中文摘要：

针对节点移动对定化实时性能的影响，通过预测节点的移动趋势平衡定位时延，提出一种基丁梯度搜索的移动协作定位算法．该算法将移动定佗拓展到时间区间上进行．存每个时间区间内，采用一跳距离测疑和胍缩感知方法构建节点距离数据，并通过预测节点移动趋势确定网络中节点间距离矩阵的边界条什；然后通过梯度搜索法求解节点在特定时间区间内的最佳估计，从而求得各节点的相对坐标；如果网络中有足够锚扩电，则可将相对坐标转换为绝对位置．仿真结果表明，与其它移动定位算法相比，提出的方案有效提高了定化精度，而且在有测距跌是的环境下也表现出较好的定位性能．

英文摘要：

Consider that the impact of mobility on real-time performance of node localization, a novel localization algorithm based on gradient search is advanced. The algorithm considers the mobile localization problem from a temporal standpoint. In each time period, the algorithm computes distance among nodes from one-hop distance by compressive sensing, and derives distance boundary condition for network distance matrix based on the prediction of movement. And then the gradient search method is employed to derive the optimal estimated position of nodes so as to calculate the relative coordinates of nodes. Finally, in condition that there are three or more anchor nodes, the global absolute coordinates is created by using the global relative coordinates. Simulation results show that compared with previous literatures, the new algorithm can achieve preferable estimation accuracy and perform well on range error.