WO2006068411A1

WO2006068411A1 - A lock-based call admission control method in which non-real-time services share bandwidth

Info

Publication number: WO2006068411A1
Application number: PCT/KR2005/004420
Authority: WO
Inventors: Dong-Hoi Kim; Pyeong-Jung Song; Kyung-Koo Jun; Seok-Hoon Kang
Original assignee: Electronics And Telecommunications Research Institute
Priority date: 2004-12-21
Filing date: 2005-12-21
Publication date: 2006-06-29
Also published as: EP1832023A1; KR100578303B1; EP1832023A4

Abstract

The present invention relates to a lock-based call admission control method in which non-real-time services for optimizing system performance and satisfying quality of service (QoS) for multi-services share bandwidth in a next generation mobile communication system. In the method, a) respective calls according to multi-service classes of real-time or non-real-time services are received, b) bandwidth reconfiguration for borrowing bandwidth from low priority services or lending the bandwidth to high priority services according to resource availability is performed, c) a lock is formed, in which the non-real-time services share the bandwidth with each other after performing the bandwidth reconfiguration, and d) the non-real-time services are allowed to share the bandwidth in a time division manner.

Description

A LOCK-BASED CALL ADMISSION CONTROL METHOD IN WHICH NON-REAL-TIME SERVICES SHARE BANDWIDTH

Technical Field

[1] The present invention relates to a lock-based call admission control method in which non-real-time services for optimizing system performance and satisfying quality of service (QoS) for multi-services share bandwidth in a next generation mobile communication system.

Background Art

[2] A reconfiguration method is a representative method among conventional call admission control methods of a mobile communication system. The reconfiguration method is for previously allocating minimum and maximum bandwidths satisfying quality of service (QoS) in traffic classes. .

[3] By using the reconfiguration method, the bandwidth of various services may be varied from a maximum value to a minimum value. That is, an operation obtaining bandwidth from services having low priority or lending bandwidth to a service having high priority is performed according to resource availability at an arbitrary moment.

[4] For an example, a new service continues at a minimum bandwidth when the new service does not obtain a maximum bandwidth in an arbitrary base station cell. For another example, when the high priority service may not obtain the minimum bandwidth from the arbitrary base station cell, the higher priority service deprives bandwidth by reducing the bandwidth of the service having lower priority to a minimum quality level.

[5] Korean patent application No. 2002-0001238 (filed on January 9, 2002) discloses an invention entitled "Method for admitting call in mobile communication system."

[6] This patent is similar to the reconfiguration scheme, in which a method for admitting new and handover calls according to the QoS in a base station of the mobile communication system is disclosed.

[7] In this patent, in further detail, there are three steps for efficiently using limited bandwidth resources: a first step for determining if there are available channel resources upon request for establishing a call requiring a QoS guarantee; a second step for determining if a remaining bandwidth may support a data rate required to guarantee the QoS when there are available channel resources; and a third step for allocating the channel resources and base station switch and transmitting a call accept response signal to a base station controller when the data rate is supported.

[8] In addition, Korean patent application No. 2002-0078138 (filed on December 10, 2002) discloses an invention entitled "Method for assigning resources of base station controller in the mobile communication system."

[9] This patent discloses a base station controller resource allocation method in a mobile communication system, the base station controller resource allocation method for allocating selector resources of the base station controller by considering various call types having different selector resource requirement when the mobile station requests the call establishment in the mobile communication system.

[10] In further detail, a call accessing rate is improved by efficiently using the given selector resources since a call is allocated to a process having a minimum amount of available resources among processors in which the amount of available resources of the base station is greater than the resource requirement of the mobile station when the mobile station requests the call establishment in the mobile communication system. Accordingly, a call blocking rate is reduced.

