Performance Analysis of Resource Allocation Algorithm Based on Simple Particle Swarm Optimization for Device to Device Communication System

Abstract

The development of Device-to-Device (D2D) communication as an alternative communication scheme continues to grow. D2D communication enables each device to communicate directly with other devices without going through the Evolved Node B (eNB). These days, D2D communication can be implemented in conventional cellular communication, and use the same spectrum as the conventional cellular user (underlay). This underlay scheme can improve the cell’s spectrum efficiency, but the interference level that happens in the cell is increased. Because D2D communication and cellular communication use the same spectrum, there will be interference between D2D user equipment (DUE) and Cellular user Equipment (CUE). A well-designed radio resource allocation is needed to reduce the interference level, while maintaining the overall performance of the cell. In this research, Simplified Particle Swarm Optimization (SPSO) is proposed to overcome this problem. SPSO is a PSO-based algorithm with a limited number of iterations designed to halt calculations when the PSO algorithm cannot find a solution. If SPSO exceeds the limit iteration, a greedy algorithm is executed to do the allocation process. From the simulation, the SPSO algorithm can achieve 1.3310 × 10⁸ bps, 12.3239 bps/Hz, 2.1328 × 10³ bps/Watt and 92% on total sumrate, spectral efficiency, power efficiency, and system fairness respectively. These number is better if compared with the conventional greedy allocation algorithm. The total sumrate, spectral efficiency, and power efficiency are increased by 0.9%, 0.74%, and 0.95% in average datarate, spectral efficiency, and power efficiency respectively. Meanwhile the SPSO’s system fairness is decreased by 1.65% compared with the conventional greedy algorithm.