Next generation cellular architectural hierarchy imposes that a pre-fixed group of Node Bs be connected to a Radio Network Controller (RNC) and a pre-designed set of such RNCs be then assigned to a Serving GPRS support Node (SGSN). In addition to that, the set of RNCs is also wired to one Mobile Switching Center (MSC). In case of a single-homed network, the operating cost, comprises the cabling or link cost and the cost incurred for supporting handoff whenever an mobile user (MU) visits a different RNC (horizontal/simple handoff) or a different SGSN or MSC (vertical handoff/complex handoff). In contrast, during the deployment phase, provision could be kept where a Node B could be connected to any of the RNCs and thereafter any RNC could be assigned to any of the SGSNs/MSCs, if needed. This concept of multi-homing allows nodes to be assigned adaptively to the switches, particularly, under constrained situation. In this work, we present a framework for multi-homing of cells to switches and endeavor to present some evolutionary optimization techniques for minimizing the total cost of resource management. We have proposed a novel meta-heuristic based approach, that evolves into a complete state-space based search tool namely, Best Contributors Search (BCS). Computational results indicate that the proposed BCS algorithm out-performs some other well-known heuristics such as matrix-based technique, Optimal Dual Home (ODH), Tabu Search (TS) and Ant Colony optimization (ACO) based methods cited so far.