基于LANDMARC的室内定位算法改进

Improved Indoor Location Algorithm Based on LANDMARC

在无线通信领域,LANDMARC算法以其低廉的造价成本,成为目前应用最为广泛的室内定位算法之一。而在实际应用中,由于RFID阅读器读取数值的离散性,导致计算中存在异常关联度,严重影响了定位稳定性;同时受到多径效应的影响,LANDMARC算法定位过程中获取的信号强度值存在偏差,易产生较大定位误差。对此,提出一种改进的LANDMARC算法,在原算法流程中引入自修复机制。该机制分为关联度修复和坐标修复两个阶段,关联度修复阶段在关联度计算完成后,生成修复值α,对引发定位过程紊乱的关联度异常值进行替换,保证算法的稳定性;而坐标修复阶段则在获取目标的理论位置后,通过最近邻居标签的平均偏差值产生修复值β,来修正目标标签的理论位置,以减少算法定位误差。仿真结果表明,该算法能够有效提高定位精度,同时增强了定位稳定性,使LANDMARC算法具有自我修复能力。

In the field of wireless communications,LANDMARC is one of the most widely used indoor localization algorithms due to its lowcapital cost.However,in the actual application,because of the discreteness generated by RFID reader's reading numerical value,abnormal correlation degree exists in the calculation,which seriously influences the positioning stability.At the same time,under the influence of the multipath effect,deviation exists in signal strength value in the process of positioning by LANDMARC,which is easy to produce larger positioning error.For this,we propose an improved LANDMARC where selfrepairing mechanism is introduced to the process in the original algorithm,which can be divided into two phases,namely,relational degree repairing and coordinate repairing.After the correlation calculation is completed,repairing value α will be generated in the correlation repairing phase,to replace the correlation degree abnormal value which leads to disorder of positioning process,ensuring the stability of the algorithm;and the coordinate repairing phase,after obtaining target's theory position,will generate repairing value β by the mean deviation value of nearest neighbors label,to modify the theory location of target label in order to reduce algorithm positioning error. The simulation shows that the improved algorithm can effectively improve the positioning accuracy,and at the same time enhance the positioning stability,making LANDMARC have ability to repair itself.