A Knowledge-Based Approach to Capacity Allocation for Multiservice Packet Networks

Artificial intelligence and expert system technologies offer considerable promise for addressing many of the problems associated with the performance and management of multiservice communication networks. Such networks provide integrated digital services by packetizing all traffic and deploying a packet-switched network. Each service demand is characterized by its own resource requirements, protocol procedures, and performance objectives. This paper deals with the application of expert systems and intelligent control to the transmission link capacity allocation problem. Optimal capacity allocation is the process of assigning link capacities for each service in a communication network to maintain the projected performance objectives.