神经元网络自适应时延同步及保密通信应用

Adaptive synchronization in neural network with time-delay and application in secure communication

考虑到时延在神经系统的普遍存在性,基于Lasalle不变性原理提出了一种新的自适应时延的同步方法.该方法的同步反馈控制强度不仅依赖于初始位置,还依赖于系统的时间延迟和反馈控制系数.利用Hindmarsh-Rose(H-R)神经元网络进行数值仿真验证其有效性和可行性,最后将该方法应用到保密通信系统中,实现保密通信.

As the delay exists in the neural system, in this paper, an adaptive time-delay synchronization method was presented based on the Lasalle invariance principle. The feedback control strength of this method not only depends on the initial position, but also depends on the system time-delay and coefficient of feedback control- Moreover, the Hindmarsh-Rose （H-R） neural networks were used to verify its effectiveness and feasibility. At last, the method was applied to secure communication systems, achieving secure communication.