离散粒子群优化在垂直分层空时系统检测中的应用

为了改善垂直分层空时系统串行干扰抵消算法的性能,提出了一种离散粒子群检测算法(DPSO-DA).该算法将垂直分层空时系统中的最优检测视为组合优化问题,根据最大似然检测公式构造DPSO算法的适应度函数,利用DPSO算法来解决该组合优化问题,从而寻找最优解.针对DPSO-DA 有可能出现早熟现象,进一步提出了一种混合离散粒子群检测算法 (HDPSO-DA).HDPSO-DA 对 DPSO-DA 的进化方程进行了重新设计,在搜索中以一定变异概率对选中的粒子进行变异,进一步改善了DPSO-DA的性能.理论分析和仿真结果表明,当误码率为10-3时,与基于最小均方误差准则的串行干扰抵消算法相比,DPSO-DA 和 HDPSO-DA 可获得约3dB 和 5dB的增益,且具有更低的复杂度.

Discrete Particle Swarm Optimization for Vertical Bell-Labs Layered Space-Time System Detection

A discrete particle swarm detection algorithm(DPSO-DA) is proposed for the signal detection of the vertical bell-labs layered space-time(V-BLAST) system to improve the performance of the ordered successive interference cancellation(OSIC) algorithm.The optimal detection in V-BLAST system can be formulated as a combinational optimization problem and then the fitness function of the DPSO algorithm is constructed in terms of the maximum likelihood detection formula.The proposed DPSO algorithm is then used to solve the combinational optimization problem to get an optimal solution. Attempts have been made at avoiding the premature convergence of the DPSO-DA, and a hybrid discrete particle swarm detection algorithm(HDPSO-DA) is proposed.The HDPSO-DA is obtained from DPSO-DA by redesigning the evolution equation of the DPSO-DA and introducing the mutation operator of the genetic algorithm. The performance of the HDPSO-DA is improved.Theoretical analysis and simulation results show that the both the DPSO-DA and the HDPSO-DA acquire about 3dB and 5 dB gain over the OSIC algorithm based on the minimum mean square error criterion at a bit error rate of 10-3 and have lower computational complexity.