KR100257933B1 - Traffic Shaping Method for Compliance with Maximum Cell Rate in Asynchronous Transmission Mode Networks - Google Patents

Abstract

PURPOSE: A traffic shaping method for maintaining maximum cell transmission rate in ATM network is provided to improve an efficiency of ATM apparatus and network by generating only optimal cells satisfying a maximum cell transmission rate, in case that the output of the ATM apparatus is time multiplexed in TDMA. CONSTITUTION: An initial sequence set of a cell stream inputted from external and a rotation sequence set are created. The rotation sequence set includes the initial sequence sets. It is determined if there is a sequence merged with an optimal sequence among the rotation sequence set. If so, the output sequence and the rotation sequence are merged to generate a new output sequence. Otherwise, the control process detects a sequence merged with an optimal sequence among compatible sequences obtained by rotating collision slots of the rotation sequence, and merges the detected sequence with the output sequence.

Description

Traffic Shaping Method for Compliance with Maximum Cell Rate in Asynchronous Transmission Mode Networks

The present invention relates to a traffic shaping method for enabling an asynchronous transmission mode apparatus to comply with a maximum cell rate previously negotiated with a network when transmitting data to an asynchronous transmission mode network.

Asynchronous Transfer Mode (ATM) network supports various types of traffic such as voice, data, and video, and guarantees a certain level of quality of service (QoS) for these traffic.

In order to guarantee the quality of service (QoS) required by the asynchronous transmission mode network, the function of monitoring and controlling the traffic in the asynchronous transmission mode network is called traffic management. Traffic shaping is a function that adjusts to conform to the agreement negotiated between the network and the network.

During such traffic shaping, the peak cell rate (PCR), which is one of the contracts for which traffic to be transmitted is negotiated between the user and the network in an environment where the output of the asynchronous transmission mode device is multiplexed by a time division muitiplexing (TDM) method. ) To satisfy traffic parameters.

The maximum cell rate is a critical traffic parameter required in the main service category of an asynchronous transmission mode network such as a constant bit rate (CBR) or a variable bit rate (VBR), and its compliance is essential.

In addition, the asynchronous transmission mode network may discard or tag when a traffic source outputs a nonconforming cell for a pre-negotiated maximum cell rate (PCR) traffic parameter. Since the efficiency of the asynchronous transmission mode (ATM) device that outputs a cell can be reduced, there is a need for asynchronous transmission mode devices that generate only an optimal cell to increase the efficiency of the device and the network.

As a general technique for satisfying the conventional maximum cell rate, there is a buffering method, a spacing method, and the like.

The buffering method is a method of sequentially storing cells input to a traffic shaping apparatus in a buffer for a predetermined time and releasing them to satisfy a condition of a maximum cell rate.

Such a buffering method emits an output suitable for the maximum data rate negotiated flexibly, and there is a problem in that there is no cell loss but it is not efficient when the amount of traffic input to the traffic shaping device is large.

In the spacing method, the theoretical cell emission time is calculated by space so that the maximum delay variation is transmitted within the allowable value, and the input cell is discarded if the maximum delay variation is not transmitted within the allowable value.

The traffic shaping apparatus by the spacing method emits a cell that satisfies the maximum delay variation and satisfies the maximum cell rate, but there is a problem that information may be lost due to cell loss by discarding the cell in certain cases. .

Meanwhile, an asynchronous transmission mode apparatus for multiplexing a cell input from a conventional user and transmitting it to an asynchronous transmission mode network. The apparatus includes a multiplexer, a slot allocation ring, and a SONET (Synchronous Optical NETwork) / SDH (Synchronous Digital Hierarchy) framer. In such a device, a new cell allocation method is required because the user is assigned a method of allocating a cell to an entry of a slot allocation ring.

Therefore, the present invention devised to meet the demand as described above, the efficiency of the device and the network by generating only the best suitable cell that satisfies the maximum cell rate in the environment where the output of the asynchronous transmission mode device is multiplexed by the time division multiplexing method. The purpose of the present invention is to provide a traffic shaping method for improving the performance.

1 is a configuration example showing an asynchronous transmission mode device to which the present invention is applied.

2 is an explanatory diagram of sequence rotation and assigned time slot rotation used in the present invention.

3 is an explanatory diagram for sequence merging used in the present invention.

4A to 4C are flowcharts illustrating a traffic shaping method for observing a maximum cell rate in an asynchronous transmission mode network according to an embodiment of the present invention.

