Genetic algorithms applied to cellular call admission: local policies

It is well known that if a stochastic service system (such as a cellular network) is shared by users with different characteristics (such as differing handoff rates or call holding times), the overall system performance can be improved by denial of service requests even when the success capacity exists. Such selective denial of service based on the system state is defined as the call admission. A previous paper suggested the use of genetic algorithms (GAs) to find near-optimal call admission policies for cellular networks. In this paper, we define local call admission policies that make admission decisions based on partial state information. We search for the best local call admission policies for one-dimensional (1-D) cellular networks using genetic algorithms and show that the performance of the best local policies is comparable to optima for small systems. We test our algorithm on larger systems and show that the local policies found outperform the maximum packing and best handoff reservation policies for the systems we have considered. We find that the local policies suggested by the genetic algorithm search in these cases are double threshold policies. We then find the best double threshold policies by exhaustive search for both 1-D and Manhattan model cellular networks and show that they almost always outperform the best trunk reservation policies for these systems.