Realization of GA-NN Blind Equalization Algorithm
Hu Yong-sheng, Yang Ling-ling, Jianlei Li · 2009
In order to overcome the shortcomings of standard BP algorithm, such as low convergence speed and local optimum, since GA is a random searching method with global characteristics, the paper adopts GA to NN training. By analyzing results of an example, the essay draws conclusion that the equalization effect of GA-NN has been greatly improved, convergence process has also been quickened, besides that , residual error as well as BER has been reduced greatly.