* Explanation of symbols for the main parts of the drawings

110: multiplexer 120: SONET / SDH framer

130: transmission module 140: slot assignment ring

The method of the present invention for achieving the above object, in the traffic shaping method applied to the asynchronous transmission mode device, generating an initial sequence set and a rotation sequence set of the initial sequence of the cell stream input from the outside, A first step of determining whether there is a sequence merged with the output sequence in an optimal state among the sequences in the set; A second step of generating a new output sequence by merging the output sequence and the rotation sequence if there is a sequence merged with the output sequence in an optimal state as a result of the determination of the first step; And a sequence capable of merging with the output sequence in an optimal state among compatible sequences obtained by slot rotation of collision slots of the rotation sequence if there is no sequence merged with the output sequence in an optimal state as a result of the determination of the first step. And a third step of merging the output sequence with the output sequence.

Hereinafter, a preferred embodiment according to the present invention will be described in detail with reference to FIGS. 1 to 4C.

1 is an exemplary configuration diagram illustrating an asynchronous transmission mode apparatus to which the present invention is applied.

As shown in the figure, the asynchronous transmission mode device, the multiplexer 110 for multiplexing and outputting a plurality of cells input from the outside, and the output of the multiplexer 110, the SONET (Synchronous Optical NETwork) / SDH (Synchronous SONET / SDH framer 120 converts and outputs a signal that satisfies Digital Hirarchy, and a transmission module 130 that converts and outputs the output of the SONET / SDH framer 120 into a signal suitable for an asynchronous transmission mode network. The slot allocation ring 140 controls the time division multiplexing process of the firearm 110.

The multiplexer 110 multiplexes a plurality of incoming cell streams to produce one output stream.

In this case, the multiplexer 110 performs a time division multiplexing function using the control logic of the slot allocation ring 140.

Meanwhile, each entry of the slot allocation ring 140 stores a pointer indicating an input terminal or a connection currently being multiplexed.

The multiplexer 110 binds one cell of the input stage indicated by the current active entry in the slot allocation ring 140 to the output stage, and then combines the currently active entry into the next entry in the slot allocation ring 140. Multiplexing is performed by repeating the above process by changing to.

Since the output stream of the multiplexer 110 has a ring-shaped control structure for allocating each time slot, a cell pattern having a predetermined time is repeatedly displayed.

In the asynchronous transmission mode network apparatus to which the present invention is applied, the number of entries occupied by each input in the slot allocation ring 140 for the maximum cell rate band allocation of each input is proportional to the maximum cell rate.

For example, if the transmission module 130 that transmits the final output cell stream to the SONET or SDH physical layer and transmits has a transmission rate performance of 155.52 Mbps and the total number of entries in the slot allocation ring 140 is 4800. The pure asynchronous transmission mode layer bandwidth excluding SONET / SDH frame load divided by the total number of entries in the slot allocation ring 140, approximately 149.76 Mbps / 4800 = 31.2 Kbps, is the minimum bandwidth allocation unit.

A minimum bandwidth allocation unit (Unit Bandwidth) UnitBW the asynchronous transmission mode device la and any entry connection is to be allocated to the slot allocation if it requires a predetermined bandwidth R _pcr ring 140, the number N _TE is then of ( Expressed as 1).

Symbol in Equation (1) above Is an indication that the nearest integer greater than the calculated value is taken if the calculated value in the symbol is not an integer.

The asynchronous transmission mode device allocates the desired maximum cell rate bandwidth by assigning an entry of N _TE , which is the result of Equation 1, to a specific input, but T _pcr = 1 / R _pcr , which is the time interval between transmission cells, and the maximum delay variation. In order to satisfy τ, the Cell Delay Variance Tolerance (CDVT), i.e., shaping to satisfy the Generic Cell Rate Algorithm (GCRA) algorithm for determining the suitability of the traffic, the allocation position must be properly placed within the slot allocation ring 140. do.

Meanwhile, definitions of terms used in the present invention are as follows.

First, an input sequence is a pattern of a finite number of time slots in which cells of a specific input stage appear repeatedly in an output stream.

The output sequence is the time slot pattern at the output when all input streams present in the system are multiplexed and included in the output stream.

In the traffic shaping apparatus of the asynchronous transmission mode to which the present invention is applied, the number of internal time slots of the input sequence and the output sequence is equal to the number of entries of the slot allocation ring 140, and the output stream has a form in which the output sequence is repeated.

