Algorithms for finding optimal offered load in wireless communication networks

In this paper we study the problem of finding the input load that optimizes the total throughput in circuit switched, multihop, wireless networks that are subject to QoS constraints on blocking probability. By doing so we are able to “size” the network capability for a given admission-control policy, thereby determining how much traffic the network can support as well as the offered loads that achieve this optimum. A “projection algorithm”, based on the use of Lagrangian optimization techniques and heuristics, is developed, and its behavior is investigated. Extensive computational results demonstrate that this algorithm provides reliable convergence to optimal solutions