MULTI-STEP PREDICTION OF FREQUENCY HOPPING SEQUENCES BASED ON BAYESIAN INFERENCE

摘要：

　　According to the chaotic characteristics of frequency hopping (FH) sequences and the short-term predictability of Chaos,this paper presents an improved Bayesian network predictive model applied to FH sequences prediction.Firstly,the model regards the entire reconstructed phase space as a prior data information; Then,according to the characteristic of FH sequences which consist of multiple frequency points,it constructs a local Bayesian network with the mutual information and an algorithm for Markov boundary; Finally,it achieves the multi-step prediction of FH by using the posterior inference algorithm.Theoretical results and large number of experiments show that the proposed Bayesian network predictive model has steady,real-time,effective and high-precision multi-step prediction ability,especially in small data set.Thus this model provides a novel method for the research and application of FH sequences prediction.

关键词： FH sequences Bayesian network phase space multi-step prediction

作者: Wensheng Wang Youlong Yang Yanying Li

作者单位: School of Science,Xidian University,Xian 710071,China Department of Mathematics,Baoji University of Arts and Sciences,Baoji,China

会议类型: 国际会议

会议名称: 2013IET International Conference on Information and Communication Technologies(IETICT2013)2013IET信息与通信技术国际会议

会议地点: 北京

会议语种:英文

页码: 75-80

在线出版日期: 2013-04-27（万方平台首次上网日期，不代表论文的发表时间）