In order to satisfy the maximum cell rate (PCR) related traffic characteristics of each input stream, the N _TE time slots described above must be allocated in the input sequence of the input stream (required). Bandwidth allocation), each allocated time slot should be properly arranged to maintain the interval of T _pcr within the margin of error of τ.

If the total number of time slots present in the input sequence is N _seq , then the input sequence containing time slots allocated at intervals of (N _seq / N _TE ) ideally meets the above conditions. Satisfies.

However, since the result of Equation 1 is not an integer value, among the total N _TE time slots, b = (N _seq mod N _TE ) time slots are T _b. = Are allocated with a time slot interval of and the remaining a = (N _TE -b) time slots are T _a = These sequences are considered ideal sequences by assigning them at intervals of two, which is called the initial sequence.

Symbol above Is a symbol that discards all the decimal points from the calculated value and obtains the integer value.

A time slot whose interval with the time slot transmitted immediately before time in the allocated time slots in the initial sequence is T _a is called a type A time slot and T _b Is called a type B time slot.

The set of all initial sequences that can be obtained as a type A time slot and b type B time slots is called an "initial sequence set", and the number of elements in the set is N _TE ! / ( a! * b!)

For example, an initial sequence set S1 having three type A time slots and two type B time slots is (AAABB, AABAB, AABBA, ABBAA, ABABA, ABAAB, BBAAA, BABAA). , BAABA, BAAAB).

A sequence Si created by right rotating an initial sequence S by i (0 ≦ i ≦ T _b ) is called a rotation sequence of an initial sequence S. The set of all sequences that can be obtained using such sequence rotation is obtained. Referred to as "rotation sequence set".

The reason why the value of i is limited to 0 ≦ i ≦ T _b is that when any sequence in the initial sequence set is rotated by more than T _b, it is matched with another initial sequence (BBAAA, which is an element of the initial sequence set described above). If you turn right by 2 * T _b, it is the same as other initial sequence AABBA).

Allocated time slots in the initial sequence, Δ =, is the number of transforms of the maximum delay variation tolerance (CDVT) in time slots from the current position, that is, the theoretical cell emission time (TET). Maximum Delay Variance Allowance (CDVT) / Time corresponding to one time slot Left turn is possible.

The maximum number of time slots in which an assigned specific time slot can turn left within the maximum delay variation tolerance is indicated by the mobility M of the time slot.

Then, turning any assigned time slot left at the current position is called time slot rotation. In this case, the mobility of the moved time slot is reduced by the number of movements, but the mobility of the time slot adjacent to the time slot is increased by the number of movements.

The new sequence Sc, generated by rotating the internally allocated time slots of any sequence S, is called the compatible sequence of sequence S.

2 is an explanatory diagram of sequence rotation and assigned time slot rotation used in the present invention.

As shown in Fig. 2, (a) represents the first input sequence, (b) represents the first rotation sequence rotated by the first input sequence (a), and (c) represents the second input sequence. (d) represents the first compatible sequence rotated the assigned time slot of the second input sequence (c), (e) represents the third input sequence, and (f) represents the third input sequence (e) The assigned time slot represents the rotated second compatible sequence, and the number displayed in each time slot represents the mobility.

The first rotation sequence b is a rotation sequence in which the first input sequence a is rotated by two time slots.

The first compatible sequence (d) rotates the first and second time slots of the second input sequence (c) by one time slot to the left so that the mobility is 1, respectively, and the second and fourth times The slot is fixed but the mobility changes to 3 due to the movement of the first and third time slots, respectively.

The second compatible sequence (f) rotates the first time slot of the third input sequence (e) by one time slot to the left so that the mobility should be 2, but rotates the second time slot by two time slots. For this reason, the mobility becomes 4, and the mobility of the second time slot is rotated by two time slots while the mobility is 4 by the rotation of the first time slot, so the final mobility becomes 2.

3 is an explanatory diagram for sequence merging used in the present invention.

Creating two new sequences by nesting two sequences, S1 and S2, is called "sequence merging."

If time slots of the same position in S1 and S2 are allocated at the same time, S1 and S2 cannot be merged and conflict with time slots that disable sequence merging. time slot).

Meanwhile, an evaluation criterion is needed to determine which result is more optimal among the results obtained because the sequence merging that satisfies the maximum cell rate (PCR) required by each connection is possible in various arrangements.

These criteria are expressed as formulas for ease of implementation and are called cost functions.