[11] However, in conventional call admission control methods, bandwidth is problematically allocated to services having lower priority according to traffic characteristics of the services even though the services are not used. Disclosure of Invention Technical Problem

[12] The present invention has been made in an effort to provide a lock-based call admission control method in which non-real-time services share bandwidth, the lock- based call admission control method for minimizing bandwidth allocation to respective services and increasing a total system availability by allocating the bandwidth to a lock according to traffic characteristics of the non-real-time services and allowing the non- real-time services to share the bandwidth. Also, the present invention has been made in effort to provide a lock-based call admission control method in which non-real-time services share bandwidth, the lock-based call admission control method for guaranteeing service continuation by allocating minimum bandwidth which may not be deprived by the real-time services having high priority to the non-real-time services. Technical Solution

[13] In an exemplary lock-based call admission control method in a mobile communication system, a) respective calls according to multi-service classes of real-time or non-real-time services are received, b) bandwidth reconfiguration for borrowing bandwidth from low priority services or lending the bandwidth to high priority services according to resource availability is performed for the respective received calls, c) a lock is formed, in which the non-real-time services share the bandwidth with each other after performing the bandwidth reconfiguration, and d) the non-real-time services are allowed to share the bandwidth in a time division manner. At this time, the non- real-time services of the lock in c) receive the minimum shared bandwidth which may not be deprived by the real-time services having high priority. In c), the bandwidth is not allocated to the respective non-real-time services, but the shared bandwidth for minimizing bandwidth allocation to the non-real-time services is allocated. Advantageous Effects

[14] According to the exemplary embodiment of the present invention, by using the characteristics of the non-real-time services, total system resource availability is increased since the bandwidth allocated to the respective services is minimized by sharing the bandwidth in the time division method.

[15] In addition, according to the exemplary embodiment of the present invention, a service continuation is guaranteed since a minimum shared bandwidth is allocated to the non-real-time services sharing bandwidth in the lock method. That is, the non- real-time services receive the minimum shared bandwidth which may not be deprived by the real-time services having high priority. Brief Description of the Drawings

[16] FlG. 1 shows a diagram for representing a conventional traffic pattern of a non- real-time service.

[17] FlG. 2 shows a diagram for representing lock formation according to an exemplary embodiment of the present invention.

[18] FlG. 3 shows a flowchart for representing an operation of the lock-based call admission control method according to the exemplary embodiment of the present invention.

[19] FlG. 4 and FlG. 5 respectively show pseudo codes for describing a detailed operation of the lock-based call admission control method according to the exemplary embodiment of the present invention.

[20] FlG. 6 to FlG. 9 show flowcharts for representing detailed operations of the lock- based call admission control method according to the pseudo codes shown in FlG. 4 and FIG. 5.

[21] FlG. 10 shows a diagram for representing the maximum and minimum required bandwidth of four different service classes used for performing the simulation for performance evaluation according to first and second exemplary embodiments of the present invention.

[22] FlG. 11 to FlG. 14 respectively show blocking probabilities of new and handoff calls for each service class according to the first exemplary embodiment of the present invention.

[23] FlG. 15 to FlG. 18 respectively show blocking probabilities of new and handoff calls for each service class according to the second exemplary embodiment of the present invention.

Best Mode for Carrying Out the Invention

[24] A lock-based call admission control method in which non-real-time service according to an exemplary embodiment of the present invention shares bandwidth will be described with reference to figures.

[25] Firstly, it will be described that services belongs to one of four classes of services recommended in a universal mobile telecommunications system (UMTS) as follows. Class 1 is a conversational class for voice or video conference traffic, Class 2 is a streaming class for real-time video streaming, Class 3 is an interactive class for the world wide web (WWW) or data access, and Class 4 is a background class for email or downloading.

[26] Class 1 and Class 2 correspond to real-time services and Class 3 and Class 4 correspond to non-real-time services. FIG. 1 shows a diagram for representing a conventional traffic pattern of a non-real-time service. The call admission control method according to the exemplary embodiment of the present invention is motivated by the fact that the traffic pattern of some non-real-time services, including WWW surfing and database access traffic, has a large portion of reading time which is a time duration elapsed between the completion of a transfer of a previous request and the beginning of the transfer of a current request.

[27] As shown in FIG. 1, the reading time is inactive. That is, no bandwidth is allocated for the reading time. In general, the reading time is longer than an actual data downloading time.

