【摘要】 针对通信网络中的路由问题， 提出了一种虚令牌动态路由算法， 并用Ｈｏｐｆｉｅｌｄ 神经网络实现该算法．该Ｈｏｐｆｉｅｌｄ 神经网络， 其神经元之间的连接强度固定不变， 通信网络的物理链路值和呼叫请求被反映到神经元的偏置电流中， 易于用硬件实现．同时， 当通信网络中的业务量发生变化时， 算法能够自动调整最短路径与负载均衡之间的关系．计算机仿真结果表明， 该算法能接入更多的网络连接更多还原

【Abstract】 A dynamic routing algorithm based on virtue token is proposed to solve the routing problem in communication network, and a Hopfield neural net is used to realize the algorithm. In the algorithm, the interconnection matrix of the neural net is fixed, and the link costs and call requests of the communication network are mapped into the biases of the neurons. So the Hopfield neural net is easy to be implemented into hardware. Meanwhile, the algorithm can adjust the balance between shortest path and least load requirements according to the link load status of the network. The simulation results show that the network using this algorithm can accept more connections.更多还原