改进ABC-BP模型在数字调制识别中的应用  

Modified Artificial Bee Colony Algorithm Optimizing BP Neural Network and Its Application in the Digital Modulation Recognition

作　　者：史先铭 刘以安[1] 

机构地区：[1]江南大学物联网工程学院,江苏无锡214122

出　　处：《江南大学学报（自然科学版）》2015年第4期396-402,共7页Joural of Jiangnan University (Natural Science Edition)　

基　　金：国家自然科学基金项目(61170120)

摘　　要：为了提高数字调制信号在不同信噪比下的识别性能,将改进人工蜂群算法优化BP神经网络(MABC-BP)的模型应用于数字调制识别中;为进一步提高基本人工蜂群(Artificial Bee Colony,ABC)算法的寻优性能,对基本ABC算法中"跟随蜂"在"食物源"邻域的搜索行为进行改进。从仿真结果可以看出,当信噪比低至0 d B时,7种数字调制信号的调制识别率依然可以达到85%以上,从而证明了该方法能有效地提高数字调制信号的识别性能。Rapid development of communication technology contributes to the development of digital modulation recognition technology. In order to improve recognition performance of the digital modulation signals under different signal-to-noise ratios( SNR),the model of the modified artificial bee colony( MABC) algorithm optimizing BP neural network is applied to the digital modulation recognition. In order to improve the performance based on the standard artificial bee colony( ABC) algorithm,in the algorithm,the onlooker bees' search behavior towards food source's neighborhood is improved. Simulation result shows that the method has good performances because when the SNR is low to 0 d B,the probability to recognize seven digital modulation signals is over 85%.

关 键 词：数字调制技术 信噪比 人工蜂群算法 BP神经网络 跟随蜂 

分 类 号：TN911.3[电子电信—通信与信息系统]