[28] When resources in a cell are insufficient, it may be advantageous in terms of system utilization to group the services having the reading time into a set, and then to have them share bandwidth in a time division method such as a round robin bandwidth sharing method rather than to allocate a separate bandwidth to each service.

[29] This grouping method or the set grouped by the above method will be referred to as a lock. In addition, the time division bandwidth sharing method is for alternating the services in the lock so as to sequentially allocate bandwidth to the services for a predetermined maximum time. When the services of the lock are finished earlier than an allowed time or do not need the bandwidth at their turn, the services promptly pass the turn to a neighboring sequent service. Accordingly, the lock enables a set of the services to continue with less bandwidth than conventional allocation methods.

[30] One of the advantages of using the lock is that the system resource availability may be improved by minimizing bandwidth allocated to the respective services since a shared bandwidth is allocated to the services in the time division method by using characteristics of the non-real-time services. The other merit is that service con- tinuation is guaranteed since a minimum shared bandwidth is allocated to the non- real-time services sharing bandwidth in the lock method. That is, the non-real-time services receive the minimum shared bandwidth which may not be deprived by the real-time services having high priority.

[31] For example, when new call or handoff requests of the real-time services are received at a cell which has too little remaining bandwidth to admit the requests by the conventional reconfiguration method, a lock is formed to arrange bandwidth by dynamically grouping the non-real-time services in the lock method according to the exemplary embodiment of the present invention. At this time, multiple locks are formed in a cell, and the respective locks are coupled with the real-time services which triggered to form the locks. Accordingly, the locks having the services may be eliminated when the real-time services forming the locks are finished.

[32] In addition, the lock formation proceeds in three steps including bandwidth collection, setting condition satisfaction, and lock formation.

[33] At first, the amount of the bandwidth to be collected by the lock is calculated as in

Math Figure 1.

[34] MathFigure 1

BW collected - BW requested - BW available

[35] where BW denotes the bandwidth amount required by the ingress real-time requested service, and BW denotes the available bandwidth admitted in the cell. At this available time, the lock operation is performed when BW is greater than BW requested available

[36] The second step is to determine whether bandwidth is satisfied with setting conditions given as in Math Figures 2 and 3. [37] MathFigure 2

[38] MathFigure 3

BW unl ,ock .ed

max( \ BW i /, ^? l, ?L c [39] , where c denotes the number of the non-real-time services of which bandwidth is included in accumulation. That is, the second step is to determine whether the amount of the bandwidth added by the current bandwidth of the non-real-time services which will be locked but are not yet locked is greater than the amount of the collected bandwidth BW as shown in Math Figure 2, and to determine whether the collected bandwidth shared by lock members is greater than a maximum value of the bandwidth requested by the lock members as shown in Math Figure 3.

[40] Finally, the lock is formed by removing the bandwidth occupied by the locked services except one service BWunlocked shared by the lock members, and the bandwidth is reallocated by the request of the ingress real-time service.

[41] FIG. 2 shows a diagram for representing the lock formation according to the exemplary embodiment of the present invention.

[42] As shown in FIG. 2, the bandwidth to be allocated is 88Kbps when the class 2 service requiring 128Kbps enters a cell. Accordingly, the class 2 service obtains 128Kbps in total by obtaining 40Kbps from two class 3 and one class 4 services. At this time, the four non-real-time services in the lock share 16Kbps of a remaining class 4 in the lock in the time division method.

[43] When using the locks according to the exemplary embodiment of the present invention, a maximum wait time timew of one locked service and a maximum buffer size B which is necessary to store the packets until next bandwidth allocation may be calculated as in Math Figure 4 and Math Figure 5.

[44] MathFigure 4

L- — 1

[45] MathFigure 5

[46] where

time Θ denotes a maximum bandwidth occupation time of a locked service i, c denotes the number of locked services, and BW denotes shared bandwidth in the lock. At this shared time,

time Θ and BW are maximum upper bounds since the lock method according to the shared exemplary embodiment of the present invention allows locked services to skip their turns of bandwidth allocation in the time division manner when the locked services do not need the bandwidth. [47] When the locked service has a constant delay time , a maximum allowable delay delay

C^maxto meet the QoS is given as in Math Figure 6. [48] MathFigure 6

c

[49] ,where it is assumed that time of locked services in a lock are the same as each delay other. [50] Lock update according to the exemplary embodiment of the present invention will now be described. [51] Update on a lock occurs either when the locked services withdraw from the lock or when the non-real-time services join the lock. The lock update operation is performed by service termination or handover. [52] When the service withdrawing from the lock occupies the shared bandwidth, the service transfers the bandwidth to a sequent locked service before leaving the lock.

