| Ubiquitous high-accuracy location information is the core foundation for the construction of a smart society.Indoor positioning by mobile communication networks has natural advantages in wide-area seamless coverage.Technologies such as multiple input and multiple output antennas enable the 5G network to provide a variety of high-precision heterogeneous positioning observations such as the angle of arrival and the time of fight in addition to the time difference of arrival.The fusion of those observations in the user equipment positioning can increase the amount of information redundancy and enhance the reliability of the positioning system,which has become one of the current research hotspots in the field of positioning and navigation.However,the ultra-dense network and other features of the 5G network also bring challenges to the multi-observation bybrid positioning.The main content and innovations of this thesis are as follows:1.Aiming at the problem of non-line-of-sight mitigation under the scarcity of error distribution models in 5G network wide-area coverage,this thesis analyzes the wireless signal observation information that 5G networks can use for positioning and the impact of non-line-of-sight effects on the measurement results.The non-line-of-sight state of the channel is identified based on the two-dimensional characteristics of time delay and angle,and the initial positioning result is obtained by using the line-ofsight source.A non-line-of-sight mitigation method based on homogeneous and heterogeneous observation information selecting is proposed,evaluating the positioning capability of homogenous heterogeneous measurements based on the change rate of dilution of precision and channel state,and the positioning result is optimized by selecting non-line-of-sight measurements,while ensuring the geometry construct while reducing the nonline-of-sight measurements to improve positioning accuracy.The simulation experiment results show that the positioning accuracy is improved by 10.42%(1?)compared with the non-line-of-sight mitigation based on membership level after using the method proposed in this thesis.2.Aiming at the problem of decreased positioning accuracy caused by enhanced inter-cell interference in 5G ultra-dense network,this thesis analyzes the characteristics of 5G downlink positioning reference signal and uplink sounding reference signal and the impact of inter-cell interference on measurement accuracy.The lower bound of the error of the multi-observation hybrid positioning is derived and used.A positioning user scheduling method based on the optimization of the overall positioning accuracy is proposed,reducing the inter-cell interference by coordinating user positioning methods.An interference weighted mitigation method based on the joint correction of heterogeneous observation information is proposed.The uplink signal angle of arrival and time of flight are used to jointly modify model prediction results of the interference intensity,and the influence of observation information under strong interference on the positioning results is mitigated by the weighting method,improving the positioning accuracy.Simulation experiments show that the method proposed in this thesis improves the positioning accuracy by 28.55%(1?)compared with the interference weighted mitigation method based on the reference distance model.3.Aiming at the problem of high-precision and high-efficiency feature fusion of heterogeneous observation information,this thesis analyzes the 5G network positioning logic architecture and mode.A multi-observation fusion positioning framework is designed.The close-form solution of the feature fusion positioning of distance,distance difference and angle is proposed,a mechanism for evaluating the accuracy of positioning results based on the combined residual weighted root mean square of heterogeneous measurements has been established,and the positioning result is corrected through non-heuristic low-iterative calculation,reducing the computational time while ensuring positioning accuracy.The simulation experiment results show that compared with the decision fusion method based on geometric dilution of precision,the positioning accuracy of the feature fusion method proposed in this thesis is improved by 23.45%(l?),while the calculation time is only slightly increased;compared with the feature fusion positioning based on particle swarm optimization,the positioning accuracy of the proposed method is improved by 17.21%(1?),and the average calculation time is reduced by 14.04%.4.A 5G network multi-observation hybrid positioning technology experimental environment and a semi-physical test platform is designed and built.The non-line-of-sight mitigation method,the inter-cell interference mitigation method,the fusion positioning solution method,and the integration of the three methods are tested and analysis.The results show that the proposed methods can effectively play a role in the multi-observation hybrid positioning,improve the positioning accuracy,and the three methods are compatible with each other,and the positioning accuracy is increased by 34.51%(1?)after integrated use. |
PDF Full Text Request