SSA Improved TDOA Algorithm In UWB NLOS Environment

Nowadays,with the maturity of 5G technology and development of smart cities,people’s production and working environment is gradually moving from outdoor to indoor.Many enterprises and government agencies have increasing demands for high-precision indoor positioning.However,owing to the intricate indoor channel environment and the easy interference of obstacles in the process of signal propagation,resulting in multipath Clutter and Non-line-of-sight(NLOS)errors make the positioning system affect the positioning accuracy of the target,resulting in a significant increase in the systematic error of the positioning system.In the indoor positioning system,the positioning system using the Ultra Wide Band(UWB)signal as the transmission carrier has many advantages such as high transmission rate,strong anti-interference ability,simple structure,and high security,which has a leading position in other common indoor positioning systems with high transmission rate,strong anti-interference ability,simple architecture,high security and many other advantages,and has a broad market prospect.Among many positioning algorithms,Time Different of Arrival(TDOA)is widely used because of its high positioning accuracy,simple construction environment.In this paper,the basic concepts of UWB positioning technology.First of all,the domestic and foreign research status and the main positioning models are discussed in depth.This paper introduces the IEEE802.15.4a model of UWB system and the modulation mode of UWB signal as well as the common positioning algorithms,compares the performance of several positioning algorithms,among which TDOA algorithm is more suitable for indoor positioning and has been studied in depth.Secondly,the positioning accuracy of conventional TDOA algorithm is easily affected by the non-line-of-sight error,the problem of inaccurate positioning accuracy and low algorithm accuracy are caused.In this paper,an improved TDOA localization algorithm based on BP neural network and salp swarm algorithm is proposed,when the program runs to the later stage of the iteration,the traditional algorithm is weak in leaping out of the partial optimum.We improve the original salps algorithm to improve this question.The algorithm first uses the improved salps swarm algorithm to obtain rough estimates,then substitutes it into the BP neural network for training and learning,and finally uses the output of the BP neural network as the final positioning result.Simulation results show that the stability and robustness of the proposed algorithm are significantly improved.