Three-dimensional Indoor Positioning Based On Multi-dimension Resource Optimization

Positioning information is critical in both our daily life and emergency situations like earthquake, debris flow, etc. Cisco's report reveals that people spend almost 80% of the daily time in indoor environment. In indoor environment, building block and multipath effect result in low-precision positioning which cannot meet the demands of some data services. In order to further improve indoor horizontal and vertical positioning accuracy, the main work of this thesis is as follows:Aiming at two typical 3GPP indoor positioning scenarios, we build a 3GPP-oriented indoor positioning simulation platform which can verify the results of our proposed indoor positioning methods. Then we analyze the effects of indoor user density, small cell density and indoor small cell deployment on indoor positioning accuracy. The simulation results show that high-density small cell can improve horizontal positioning accuracy by 54.2% and vertical positioning accuracy by 35.3%, compared with that of low-density small cell; indoor small cell deployment can improve horizontal positioning accuracy by 73% and vertical positioning accuracy by 47% compared with that under outdoor deployment; while changing indoor user density has no significant effect on both horizontal and vertical positioning accuracy.From the perspective of spatial-domain resource optimization, a new positioning reference node selection metric named correlation metric (CM),is proposed, related to both the distance and positioning signal quality.CM's influence on positioning accuracy are analyzed by graph theory. On this basis, a new positioning node selection algorithm named distributed positioning node selection (DPNS) is proposed to improve CM. The simulation results show that this method can improve horizontal positioning accuracy by 56% and vertical positioning accuracy by 33.3%compared with that of the traditional one.Further, from the perspective of joint optimization of multi-dimension resources including time-domain resource, frequency-domain resource and spatial-domain resource, a new indoor positioning enhancement method,named C-P positioning enhancement method, is proposed in order to improve vertical positioning accuracy. This method utilizes CRS together with PRS for indoor positioning according to the distribution of CRS and PRS in OFDM subframe. The simulation results reveal that vertical positioning accuracy can be improved almost 30% by C-P positioning enhancement method compared with that of the single-port PRS positioning.