When the service withdrawing from the lock is the last service residing in the lock, the bandwidth is added to available bandwidth of the cell and released. However, the lock continues to exist until a lock initiator service ends. [53] The other type of lock update is a case in which the non-real-time service joins existing locks. That is, when the non-real-time service enters a cell but there is not enough bandwidth, the service is still able to be admitted to the cell by joining one of the locks which have extra capacity to share the bandwidth. [54] The non-real-time service may join existing locks when the following two conditions are met. For a first condition, the current number of locked services is required to be less than the number of locked services at the lock formation time. For a second condition, the bandwidth required by the joining service is required to be equal to or less than the bandwidth shared in the lock.

[55] When a lock satisfying these two conditions is found, the requesting service is allowed to enter the cell by joining the lock. In this case, there is no need to allocate bandwidth for the service.

[56] A reason to limit the number of services in a lock in the first condition is that bandwidth sharing by too many locked services may result in an unacceptable service delay and packet loss. Another reason to limit the number of services is that it helps to balance the number of locked services in the locks by preventing the services from concentrating onto several locks.

[57] Lock elimination according to the exemplary embodiment of the present invention will now be described.

[58] The lock according to the exemplary embodiment of the present invention may be eliminated when a lock initiator service ends at service time expiration or at handoff to another cell. When the lock is eliminated in the above case that the lock initiator service ends at service time expiration or at handoff to the other cell, simpler lock implementation and more consistent lock management may be performed with regard to formation and elimination.

[59] When the lock according to the exemplary embodiment of the present invention is removed by the expiration or the handoff of the lock initiator, the respective bandwidth of the locked services is restored to a level before the services are joined to the lock. The rest of the bandwidth is reinstated to the cell for future allocation when there is more bandwidth than needed for this restoration. At this time, referring to the above Math Figure 1, the bandwidth released by the lock elimination is required to be equal to or less than the sum of bandwidth required for the restoration.

[60] FIG. 3 shows a flowchart for representing an operation of the lock-based call admission control method according to the exemplary embodiment of the present invention.

[61] As shown in FIG. 3, in the lock-based call admission control method according to the exemplary embodiment of the present invention, calls according to real-time or non-real-time multi-service classes are received in step S310, and a reconfiguration operation for the received calls is performed in step S320. The reconfiguration operation is to borrow bandwidth from low priority services or to lend the bandwidth to high priority services according to the resource availability at the moment.

[62] The lock operation in which the non-real-time services share bandwidth is performed in step S330, and the non-real-time services share the bandwidth in the time division manner in step S340. That is, in addition to performing the reconfiguration in step S320, the lock operation in which the non-real-time services share the bandwidth by using traffic characteristics of the non-real-time service is performed, the bandwidth shared by the non-real-time services is allocated, and the non-real-time services share the bandwidth in the time division manner. The reading time described with reference to FlG. 1, one of the traffic characteristics of the non-real-time service, is a time duration elapsed between the completion of a transfer of a previous request and the beginning of the transfer of a current request. In addition, no bandwidth is allocated for the reading time since the reading time is inactive.

[63] At this time, the shared bandwidth is allocated instead of allocating the bandwidth to the respective non-real-time services, and therefore the bandwidth allocated to the respective non-real-time services is minimized.

[64] Accordingly, service continuation is guaranteed by allocating the minimum bandwidth to the non-real-time services sharing the bandwidth in the lock method. That is, the non-real-time services receive the minimum bandwidth which may not be given to real-time services having high priority.

[65] Then, it is determined whether the call continues in step S350, and steps 320 to 340 are repeatedly performed while the call continues.

