Random estimation self-learning neighbor discovery method based on time window

The invention discloses a random estimation self-learning neighbor discovery method based on a time window. The method belongs to the communication network neighborhood, and mainly solves the problems of high neighbor node discovery difficulty and long convergence time in the directional self-organizing network. According to the method, network nodes in a neighbor discovery problem are modeled into a learning automaton, then the automaton can adapt to constantly changing environments by using a random estimator and an optimization method based on a time window, and the selection probability of each sector is adjusted in real time; therefore, the neighbor discovery speed in a wireless ad hoc network scene using the directional antenna is accelerated. An MATLAB analog simulation result proves that the robustness of the algorithm in accelerating the discovery process and complex scenes is achieved. 该发明公开了一种基于时间窗的随机估计自学习邻居发现方法。该发明属于通信网络邻域，主要解决了定向自组织网络中邻居节点发现难度大，收敛时间长的问题。该方法将邻居发现问题中的网络节点建模为学习自动机，然后利用随机估计量和基于时间窗的优化