of the frequency below the second threshold for the number of arrival cells. A second less-than-frequency calculating means stored in the storage means of No. 6 and the first logarithmic threshold excess frequency from the third storing means, and a first estimation approximated by a predetermined m-th order equation A logarithmic threshold value exceeding frequency is obtained, the second logarithmic threshold value exceeding frequency is read out from the fifth storage means, and the second logarithmic threshold value approximated by the m-th order is obtained.
Of the number of cells arriving per unit time based on the second estimated logarithmic threshold excess frequency and the first estimated logarithmic threshold excess frequency. An attenuation factor calculating means for determining a rate, and reading the frequency less than the first logarithmic threshold value from the fourth storage means, obtaining a frequency less than a first estimated logarithmic threshold value approximated by the m-th order equation, The frequency less than the second logarithmic threshold value is read from the sixth storage means, and the second
Is calculated based on the second estimated logarithmic threshold frequency and the first estimated logarithmic threshold frequency, the rate of increase of the distribution of the number of cells arriving per unit time. Reading out the storage contents of the first storage means and calculating the average (average cell rate) of the number of arriving cells in the cycle T ′ based on the number of arriving cells counted in each cycle T. Average calculating means, variance calculating means for calculating the variance of the number of arriving cells in the cycle T ′ based on the average cell rate and the number of arriving cells counted for each cycle T, A cell arrival process is turned on based on the attenuation rate, the rise rate of the distribution obtained by the rise rate calculation means, the average cell rate obtained by the average calculation means, and the variance of the number of arrival cells obtained by the variance calculation means. A peak cell rate when as OFF process, the ON from OFF and OFF every time the inverse of the peak cell rate (time slot) O
N, a peak cell rate calculating means for calculating the probabilities a, d for transitioning to N, and the cell loss based on the peak cell rate, the probabilities a, d for transition, and the target cell loss rate determined by the peak cell rate calculating means. AT that satisfies the target rate
An ATM virtual path capacity setting device, comprising: ATM virtual path capacity calculation means for obtaining M virtual path capacity. 8. The ATM virtual path capacity setting device according to claim 5, wherein the storage contents of said first storage means are read out, and said first virtual path capacity setting device reads said first storage means for each of said periods T. A first rearranging unit that rearranges the number of arriving cells in a descending order and stores the same in a seventh storage unit, and reads out a storage content of the second storage unit, and reads the second arrival cell number in each cycle T And a second rearrangement unit that rearranges the first arrival cell number threshold value and the first arrival cell number threshold value lower frequency. Is obtained, the storage contents of the seventh storage means are read out instead of the first storage means, and the second arrival cell number threshold excess frequency and the second arrival cell number threshold are read out. When obtaining the less than frequency, the eighth storage is used instead of the second storage means. An ATM virtual path capacity setting device for reading out the storage contents of a storage means.