[66] FlG. 4 and FlG. 5 respectively show pseudo codes for describing detailed operations of the lock-based call admission control method according to the exemplary embodiment of the present invention. FlG. 6 to FlG. 9 show flowcharts for representing detailed operations of the lock-based call admission control method according to the pseudo codes shown in FlG. 4 and FlG. 5.

[67] The pseudo code shown in FlG. 4 and FlG. 5 provides detailed descriptions of the lock-based call admission control method according to the exemplary embodiment of the present invention, and the pseudo code is implemented at each base station. In addition, the pseudo code describes the bandwidth allocation corresponding to each service in every case.

[68] Notations used in the pseudo code shown in FlG. 4 and FlG. 5 will be described.

[69]

BW r^meq^ax and

BW mm req respectively denote maximum and minimum bandwidths required by incoming service, and BW avail denotes available bandwidth of a current cell. BW locking denotes deprived bandwidth by lock formation, and

_O τττreconfig ^{n VV} avail

denotes available bandwidth of a cell after the reconfiguration of class 3 and 4 services which are non-real-time services. [70] A detailed description of the lock-based call admission control method based on the pseudo code shown in FlG. 4 FlG. 5 will be described with reference to FlG. 6 to FlG.

9. [71] Firstly, referring to FlG. 6 and FlG. 7, the lock-based call admission control method according to the exemplary embodiment of the present invention for a new call connection will be described. [72] When a new call arrives in step S501, it is determined whether the arrived new call is the real-time service of class 1 and class 2 in step S503. [73] When the new call is the real-time service of class 1 and class 2, it is determined whether the maximum request bandwidth

BW max r.eq

is less than the bandwidth BW available in the current cell in step S505. When the avail maximum request bandwidth

BW max r.eq

is less than the bandwidth BW , the service is admitted by allocating avail

BW_n m_qax

in step S508. That is, the maximum request bandwidth BW max r.eq

of the real-time service is admitted in the corresponding cell. [74] When the maximum request bandwidth

BW max r.eq

is not less than the bandwidth BW , it is determined whether the minimum request avail bandwidth

BW mill r.eq

is less than the bandwidth BW available in the current cell in step S506. When the avail minimum request bandwidth

BW_r m_eqill

is less than the bandwidth BW , the service is admitted by allocating the minimum avail request bandwidth

BW r^meq^m

in step S509. That is, the service enters a cell at a degraded level by requesting the minimum request bandwidth

BW- [75] Accordingly, when there is sufficient bandwidth to perform the above steps, the service is accepted by allocating a corresponding bandwidth. [76] When there is insufficient bandwidth available, it is determined whether enough bandwidth to accept the services may be gathered by forming a lock. That is, when the minimum request bandwidth

BW-

is less than a sum of the bandwidth BW available in the current cell and the avail bandwidth deprived by the lock formation in step S507. At this time, when the minimum request bandwidth

BW-

is less than a sum of the bandwidth BW available in the current cell and the avail bandwidth deprived by the lock formation, the bandwidth is deprived by locking the non-real-time services currently supported in the cell in step S510, and the service is accepted by allocating the minimum request bandwidth

BW-

in step S511. That is, when the steps S507 and S510 are performed appropriately, the services are accepted with the minimum request bandwidth

. Otherwise, the services are rejected in step S512. [77] In addition, it is determined whether the new call has arrived in step S501 is the non-real-time service of class 3 and class 4 in step S504. [78] When the new call is the non-real-time service of class 3 and class 4, it is determined whether the maximum request bandwidth

is less than the bandwidth BW available in the current cell in step S513. When the avail maximum request bandwidth

D T^ max req is less than the bandwidth BW , the service is accepted by allocating the maximum avail request bandwidth

in step S515. That is, when the requested bandwidth of the service is accepted in the cell, the maximum request bandwidth

of the non-real-time service is accepted in the corresponding cell. [79] When the maximum request bandwidth

D T^ max req is not less than the bandwidth BW , the services attempt to obtain the minimum avail request bandwidth BW r^meq^m

. That is, it is determined whether there are existing locks to join in step S514. At this time, the services attempt to join the existing locks in the cell by searching for locks with capacity. Accordingly, when the lock is found, the services are accepted in step S516 without bandwidth allocation. Otherwise, the services are rejected in step S517.

