A Call Admission Control Protocol for High Speed Networks: Modeling and Analysis

C. Madubata and M. Arozullah (USA)


ATM Networks, Access Networks, Network Management, Computer Networks


This paper presents the development, analytical modeling, and analysis and performance evaluation of a Call admission Control (CAC) protocol for high speed Broadband Integrated Services Digital Networks. The protocol is suitable on-line applications and provides efficient utilization of network resources (link capacity and buffer space). The protocol also simultaneously satisfies two Quality of Service parameters, namely, Cell Loss Ratio (CLR) and delay. The protocol accommodates input sources that are either Poisson or ON-OFF with with MMPP characteristics. The desired delay and the CLR are prescribed for end-to-end operation of a connection. The end-to-end delay and CLR values are divided among the nodes of the connection. For each node of the network analytical expressions for the CLR and Delay are developed using M/D/1/K and MMPP/D/1/K queuing models. These analytical expressions are then used to obtain the maximum acceptable total input data rates for given buffer size, B cells and output link capacity C bps that satisfies the required CLR and Delay at the node. This is done by using an optimal numerical optimization method. The numerical optimization method is optimal in the sense that it allows admission of the maximum possible input data rate satisfying the desired CLR and delay for given buffer size and link capacity. Finally the CAC protocol is developed using the above results. The MATLAB codes for the protocol are available on request. Performance of this CAC protocol has been compared with a number of other CAC protocols. This protocol has better performance in a number of aspects. Call Admission Control is an important function of high-speed networks for avoiding congestion and for efficient utilization of network resources. Thus the protocol should be useful for design and operation of such networks.

Important Links:

Go Back