Considerations for obtaining the cost function proposed in the present invention are as follows.

1. Allocate a time slot so that the cells to be transmitted are as close as possible to the theoretical cell release time (TET).

2. Ensure that time slots allocated for ease of addition of connections are not dense.

3. Consider the unique traffic characteristics of the connections.

Although the asynchronous transmission mode network allows cells with variations up to the maximum delay variation tolerance, there is clearly a qualitative difference between traffic where the cell's transmission time is close to the theoretical cell emission time (TET) and traffic that is not. As the (ATM) terminal guarantees as many connections as possible, it is necessary to consider the items 1 and 2 above.

In addition, the item of 3 is to give different weights in consideration of the unique characteristics of each connection.

For example, in a connection with Δ = 10, the maximum delay variation tolerance (CDVT), the time slot shifted by one time slot from the theoretical cell emission time (TET) and one time in the connection with Δ = 1. The result of moving by the slot is the same in that the cell has moved by one time slot, but it can be estimated that the latter case is relatively out of the theoretical cell emission time (TET).

To formulate the three elements above, we define the following cost function:

If any time slot is in the i th position and the theoretical cell emission time (TET) of this time slot is j th, then the suitability cost of this time slot is defined as | ji | The suitability cost is the cost of the sum of the suitability costs of all assigned time slots in any input sequence S divided by the maximum delayed shift tolerance (CDVT) Δ, i.e., C_COST (S) = Σ (each allocated slot). Is the suitability cost of the input sequence S.

The distribution cost is the total number of time slots allocable by time slot rotation among the unassigned time slots in sequence S divided by the number n of connections that S contains. It is defined as cost D_COST (S).

Conformance costs represent the closeness of the allocated time slots to the theoretical cell emission time (TET), and the connections with large maximum delay variation tolerances (CDVTs) are reflected at a relatively low cost compared to small connections. Satisfy considerations 1 and 3.

The distributed cost represents the possibility of resolving a conflict time slot when a new connection is added and reflects consideration 2 above. The variance cost should be subtracted when included in the overall cost function, since the higher the value, the better the allocation quality.

The cost of generating a new output sequence S by merging the input sequence Si into an output sequence representing existing active connections is defined by the following polynomial:

COST (S) = α * C_COST (Si)-D_COST (S)

The α coefficient in the above cost function is a constant having a different value depending on whether shaping quality of currently active connections or weighting of new connections is added.

As shown in the figure, (a) represents a fourth rotation sequence, (b) represents a third compatible sequence, (C) represents a first output sequence having two connections, and (d) represents A third output sequence c with three compatible sequences b and two connections represents a merged second output sequence.

Now, the sequence merging will be described in detail.

The fourth rotation sequence (a) is one of the rotation sequences of the newly added connection, and the number indicated in the time slot represents the mobility.

The third compatibility sequence (b) is generated by rotating the fourth rotation sequence (a) within a mobility allowance limit by rotating a time slot that collides with the allocation time slot in the first output sequence (c) to be merged among the allocation time slots. As an example of a compatibility sequence, the number indicated in the time slot represents the mobility after rotation of the assigned time slot.

The time slot conformance cost of the third compatibility sequence (b) is 0 for the first time slot, 0 for the second time slot, 2 for the third time slot, 1 for the fourth time slot, and 5 for the fifth time slot. 0.

Therefore, the suitability cost is 1.5 when the total time slot suitability cost 3 is divided by the maximum delay variation tolerance Δ = 2.

The x mark of the time slot of the second output sequence d indicates the assignable time slot using time slot rotation, and has seven assignable time slots.

Thus, the distributed cost is 2.3, which is the allocable time slot 7 divided by the number of connections 3.

Therefore, the cost COST (S) when the second output sequence is generated becomes 12.7 when the cost of multiplicity of 1.5 is multiplied by α = 10 when α = 10, minus the variance cost 2.3.

4A to 4C are flowcharts illustrating a traffic shaping method for observing maximum cell rate in an asynchronous transmission mode network according to an embodiment of the present invention.

As shown in the figure, first the output sequence (OSEQ: Output Sequence) and the temporary storage sequence SEQ are initialized (400), and then in the standby state (401) and then the bandwidth R _pcr and the maximum indicated by the maximum cell rate from the outside. In operation 402, it is determined whether a connection request ConReq (R _pcr, Δ) including information of the delay variation tolerance Δ is received.