[80] Referring to FlG. 8 and FlG. 9, the lock-based call admission control method for a handover case according to the exemplary embodiment of the present invention will be described.

[81] Firstly, it is determined whether a new call is handover call in step S502, and it is determined whether the handover call is the real-time service of class 1 and class 2 in step S518.

[82] When the handover call is the real-time service of class 1 and class 2, it is determined whether the maximum request bandwidth

is less than the bandwidth BW available in the current cell in step S520. When the avail maximum request bandwidth

BW_r m_eqax

is less than the bandwidth BW avail , the service is accepted in step S525 by allocating the maximum request bandwidth

BW_r m_eqax

That is, the maximum request bandwidth

of the real-time service is accepted in the corresponding cell. [83] When the maximum request bandwidth BW. max req is not less than the bandwidth BW , it is determined whether the minimum request avail bandwidth

BW mm r.eq is less than the bandwidth BW avail available in the current cell in step S521. When the minimum request bandwidth

BW r^meq^m

is less than the bandwidth BW avail , the service is accepted by allocating the minimum request bandwidth

BW-

. That is, the service enters a cell at a degraded level by requesting the minimum request bandwidth

BW r^meq^m

[84] Accordingly, when there is sufficient bandwidth to perform the above steps, the service is accepted by allocating corresponding bandwidth. [85] When there is still insufficient bandwidth after performing steps 520 and 521, the minimum request bandwidth

BW-

allocation is attempted after performing the reconfiguration for degrading the bandwidth of the non-real-time services to the minimum quality level. [86] In further detail, it is determined whether the minimum request bandwidth BW. mm req is less than the bandwidth

r> _JJ7 reconfig ^{D VV} avail available in the cell after performing the reconfiguration of class 3 and class 4 of the non-real-time service in step S523. When the minimum request bandwidth

BW-

is less than the bandwidth

Ji _TXl reconfig ^{n VV} avail

, bandwidth from current class 3 and class 4 services is reconfigured in step S527. When there is enough bandwidth collected as a result of the reconfiguration, the services are admitted in step S528 by allocating the minimum request bandwidth

BW_r m_eqill

[87] When the minimum request bandwidth

BW-

is not less than the bandwidth

τ> ττi reconfig ^{D VV} avail

, it is determined whether bandwidth large enough to accept the services is collected by forming a lock. That is, when the bandwidth large enough to accept the services is collected by forming the lock, the services are accepted with the minimum request bandwidth

BW-

. Otherwise, the services are rejected. [88] In further detail, when the minimum request bandwidth

BW min r.eq is not less than the bandwidth

BW reconfig a 2vvaaiiϊl

, it is determined whether the minimum request bandwidth

BW_r m_eqin

is less than a sum of the bandwidth

τ> JJ7 reconfig ^{D VV} avail available in the cell after performing the reconfiguration of class 3 and class 4 of the non-real-time service and the bandwidth BW locking deprived by the lock formation in step S524. At this time, when

BW-

is less than a sum of

_T> _ITT reconfig ^{D VV} avail and BWlocking, the bandwidth is collected by locking the current non-real-time services in step S529, and the service is admitted in step S530 by allocating the minimum request bandwidth BW r^meq^m

. That is, when steps S524 and 529 are appropriately performed, the services are accepted with the minimum request bandwidth

BW-

. Otherwise, the services are rejected in step S531.

[89] In addition, it is determined whether the handover call in step S502 is the non- real-time service of class 3 and class 4 in step S519.

