In this paper using the dynamic pricing platform, a novel framework for downlink resource allocation in heterogeneous cellular CDMA networks is proposed. For each user we define a utility that is a function of channel status and delay condition of that individual user as well as network load status. The network utility is then defined as the summation of the users' achieved utilities. We solve downlink resource allocation problem through maximization of total network utility. This approach results in a sub-optimal base-station assignment scheme which-unlike previous work-is network optimal instead of cell optimal. We then show that optimal base-station assignment is a Multi-dimensional Multiple- choice Knapsack Problem (MMKP). Since MMKP is NP-Hard a polynomial-time suboptimal heuristic algorithm is then employed to develop an efficient base-station assignment.