基于线性插值和动态指纹补偿的分布式定位算法

A Distributed Localization Algorithm Based on Linear Interpolation and Dynamic Fingerprinting Correction

针对现有射频指纹匹配定位算法存在训练开销过大、环境信号动态改变自适应能力较差以及系统扩展性不足等问题,该文提出了一个基于线性插值和射频指纹动态补偿的分布式定位算法。该算法利用同一信标发射无线信号局部物理空间的相关性,提出了射频指纹的线性插值方法,减少了训练数据的采集工作量。此外,为自适应环境无线信号的动态变化,该算法使用少量均匀部署在定位区域内的参考节点实时监测环境信号变化,对移动节点测量的RSSI进行动态补偿,从而提高定位精度。实验结果表明,所提定位算法不仅有效减少了指纹数据库的创建工作量,而且能够较好地自适应环境变化,具有较高的定位精度。

In order to reduce the training cost and improve the robustness to environmental change for the radio-fingerprinting-based positioning systems, this paper presents a distributed localization algorithm based on linear interpolation and dynamic fingerprinting correction. It utilizes the spatial correlation of measurements sampled at adjacent reference points to obtain new fingerprints by linear interpolation and deploys a few reference nodes to dynamically compensate mobile node＇s RSSI. Extensive experiments show that the proposed algorithm not only achieve high localization accuracy, but also save training time and is a- ble to adapt to the environmental change.