一种基于Hopfield神经网络的Adhoc网络QoS路由选择算法

当前研究Adhoc网络的QoS保证主要集中于QoS路由选择。人工神经网络已成为求解大规模优化问题的一种有效方法，已经证明合适的神经网络能实时地得到问题的精确解。主要考虑在Adhoc网络中满足时延条件下的最小耗费问题，建立了一种新的Hopfield神经网络模型，给出能量函数各参数之间的关系，并证明了通过适当选取参数，网络的可行解将是渐近稳定的。计算实例表明了新网络模型的有效性。

QoS Routing Optimal Algorithm in Ad hoc Networks Based on a New Hopfield Neural Network

The studies of QoS qualification in Ad hoc networks have focused on QoS routing algorithm presently. Artificial neural network has been an effective method for solving the large-scale optimal problems and has been proven to be able to get the exact solutions to the problem. In the paper, the minimum cost under delay constraints in Ad hoc network is studied. A new Hopfield network is proposed and the neural network could be robust to the changes of the network topology. A method about the relationships between network coefficients is proposed and it proves that the feasible solution to the problem is asymptotically stable if the network coefficients can be tuned appropriately. Finally an example is presented which proves the algorithm to be effective.