A Dual-based Method for Resource Allocation in OFDMA-SDMA Systems with Minimum Rate Constraints

TL;DR

This paper introduces a dual-based resource allocation method for OFDMA-SDMA systems with minimum rate constraints, providing a benchmark and a primal-feasible solution that outperforms standard heuristics in accuracy.

Contribution

The paper develops a dual problem formulation and a primal-feasible dual-based algorithm for resource allocation, offering a new benchmark and improved performance over existing methods.

Findings

The dual-based method closely follows the dual optimal bound.

It outperforms the weight adjustment method in accuracy.

The approach provides a useful benchmark for heuristic algorithms.

Abstract

We consider multi-antenna base stations using orthogonal frequency-division multiple access (OFDMA) and space division multiple access (SDMA) techniques to serve single antenna users, where some of those users have minimum rate requirements and must be served in the current time slot (real time users), while others do not have strict timing constraints (non real time users) and are served on a best effort basis. The resource allocation problem is to find the user assignment to subcarriers and the transmit beamforming vectors that maximize a linear utility function of the user rates subject to power and minimum rate constraints. The exact optimal solution to this problem can not be reasonably obtained for practical parameters values of the communication system. We thus derive a dual problem formulation whose optimal solution provides an upper bound to all feasible solutions and can be used to benchmark the performance of any heuristic method used to solve this problem. We also derive from this dual optimal solution a primal-feasible dual-based method to solve the problem and we compare its performance and computation time against a standard weight adjustment method. We find that our method follows the dual optimal bound more closely than the weight adjustment method. This off-line algorithm can serve as the basis to develop more efficient heuristic methods.