基于RSSI技术的室内定位设备无关性研究

发布时间：2018-09-07 13:31

【摘要】：随着无线局域网(Wireless Local Area Networks,WLAN)的覆盖和4G网络的推广,人们能随时、随地、高速的接入互联网,获取自己需要的信息,诸多基于位置信息的应用也应运而生,新的问题也随之而来。由于不同移动终端硬件及实现差异,导致在定位过程中处于同一位置的移动终端采集到的无线访问接入点(Access Point,AP)信号强度(Received Signal Strength Indication,RSSI)不同,引起定位误差。为此,本论文首先在实际室内环境中采集WiFi数据,分析移动终端设备之间的差异性。在分析过程中,发现移动终端设备所获得的RSSI数值异常值,与环境瞬间变化有关。基于此,本文通过设定置信区间来修正同一AP获得的RSSI异常值,来提高定位准确度。其次,探索解决终端设备差异性问题的解决方案。通过对比不同设备在多时段所采集的RSSI数据,发现当不同设备在同一位置采集周围AP无线信号时,不同设备所采集的RSSI数值趋势大致相似。利用上述发现,借鉴加权K近邻(Weighted K-nearest Neighbor,WKNN)算法中权重因子的概念并结合Pearson相关系数,提出一种基于Pearson相似度的终端差异消除方法。在该算法中需要计算定位设备所采集的数据与位置指纹数据库内每一个采样点指纹数据之间的Pearson相关程度,并将计算所得的结果作为一个系数因子,解决不同设备所获得的RSSI数据不匹配的问题。最后,在基于Android的平台上验证了提出的基于Pearson相似度的终端差异消除方法的有效性。实验将三款不同智能终端设备分别与不同的现有算法相比较,发现本文提出的方法有效地降低了终端设备差异性,减少了室内定位误差,提高了室内定位准确性。实验结果表明,对于同构设备,基于Pearson相似度的终端差异消除方法在1.5m以内的定位精度保持在85%以上;对于异构设备,定位精度保持在70%以上。
[Abstract]:With the coverage of wireless local area network (Wireless Local Area Networks,WLAN) and the promotion of 4G network, people can access the Internet at any time, anywhere and at high speed to get the information they need. Many location-based applications have emerged, and new problems have followed. Because of the difference in hardware and realization of different mobile terminals, the wireless access point (Access Point,AP) signal intensity (Received Signal Strength Indication,RSSI (Received Signal Strength Indication,RSSI) collected by the mobile terminal in the same location is different, which results in the location error. Therefore, this paper firstly collects WiFi data in the actual indoor environment and analyzes the differences between mobile terminal devices. In the process of analysis, it is found that the outliers of RSSI obtained by mobile terminal devices are related to the instantaneous change of environment. Based on this, this paper modifies the RSSI outliers obtained by the same AP by setting the confidence interval to improve the localization accuracy. Secondly, explore the solution to the terminal equipment difference problem. By comparing the RSSI data collected by different equipments in different time periods, it is found that when different devices collect the surrounding AP wireless signals in the same position, the RSSI values collected by different devices are roughly similar. Based on the above findings, using the concept of weight factor in weighted K nearest neighbor (Weighted K-nearest Neighbor,WKNN) algorithm and Pearson correlation coefficient, a terminal difference cancellation method based on Pearson similarity is proposed. In this algorithm, we need to calculate the Pearson correlation between the data collected by the location device and the fingerprint data of each sampling point in the location fingerprint database, and take the calculated results as a coefficient factor. To solve the problem of RSSI data mismatch obtained by different devices. Finally, the effectiveness of the proposed terminal difference cancellation method based on Pearson similarity is verified on the platform of Android. By comparing three different intelligent terminal devices with different existing algorithms, it is found that the method proposed in this paper can effectively reduce the difference of terminal equipment, reduce the indoor positioning error and improve the accuracy of indoor positioning. The experimental results show that for isomorphic equipment, the location accuracy of the terminal difference cancellation method based on Pearson similarity is more than 85% within 1.5m, and that of heterogeneous equipment is more than 70%.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN925.93

【参考文献】