Ad-Hoc网络中基于状态转换概率的中继选择算法研究

在多跳Ad-Hoc网络中，随着节点的增加以及传输跳数的变化，网络状态数目将不断增大，随之带来的路由计算将变得十分复杂，从而严重影响系统的整体性能。如何在海量的网络转换状态形势下，快速选择最有效的状态进行下一跳数据传输是关键。创新地提出了基于状态转换概率的中继选择策略，结合节点地理信息和信道环境信息计算出邻居节点可能的到达概率，并选择概率大者进行数据传输，从而提高数据传输性能。同时，为了降低系统运算复杂度并节省系统能量，采用了变换的metropolis选择准则，以模拟退火的贪心搜索逐级去除那些小概率的传输状态，从而大幅度降低了运算空间。仿真给出了算法参数对运算速度与成功率的影响。同时，也表明了该算法在网络拓扑变化时对系统能耗和失败概率的增长均有较好的控制。

Research on relay selection algorithm based on state transition probability in Ad-Hoc networks

In the multi-hop Ad-Hoc networks, the number of network states will increase as the number of hops and nodes increases. Then, the computation will become very complicated and the system performance will be affected seriously. In the condition of massive network states, it is crucial to choose the most effective state to the next hop. So, we innovatively propose the relay selection strategy based on the state transition probability. First, we compute the transition probability of each neighbor according to the geographic information and channel environment information. After that, the state with max transition probability will be selected as the next transmission state. Thus, the system performance can be improved. In order to reduce the state space and save system energy, we adopt the transformed Metropolis criterion and remove the states with small probabilities by the greedy search of simulated annealing. The simulation gives the influence on the operation speed and the success rate with the algorithm parameter setting. At the same time, it also shows that the algorithm can control the growth of the energy consumption and the failure probability when the network topology changes.