Collaborative Localization Of Network Vehicles Based On Approximate Messaging Passing

With the evolution of 5G commercial use,intelligent networked vehicles will be widely used in intelligent transportation systems.Equipped with on-board sensors and communication equipment,intelligent networked vehicles are connected to modern communication network and intelligent information exchange can be realized.The location information of networked vehicles plays an important role in many applications such as automatic driving.High-precision localization can be achieved by cooperative localization.This dissertation makes an in-depth localization research on the design of cooperative localization algorithm,the non-line-of-sight identification and the parameter estimation in cooperative localization.The main contributions are as follows:(1)The design of varied types of ranging measurements and non-line-of-sight cooperative localizers:a three-dimensional universal cooperative localizer(3D UCL)is proposed for networked vehicles in 3D space under varied types of ranging measurements including time-of-arrival(TOA),received signal strength(RSS),angle-of-arrival(AOA),and Doppler frequency.Its core idea is to exploit generalized approximate message passing(GAMP)to resolve the 3D cooperative positioning problem after converting it as a generalized linear mixing problem.Unfortunately,the positioning performance of 3D UCL is severely degraded by the inaccurate ranging measurements from the non-line-of-sight(NLOS)links.Therefore,a 3D geographical information enhanced UCL(3D GIE-UCL)is developed by combining 3D UCL with a NLOS identification mechanism assisted by geographical information.Finally,3D UCL is accelerated by graphics processing unit(GPU)parallelization,particle reduction and message censoring.3D GIE-UCL is accelerated by particle reduction and anchor upgrading.Simulation results validate state-of-the-art positioning performances and cooperative gains of both 3D UCL and 3D GIE-UCL after comparing them with existing cooperative localizers.3D UCL and GIE-UCL show 241x and 3.3x speedup after adopting the acceleration techniques,respectively.(2)The design of cooperative localization considering known and unknown parameters:based on 3D AOA measurements,cooperative localizers are designed with known or unknown Euler angles.Two types of localizers are proposed based on GAMP and least-square(LS)method,which work under(un-)quantized 3D AOAs,(un-)known Euler rotation angles,and ?/2?-periodic azimuth AOAs.Firstly,when the Euler angles are known,GAMP localizer is obtained by categorizing the 3D AOA-CL problem as a generalized linear mixing one,and resolving the latter by GAMP,whose mean-and-variance messages involving(un-)quantized 3D AOAs are evaluated by importance sampling technique.Type-? LS localizer is developed independently under ?/2?-periodic azimuth AOAs.Under the 2?-periodic case,there exist tan/cos-relationships between AOAs and xyz minus results for two neighboring VANET nodes.Therefore,loss function with square error is defined to measure how the above relationships are violated.Its value is minimized through optimizing vehicle positions by the gradient-descent method.Under the ?-periodic case,additional 0/1 integers are introduced to indicate the front-or-back impinging information.They are relaxed as continuous variables ranging into[0,1),and then reconstructed by the grandient descent(GD)method jointly with vehicle positions.Secondly,when the Euler angles are unknown,the expectation-maximization(EM)framework is combined with GAMP,where vehicle positions and Euler angles are alternatively updated through one-step GAMP iteration and maximizing conditional probability distribution functions(pdfs)expected over xyz-minus variables,respectively.3D Type-? LS localizer is developed to iterate between estimating Euler angles by maximizing the probability function between 3D AOA measurements and Euler angles,and recovering vehicle positions by 3D Type-I LS localizer.Simulation results validate that the proposed two localizers outperform existing localizers.GAMP-type localizers outperform LS-type ones,and are capable of handling nonlinear quantization losses.Vehicle positions and Euler angles are alternatively recovered with the progress of outer-iteration in EM-GAMP and Type-II LS localizers.