基于遗传算法的直接盲信号检测

Direct blind equalization using genetic algorithms

针对现有的盲均衡遗传算法收敛速度慢,难以准确地收敛到最优解的问题,借助接收数据阵补投影算子对待测发送序列向量的零化作用,把信号盲检测问题转化为整数约束下的二次优化问题,提出了一种基于遗传算法的直接盲信号检测的方法.所用遗传算法的特点是,适应度函数是直接针对被检测序列构造的,编码空间与遗传空间一致,搜索空间是离散、有限的.仿真结果表明,该算法收敛速度快、性能好.

Focusing on the problems existed in some blind equalization genetic algorithms that good astringency and converge to optimal solution are hard to achieve, this paper takes the complementary projection operator of a received date matrix as annihilator of a transmitted sequence vector to be detected, and then formulates the blind detection problem into a quadratic optimization with integral constraints. A novel direct blind equalization based on genetic algorithms is proposed. The character of this algorithm is that the fitness function is formatted with the transmitted sequences, the encoding space is identical to the genetic space, and the searching space is discrete and finite. Simulations show that the algorithm achieves good astringency and satisfactory performance.