[90] When the handover call is the non-real-time service of class 3 and class 4, it is determined whether the maximum request bandwidth

BW r^™eq^*

is less than the bandwidth BW available in the current cell in step S532. When the avail maximum request bandwidth

is less than the bandwidth BW avail , the services are admitted in step S535 by allocating the maximum request bandwidth

BW r^meq^aκ

. That is, when the requested bandwidth of the service is accepted in the cell, the maximum request bandwidth

BW-

of the non-real-time service is accepted in the corresponding cell. [91] When BW_r m_eqax

is not less than BW avail , the services attempt to obtain the minimum request bandwidth

BW_r m_eqin

That is, when

BW_r m_eqax

is not less than BW , it is determined whether the minimum request bandwidth avail

BW_r m_eqm

is less than the BW available in the current cell in step S533. When avail max

is less than BW , the services are accepted in step S536 by allocating the minimum avail request bandwidth

BW mm r.eq

[92] When respective

BW max r.eq

and

BW r^meq^m are not less than BW avail in steps S532 and S533, it is determined whether there are existing locks to join in step S536. At this time, the services attempt to join the existing locks in the cell by searching for locks with capacity. Accordingly, when the lock is found, the services are accepted in step S537 without bandwidth allocation. Otherwise, the services are rejected in step S538.

[93] In FlG. 10 to FlG. 18, performance evaluation between the lock-based call admission control method according to the exemplary embodiment of the present invention and the conventional call admission control method based on the reconfiguration method is performed. In the exemplary embodiment of the present invention, the round robin method is used when the bandwidth is shared in the time division method.

[94] Firstly, 19 hexagon-shaped cells are formed for a simulation environment for performance evaluation. Each cell has IKm radius and may support up to 2Mbps bandwidth. A path-loss model and a shadowing model are used for the channel environment. At this time, each terminal performs 4 services having different service classes at the same time.

[95] FIG. 10 shows a diagram for representing the maximum and minimum bandwidth of four different service classes used for performing the simulation for performance evaluation according to first and second exemplary embodiments of the present invention.

[96] FIG. 11 to FlG. 14 respectively show blocking probabilities of new and handoff calls for each service class according to the first exemplary embodiment of the present invention. FIG. 15 to FlG. 18 respectively show blocking probabilities of new and handoff calls for each service class according to the second exemplary embodiment of the present invention.

[97] Simulations according to the first and second exemplary embodiments of the present invention are different from each other in terms of the ratio of the numbers of simulated services of each class. That is, in the first exemplary embodiment of the present invention, the ratio among class 1, 2, 3, and 4 is 1:1:1:1, and in the second exemplary embodiment of the present invention, the ratio is 4:1:2.5:2.5. In addition, an arrival rate λ of each terminal follows exponential distribution, and locations (angle θ, and speed δ) of respective terminals are uniformly distributed following the exponential distribution with a mean of 0.005 (sec).

[98] As shown in FlG. 11 to FIG. 14, the performance of the lock-based call admission control method according to the exemplary embodiment of the present invention and the performance of the conventional configuration method are evaluated by comparing their blocking probabilities with each other. The lock-based call admission control method according to the exemplary embodiment of the present invention includes the lock function and functions of the conventional reconfiguration method as well. [99] As shown in FlG. 10, for all classes 1, 2, 3, and 4, the lock-based call admission control method according to the exemplary embodiment of the present invention decreases the blocking probabilities of the non-real-time services, compared to the conventional reconfiguration method, as the call arrival rate λ is increased. While the blocking probabilities of the non-real-time services are increased as the real-time services are increased according to the conventional reconfiguration method, the blocking probabilities of the non-real-time services are reduced according to the exemplary embodiment of the present invention because the non-real-time services are admitted since the new non-real-time services join the lock as the real-time service is increased.

[100] As shown in FlG. 15 to FlG. 18, the performance of the lock-based call admission control method according to the exemplary embodiment of the present invention and the conventional configuration method is evaluated by comparing their blocking probabilities with each other.

[101] For example, in a case of class 2, the first exemplary embodiment of the present invention shown in FlG. 11 to FlG. 14 has increased blocking probabilities compared to the second exemplary embodiment of the present invention shown in FlG. 15 to FlG. 18. That is, the first exemplary embodiment of the present invention has the blocking probabilities increased by 15% compared to the second exemplary embodiment of the present invention because the class 2 service is a service requesting maximum bandwidth, the number of users of class 2 service is a quarter of the total of users since the ratio of classes 1, 2, 3, and 4 is 1:1:1:1 according to the first exemplary embodiment of the present invention shown in FlG. 11 to FlG. 14, and the number of the users of class 2 is one tenth of the total of the users since the ratio of classes 1, 2, 3, and 4 is 4: 1 :2.5:2.5 according to the second exemplary embodiment of the present invention shown in FlG. 15 to FlG. 18.

