期刊名称：International Journal of Security and Its Applications
印刷版ISSN：1738-9976
出版年度：2016
卷号：10
期号：2
页码：95-104
DOI：10.14257/ijsia.2016.10.2.09
出版社：SERSC
摘要：The project of ant colony algorithm optimization neural network combining blind equalization algorithm is proposed. The better initial weights of neural networks are provided because of the randomness, ergodicity and positive feedback of the ant colony algorithm. And then, a combination of optimal weights are found through BP algorithm, which is fast local search speed. Thus blind equalization performance is improved. Computer simulation show that, the novel blind equalization algorithm speeds up the convergence rate, reduces the remaining steady-state error and bit error rate, which is compared with the Neural Network Blind Equalization Algorithm(NNBE) and Genetic Algorithm optimization Neural Network Blind Equalization Algorithm(GA-NNBE) .
