中文摘要：

移动 Ad Hoc 网络中，节点之间的路径长度是构建 QoS 路由的一个重要参数，提出一种基于移动平均（MA）模型的网络路径长度预测方法。首先根据赤池信息量准则（AIC）确定 MA 模型的阶，然后对路径长度的历史和当前数据进行时间序列分析，找出之间的自相关性，从而预测未来路径长度。建立仿真实验，在 AODV 和DSR 网络协议下，节点分别执行 MHG 和 RPGM 移动模型，通过提出的方法对各种场景下的路径长度进行预测，并根据性能指标评估预测性能。结果表明，提出的方法能够有效地预测未来路径长度。

英文摘要：

The path length between nodes is an important parameter to build QoS routing in mobile ad hoc network,therefore a method base on Moving Average （MA）model that predict path length of network is proposed.Firstly,the order of MA model is determined according to Akaike Information Criterion （AIC）.Secondly,the time series of the historical and current path length data is analyzed to find out the self-correlation so as to predict the future path length.In simulation experiments,the nodes execute MHG and RPGM mobile models with AODV and DSR network protocol respectively,the proposed method is been used to predict the path length in different scenario and the performance of prediction is evaluated according to performance index.Experimental results show that the proposed method can predict the future path length effectively.