As a result of determining whether a connection request has been received, if the connection request is not received, it repeats from the waiting state 401, and if the connection request is received, the initial sequence set SI = ｛SI ₁ , SI ₂ , SI ₃ ,. , SI _j ) is generated, and the mobility for the assigned slots of the generated initial sequences is calculated (403).

Later, the variable COST is set to infinity, the variable J is set to 1 (404), and it is determined whether the variable J is equal to or smaller than j (405).

As a result of determining whether the variable J is equal to or smaller than j, if the variable J is equal to or smaller than _j, the rotation sequence set S = SS ₁ , S ₂ ,. , S _m 하고 and compute the mobility for the assigned slots of the generated rotation sequences (406).

Subsequently, after setting variable M to 1 (407), it is determined whether variable m is greater than or equal to M (408), and if m is not equal to or greater than M, after adding 1 to J (409), variable j Repeat from step 405 to determine whether is equal to or greater than J, and if m is equal to or greater than _M , whether the variable COST is greater than the merge cost value MergeCost (OSEQ, S _M ) of the output sequence OSEQ and rotation sequence S _M. Determine (410).

As a result of determining whether the variable COST is greater than the merge cost value of the output sequence and the rotation sequence, if the variable COST is large, the merge cost value MergeCost (OSEQ, S _M ) is placed in the variable COST, and S _M is placed in the SEQ, which is a variable for storing the temporary sequence (411).

Subsequently, after adding 1 to the variable M (412), the process is repeated (408) to determine whether the number M is equal to or smaller than m.

As a result of determining whether the variable COST is greater than the merging cost value of the output sequence and the rotation sequence, if it is not large, 1 is added to the variable M (412), and then the process of determining whether the variable M is equal to or smaller than m is repeated (408). .

On the other hand, as a result of determining whether j is equal to or greater than J, if j is not equal to or greater than J, it is determined whether the variable COST is not infinite (413).

As a result of judging whether the variable COST value is not infinite, if the variable COST is not infinite, the output sequence OSEQ and the temporary storage sequence SEQ are merged (Merge (OSEQ. SEQ)) and the new output sequence is placed in OSEQ (433). Repeat from state 401.

As a result of determining whether the variable COST value is infinity, if the variable COST value is infinity, J is set to 1 (414), and it is determined whether j is equal to or greater than J (415).

As a result of judging whether j is greater than or equal to J, if j is not greater than or equal to J, it is determined whether the variable COST is not infinite (432), and if the variable COST is infinity, the connection request ConReq (R _pcr, Δ) is rejected. After (434). Iterates from the wait state 401, and if the cost function value is not infinite, merges the output sequence OSEQ and the temporary storage sequence SEQ (Merge (OSEQ, SEQ)), puts a new output sequence in the OSEQ (433), and then waits. Repeat from (401).

On the other hand, as a result of judging whether j is equal to or greater than J, when j is equal to or greater than _J, the rotation sequence set S = ｛S ₁ , S ₂ ,. After generating S _m , and calculating mobility for the allocation slots of the generated rotation sequences (416), the variable M is set to 1 (417) and it is determined whether m is equal to or greater than M (418). .

As a result of determining whether m is greater than or equal to M, if m is not equal to or greater than M, after adding 1 to J (419), the process is repeated from the determination process 415 to determine whether j is greater than or equal to J, and m is M Greater than or equal to a set of slots C = ｛C1, C2,... That collide with the output sequence OSEQ in rotation sequence S _M. After generating Cn '(420), the variable N is set to 1 (421).

Subsequently, it is determined whether n is greater than or equal to N (422), and if n is greater than or equal to N, it is determined whether the variable COST is greater than the merge cost value MergeCost (OSEQ, S _M ) of the output sequence OSEQ and compatible sequence S _M. Determine (425).

Determination of whether the variable COST is greater than the merge cost value MergeCost (OSEQ, S _M ) of the output sequence OSEQ and the compatible sequence S _M , if greater, place the merge cost value in the variable COST and the compatible sequence S _M in the output sequence OSEQ (426 After adding 1 to the variable M (427), repeat the process of determining whether m is greater than M (418), and if it is not large, add 1 to the constant variable M (427), then determine whether m is greater than M. The determination process is repeated from 418.

On the other hand, as a result of determining whether n is equal to or greater than N, if n is not equal to or greater than N, the variable K is set to 1, the mobility of the collision time slot C _N is calculated and placed in the variable L (423), and L is K Determine if greater than or equal (424), add 1 to M if not equal to (427), then repeat from decision process 418 if m is greater than or equal to M, if conflict time slot C _N It is determined whether there is no collision with the allocated time slots of the output sequence OSEQ at the slot rotated by this K (428).

