Research On The Call Admission Control Strategy For Systems With Multi-Rate VBR Services

B-ISDN is actually a multi-rate operation system, which aims to provide users with multifarious services, such as voice, video, e-mail transfers, as well as computer files and other forms of information, making the amount of communication networks information rapidly growing. On the other hand, because of the limited network resource, compression coding technology is widely used, inducing the majority of on-line transmission of real-time information flow is VBR (variable bit rate) service. Various VBR service have different QoS requirements. When a number of statistical multiplexing media streams goes into the system, it is important to determine whether the system can ensure their respective requirements of QoS, or under their QoS constraints conditions, the maximum number of media steams will be permitted into the channel. Therefore, CAC (call admission control) is essential. Call admission control system based on multi-rate VBR service has become an important research subject.This article first introduced some basic concepts of multi-rate VBR system and CAC, and analyzed the typical resource management and call admission control strategy. we pointed out the method of calculating the call loss probability of completed sharing, part sharing, and mixed sharing, as well as the advantages and disadvantages based on various strategies. as an important statistical properties, the control parameter in peak method can only be integer, which would cover the average rate of various service information when used in multi-rate VBR service system, It is difficult to allocate means and results to network bandwidth for a better performance. To fix this problem, we simplified the real-time VBR stream with ON-OFF model, and put forward a call admission control strategy that based on the average bandwidth needs of on-line various VBR service and expectations of the packet loss rate. The strategy just needs a dynamic adjustment of the control parameters - capacity scaling factor, and other factors to decide whether to accept a call, such as the product of scaling factor, capacity and physical capacity, and all types of service connections online and the type of call request . then following the method of calculating the probability of call loss and packet loss rate, as well as the solution of capacity scaling factor, it also compare with the performance of peak method at the same time.The results showed that in CAC strategy with average bandwidth measure, the computational efforts required to determine the control parameters was less than peak method, leading to a better overall performance, because of the fact that the average information rate of VBR source could be any real number, which allowing a higher rate of packet loss, the probability of call accepted for big bandwidth demanded service increased, so did the throughput of the system. the strategy achieved the goal of improving, with values in researching and applying.