Research On Anomaly Aware WSNs Node Localization Algorithms In Complex Environments

With the emergence of various low-power,low-cost,multi-functional miniature sensors,Wireless Sensor Networks have gained tremendous development.As an important supporting technology for WSNs,node localization has always been a hot research area for scholars at home and abroad.The premise of many WSNs applications to work effectively is to obtain enough accurate location information,which conveys the connection between the data perceived by the sensor and the place where it occurs.Without this key information,the value of WSNs applications will be greatly reduced.At present,scholars have also proposed many loalization algorithms suitable for various application scenarios,which are usually divided into range-based localization algorithms and range-free localization algorithms.Usually,the range-based localization algorithms have higher localization accuracy and a wider range of applications.Therefore,this paper focuses on the research of range-based localization algorithms.Generally,in practical applications,due to the influence of factors such as complex environments,hardware anomalies,and network attacks,the ranging information obtained is often incomplete and inaccurate,resulting in reduced localization accuracy.To address this challage,two WSNs localizaton algorithms are designed based on low-rank matrix decomposition technology in this paper.The main research contents and innovations of this article are as follows:1.Aiming at the problem of incomplete and inaccurate ranging information,a noise-immune localization algorithm via low-rank matrix decomposition?Lo CMD?for wireless sensor networks is proposed.This algorithm introduces the popular low-rank decomposition technology in the field of machine learning,which can complete the missing part of the ranging information and avoid the problem of limited scalability caused by solving the kernel norm minimization matrix completion problem.At the same time,we use the Mixture of Gaussians model to fit arbitrary distributed noise in ranging information.Experiments show that this algorithm can accurately estimate the missing ranging information between nodes in complex environments and obtain good node localization results.2.Aiming at the problem that abnormal nodes may exist in the localization process,an anomaly-aware node localization?ANLo C?algorithm is proposed.The algorithm is based on the Lo CMD algorithm and uses the?_2,1-norm to smooth the structured anomalies in the ranging information.It can achieve accurate node locazilation information and detect the node anomalies,which provides a scientific basis for the maintenance and management of WSNs.Experiments show that the algorithm can achieve excellent localization results in complex environments and can detect abnormal nodes well.Finally,this paper also extends ANLo C to the problem of node localization in large-scale scenarios.Simulation experiments show that this extended algorithm can also perform well.