Research On Multi-source Information Sensing Based WLAN Indoor Localization Accuracy Bound

With the development of wireless communication technology,the demand for wireless localization has been significantly studied.In outdoor environment,the localization system such as Global Positioning System(GPS)has been widely used.In indoor environment,the localization performance of outdoor localization systems including GPS deteriorates dramatically due to complex indoor structures and multipath effect.With the large-scale deployment of Wireless Local Area Network(WLAN),WLAN fingerprinting based localization has drawn much attention which makes use of available WLAN infrastructures.Meanwhile,due to the significant practical guidance for localization databased construction and optimization,the WLAN indoor localization accuracy bound is widely studied by researchers.To overcome the problems such as poor stability and limited scope of application exsiting in previous research,the relations between signal distribution characteristics and WLAN indoor localization accuracy bound are firstly constructed in this thesis.Based on this,the user behavior characteristics are introduced to derive the relations between multi-source information characteristics and localization accuracy bound.In all,three contributions of this thesis are listed as follows.1.The relations between signal distribution characteristics and WLAN indoor localization accuracy bound are constructed.Firstly,the process of WLAN localization is analogized into the one of information propagation in a parallel Gaussian noisy channel,and then the relations between the channel capacity of analogical communication system and the corresponding localization error are formulated.After that,the relations between signal distribution characteristics and WLAN indoor localization accuracy bound are built to obtain the localization accuracy bound of target environment.Finally,based on simutaled and real data,the performance of proposed approach is validated.2.Based on the proposed estimation approach of localization accuracy bound,the localization database is optimized to reduce the construction and storage overhead of database.First of all,by analogizing the database construction process into the information propagation process in a lossy channel,the relations between sample capacity and localization error are formulated from an information-theoretic view.After that,by selecting an appropriate time interval,the minimum number of sampled signal under the given expected localization accuracy is determined.Finally,the performance of proposed approach is validated with real sampled data.3.Combining signal distribution characteristics with user behavior characteristics,the multi-source information sensing based WLAN indoor localization accuracy bound is derived.In concrete terms,first of all,the process of indoor WLAN positioning is analogized into the information propagation process in colored Gaussian noisy channel.Then,the motion behavior of the pedestrian in indoor environment is considered to formulate the relations between the information entropy of the belonging regions of users and the corresponding localization eroor.After that,the positioning error bound is derived out based on the channel capacity constraint with respect to the information entropy.Finally,experiments wth real sampled data are performed to verify the performance of proposed approach.