Evolutionary algorithms for multiobjective optimization of frequency reuse schemes in CoMP-based MIMO-OFDMA networks

Abstract

Interference cancellation techniques based on soft frequency reuse (SFR) and coordinated multipoint transmission (CoMP) are deemed to play a fundamental role in the next generation of cellular deployments (B4G/5G systems). Both have shown great potential to satisfy the ambitious requirements of modern wireless cellular systems in terms of area spectral efficiency whereby users located anywhere in the cell, even at the edge, should be able to obtain reasonably large throughput. The proper selection of the SFR configuration parameters is a crucial task to reach a satisfactory network performance. This paper proposes a framework with a per-cell SFR configuration instead of the conventional global SFR configuration. An optimization tool able to find these configurations is needed since the exhaustive search, in this context, is computationally prohibitive. To this end, an evolutionary (i.e. genetic) algorithm is suitably adapted to the multiobjective problem of trying to simultaneusly optimize overall and cell-edge user throughputs. Numerical results demonstrate that the proposed framework outperforms the conventional setups both in terms of total capacity and fairness while remaining computationally feasible.