A Cost-Based Approach for Base Station Assignment in Mobile Networks with Limited Backhaul Capacity

Existing base station (BS) assignment methods in cellular networks are mainly driven by radio conditions since it is assumed that the limiting factor is the air interface. This assumption has been proven to be very reasonable when circuit voice was the dominant service and backhaul capacity provisioning in BS accounting for the peak rate of the radio link was an economically feasible option. However, as traffic continues increasing and enhanced air interfaces are been deployed, a growing concern is that also the mobile backhauling can become a bottleneck in certain deployment scenarios. In this paper we present a BS assignment algorithm designed to optimize radio resource usage while simultaneously considering potential capacity limitations in mobile backhauls. The BS assignment problem is formulated using a cost-based framework and mapped into a Multiple-Choice Multidimensional Knapsack Problem (MMKP) that is solved by means of a heuristic algorithm with polynomial time. The proposed algorithm is evaluated and compared to two schemes based exclusively on radio conditions. Simulation results show that the proposed algorithm can assign more users than existing algorithms without violating radio and transport constraints in partially backhaul limited scenarios.