As a result of determining whether there is no collision with the allocated time slots of the output sequence OSEQ, if there is a collision, after increasing K by 1 (429), it is repeated from the determination process 424 of whether L is equal to or greater than K, and the collision is repeated. If none, change the allocation position of the collision time slot C _{N to} the slot rotated position by K and recalculate the mobility of the allocation slots in S _M (430), then increase the variable N by one and determine whether n is equal to or greater than N. The determination process is repeated from 422.

The present invention described above is capable of various substitutions, modifications, and changes without departing from the spirit of the present invention for those skilled in the art to which the present invention pertains. It is not limited to the drawing.

The present invention as described above, because the transmitted data always meets the maximum cell rate traffic parameters previously negotiated with the network, it is highly likely that the data is safely delivered to the destination, resulting in improved network and user traffic management capability Reliable communication is possible and it is very efficient in that there is no load on the device in the data transmission state because the calculation is performed only at the initial connection establishment stage.

Claims (4)

In the traffic shaping method applied to the asynchronous transmission mode device, A first step of generating an initial sequence set of an externally input cell stream and a rotation sequence set of initial sequences, and determining whether there is a sequence merged with an output sequence in an optimal state among the sequences in the generated rotation sequence set; A second step of generating a new output sequence by merging the output sequence and the rotation sequence if there is a sequence merged with the output sequence in an optimal state as a result of the determination of the first step; And As a result of the determination in the first step, if there is no sequence merged with the output sequence in an optimal state, a sequence that can be merged with the output sequence in an optimal state among compatible sequences obtained by slot rotation of the collision slots of the rotation sequence is determined. Third step of finding and merging with output sequence Traffic shaping method for observing the maximum cell rate of the asynchronous transmission mode network made, including. The method of claim 1, The first step is, A fourth step of initializing the output sequence and the temporary storage sequence variable and waiting to receive a connection request; A fifth step of generating an initial sequence set for an externally input cell stream and calculating mobility for an allocation slot of an initial sequence, and then setting an infinite variable for cost value storage; Obtaining a rotation sequence set for each initial sequence of the initial sequence set and calculating mobility for the assigned slots of the rotation sequence; And A seventh step of obtaining a merge cost of an output sequence and a rotation sequence and determining whether the obtained cost is infinite Traffic shaping method for observing the maximum cell rate of the asynchronous transmission mode network made, including. The method of claim 2, The seventh step, An eighth step of selecting any rotation sequence from the rotation sequence set; A ninth step of calculating a merging cost value of the selected rotation sequence and output sequence; A tenth step of determining whether a merge cost value is smaller than a variable for storing a predetermined cost value; As a result of the determination in the tenth step, if the merging cost value is smaller than the variable value for storing the predetermined cost value, the variable value for storing the predetermined cost value is updated to the merging cost value and then repeatedly performed from the eighth step. Eleventh step; A twelfth step of repeating from the eighth step if the merging cost value is larger than a variable value for storing a predetermined cost value as a result of the determination of the tenth step; And The thirteenth step of determining whether the variable value for storing the cost value is infinite Traffic shaping method for observing the maximum cell rate of the asynchronous transmission mode network made, including. The method according to any one of claims 1 to 3, The third step, A fourteenth step of generating a set of rotation sequences of the initial sequence and calculating mobility for the assigned slots of the rotation sequences; Obtaining a set of allocation time slots in which collisions with an output sequence occur among the allocated time slots of the generated rotation sequence; A sixteenth step of moving the allocation position of each collision time slot to a time slot in which collision can be avoided within a range of mobility; A seventeenth step of calculating a merging cost value and determining whether it is smaller than a variable value for storing a predetermined cost value; As a result of the determination in step 17, when the merge cost value is smaller than the variable value for storing the predetermined cost value, the variable sequence for storing the cost value is changed to the merge cost value and a new sequence of the temporary storage sequence variable is obtained. 18th step to switch to; A nineteenth step of merging and outputting the temporary storage sequence and the output sequence; And As a result of the determination of step 17, when the merging cost value is larger than a variable value for storing a predetermined cost value, step 20 of rejecting the connection Traffic shaping method for observing the maximum cell rate of the asynchronous transmission mode network made, including.