[102] Accordingly, the bandwidth of other service classes is increased because the reduction of the users of class 2 requesting high bandwidth decreases the bandwidth of class 2, and therefore the blocking probabilities of the other service classes are reduced. That is, the new and handover call blocking probabilities according to the exemplary embodiment of the present invention are reduced to less than the same of the conventional reconfiguration method.

[103] While this invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

[104] Throughout this specification and the claims which follow, unless explicitly described to the contrary, the word "comprise" or variations such as "comprises" or "comprising" will be understood to imply the inclusion of stated elements but not the exclusion of any other elements.

[105]

[106]

Claims

[1] A lock-based call admission control method in a mobile communication system, comprising: a) receiving respective calls according to multi-service classes of real-time or non-real-time services; b) for the respective received calls, performing bandwidth reconfiguration for borrowing bandwidth from low priority services or lending bandwidth to high priority services according to resource availability; c) forming a lock in which the non-real-time services share the bandwidth with each other after performing the bandwidth reconfiguration; and d) allowing the non-real-time services to share the bandwidth in a time division manner.

[2] The lock-based call admission control method of claim 1, wherein b) to d) are repeatedly performed when the call continues.

[3] The lock-based call admission control method of claim 1, wherein, the non- real-time services of the lock in c) receive the minimum shared bandwidth which may not be deprived by the real-time services having high priority.

[4] The lock-based call admission control method of claim 1, wherein, in c), the bandwidth is not allocated to the respective non-real-time services, but the shared bandwidth for minimizing bandwidth allocation to the non-real-time services is allocated.

[5] The lock-based call admission control method of claim 1, wherein a traffic pattern of the non-real-time service is one of a world wide web (WWW) surfing traffic pattern and a database access traffic pattern.

[6] The lock-based call admission control method of claim 1, wherein a traffic pattern of the non-real-time service has a reading time longer than a data downloading time.

[7] The lock-based call admission control method of claim 1, wherein the lock means a method for grouping services having a reading time into a set when there are insufficient resources in a cell.

[8] The lock-based call admission control method of claim 1, wherein c) is performed when new call or handoff requests of the real-time services are received at a cell which has too little remaining bandwidth to admit the requests by the conventional reconfiguration method.

[9] The lock-based call admission control method of claim 1, wherein, when the services of the lock in c) are finished earlier than an allowed time or do not need the bandwidth at their turn, the services pass the bandwidth to a subsequent neighboring service. [10] The lock-based call admission control method of claim 1, wherein the lock having the services is deleted when the real-time services forming the lock in c) are finished. [11] The lock-based call admission control method of claim 1, wherein the lock is formed in c) when a request bandwidth is greater than a bandwidth available in a current cell. [12] The lock-based call admission control method of claim 1, wherein the lock is formed in c) when the amount of the bandwidth added by the current bandwidth of the non-real-time services which will be locked but is not yet locked is greater than the amount of collected bandwidth, and bandwidth shared by lock members is greater than a maximum value of the bandwidth requested by the lock members. [13] The lock-based call admission control method of claim 1, wherein the lock formed in c) is updated when the services in the lock withdraw from the lock or the non-real-time services join the lock. [14] The lock-based call admission control method of claim 1, wherein the shared bandwidth is given to services in a subsequent lock when the services withdrawn from the lock occupy the shared bandwidth when the services are withdraw from the lock. [15] The lock-based call admission control method of claim 1, wherein the lock in c) is updated when the non-real-time services satisfying a predetermined condition join the existing lock. [16] The lock-based call admission control method of claim 1, wherein the service is admitted in the cell by joining one of the locks having extra capacity to share the bandwidth when the non-real-time service in c) enters a cell but there is not enough bandwidth. [17] The lock-based call admission control method of claim 1, wherein the lock in c) is deleted when a lock initiator service ends at service time expiration or is handed over to another cell.