室内定位RSSI空间建模与接收设备偏差研究

发布时间：2018-04-21 08:52

本文选题：WIFI定位 + 接收信号强度值　；参考：《西南交通大学》2017年硕士论文

【摘要】：随着当代信息技术的发展,位置信息的获取应用在人们生活的各个方面,基于位置的服务(LBS)逐渐由室外转向了室内。近年来,无线局域网技术得到了快速的发展,同时用户终端设备趋于智能化,无线定位成为了室内定位的主要方式。无线信号室内空间传播具有很高的复杂性,传统交会定位方式并不适用,因此,本文主要研究内容为基于位置指纹的WIFI室内定位。本文分析了 RSSI(Received Signal Strength Indication)空间分布特性,建立基于RSSI空间分布的Radio Map,并通过位置指纹定位算法分析Radio Map的可靠性。利用线性回归修正异构设备RSSI差异问题,对校正定位前后进行了精度分析与评定。在RSSI空间建模阶段,首先,通过室内环境下RSSI采集试验,验证了复杂环境下RSSI概率分布特性与空间可区分性。然后,分析了 RSSI特征提取与离散指纹库建立两个阶段噪声,针对RSSI观测噪声采用奇异值剔除和高斯滤波,提高了 RSSI信号信噪比和可用性,利用邻域均值算法对RSSI位置指纹库进行滤波,提高了 RSSI与空间的匹配程度。接着,利用三次曲面插值算法,对离散型位置指纹库进行了内插,建立连续型RSSI空间模型Radio Map,提高了离散指纹库建立的工作效率。最后,分别采用确定型与概率型指纹识别算法进行了定位试验,分析了两者定位精度,并验证了Radio Map建立的可靠性与正确性。在定位阶段,利用室内环境下异构设备动态与静态RSSI观测序列,分析了异构设备之间存在的RSSI偏差,说明了其大小。对异构设备同步动态RSSI序列进行了相关性分析,论证了异构设备RSSI之间存在正相关的线性关系,利用线性回归对异构设备RSSI之间进行了线性关系建模,并标定了模型参数。通过RSSI修正前后的定位试验,说明了异构设备RSSI修正方法的正确性。
[Abstract]:With the development of modern information technology, location information acquisition and application in all aspects of people's lives, location-based services gradually shifted from outdoor to indoor. In recent years, wireless local area network (WLAN) technology has been developed rapidly. At the same time, user terminal equipment tends to be intelligent, wireless positioning has become the main way of indoor positioning. Wireless signal propagation in indoor space has a high complexity, traditional rendezvous location method is not applicable, therefore, the main content of this paper is WIFI indoor location based on location fingerprint. In this paper, the spatial distribution characteristics of RSSI(Received Signal Strength indication are analyzed, the Radio map based on RSSI spatial distribution is established, and the reliability of Radio Map is analyzed by the location fingerprint location algorithm. Using linear regression to correct the RSSI difference of heterogeneous equipment, the accuracy analysis and evaluation are carried out before and after calibration. In the stage of RSSI spatial modeling, the characteristics of RSSI probability distribution and spatial separability in complex environment are verified by the RSSI collection experiment in the indoor environment. Then, the RSSI feature extraction and the establishment of discrete fingerprint database are analyzed. The singular value elimination and Gao Si filter are used to improve the signal-to-noise ratio and availability of RSSI signal. The neighborhood mean algorithm is used to filter the RSSI location fingerprint database, which improves the matching degree between RSSI and space. Then, the interpolation algorithm of cubic surface is used to interpolate the discrete position fingerprint database, and the continuous RSSI spatial model Radio map is established, which improves the working efficiency of the discrete fingerprint database. Finally, the localization experiments are carried out by using the deterministic and probabilistic fingerprint identification algorithms, and the accuracy of the two algorithms is analyzed, and the reliability and correctness of the Radio Map are verified. In the positioning stage, using the dynamic and static RSSI observation sequences of heterogeneous devices in indoor environment, the RSSI deviation between heterogeneous devices is analyzed and its size is explained. The correlation analysis of synchronous dynamic RSSI sequences of heterogeneous devices is carried out, and the positive linear relationship between RSSI of heterogeneous devices is demonstrated. The linear relationship between heterogeneous devices RSSI is modeled by linear regression, and the model parameters are calibrated. The correctness of the RSSI correction method for heterogeneous equipment is demonstrated by the positioning test before and after RSSI correction.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN925.93

【参考文献】