网络中高速信息移动节点位置预测仿真

Simulation of Location Prediction of High Speed Mobile Node in Network

对网络中高速信息移动节点位置进行预测,能够提升网络数据采集和信息转发的质量。对移动节点位置进行预测时,由于节点位置的真实值和预报值存在误差,导致得到的移动节点运动累计时间过长。传统方法利用均值漂移和联合粒子滤波,获取移动节点运动累计时间,但过程复杂,导致节点位置预测精度偏低。提出基于马尔科夫链的网络中高速信息移动节点位置预测方法。融合于ARIMA模型预测出节点下一时刻的地理位置,获取节点的累计保持时间预测值,计算出节点位置真实值和预报值之间的均方误差,得到高速信息移动节点运动累计保持时间,融合于马尔科夫链原理计算出高速信息移动节点到达所有位置的概率分布,完成对网络中高速信息移动节点位置预测。仿真证明所提方法对高速信息移动节点位置预测更准确。

In this paper, the author proposes a location prediction method of high-speed information mobile node in network based on Markov chain. Firstly, geographic location of node on next moment integrated with ARIMA model was predicted and prediction value of node cumulative holding time was acquired. Then mean square error between true value and prediction value of node location was worked out and movement cumulative holding time of high-speed information mobile node was acquired. Moreover, the probability distribution of high-speed information mobile node arriving all location integrated with the Markov chain theory was worked out. Simulation results show that the method can predict the location on high-speed information mobile node more accurately.