A performance bound on dynamic channel allocation in cellular systems: equal load

We model a cellular network as a more general multiple service, multiple resource system. We define the “state” of the system as the number of calls currently carried in each cell. We restrict ourselves to channel allocation policies that place restrictions on the global state of the system, are allowed immediate global channel reallocation, and ignore handoffs. Maximum packing and fixed allocation are considered as special cases of such policies. Under uniform load conditions, we prove that throughput is increasing and concave with respect to increases in load or capacity, under maximum packing or fixed allocation. We propose that the optimal policy, in the considered class, varies from maximum packing at low loads to fixed allocation at high loads. This policy is often impractical to implement, but can be considered as a performance bound on practical systems. The analytical results are investigated numerically using a simple seven cell linear network