Research On Multi-target Location Technology Based On RA-RSSI Fingerprint Database

With the continuous development of wireless communication technology,various application services based on location emerge in an endless stream.Owing to the advantages of low power consumption and low cost,wireless sensor networks(WSNs)technology has been widely used in smart homes,environmental monitoring,military reconnaissance,medical treatment and other fields.In addition,WSNs can also work in a special environment that humans cannot directly reach in virtue of their wide distribution and wireless connection methods,and achieve target location,data collection and processing.WSNs can continuously collect information in the environment through many sensor nodes.The location information of the monitoring target plays a key role in various application scenarios.The acquired data must be related to the location of the nodes to make sense.Therefore,the WSNs target node location technology has become an important topic that has widely attracted attention.Compressed sensing technology is a new sampling technique in the field of signal processing.It can accurately restore the original signal through a small amount of sampled data.It is of more practical value in the modern information society where the amount of data is soaring.The research of applying CS technology to WSNs multi-target location problem has received extensive attention from scholars,and it has also achieved certain results.In the multi-target localization problem of WSNs based on CS,the localization effect is closely related to the signal received by the random deployed reference nodes in the use of Received Signal Strength Indicator(RSSI).Furthermore,the multi-target positioning problems are mostly based on the unknown number of nodes in actual situation.Therefore,the research of RSSI fingerprint database and reconstruction algorithm are put forward,RA-RSSI fingerprint database is designed and VS-GMP algorithm is proposed.They make the location system able to resist abnormal receiving signals strength indicator from damaged anchor nodes with ensuring the reliability of the node's sending and receiving signals,reduce the amount of location calculations and improves the location effect in the case where the number of targets is unknown.In the design of RA-RSSI fingerprint database,an interpolation optimization algorithm is proposed to establish and update the fingerprint database.The interpolation function is divided into two parts: deterministic attenuation function and random variation function.The former function can be solved by a small number of measurement values combined with the fitting method,and the latter function is solved by Kriging interpolation method.In order to select the damaged anchor nodes,a new voting mechanism is proposed to determine the credibility of each anchor node,and the change rule of online standard deviation is used to determine the degree of damage.Then,the corresponding data in the fingerprint database can update.In the VS-GMP algorithm,an idea of variable step size combined with double thresholds is proposed.It can adaptively adjust the step size according to residuals,and uses double thresholds to strictly control the iteration conditions.The simulation results show that,the proposed RA-RSSI fingerprint database can update the data in time and avoid the serious location error caused by the damage of the anchor node.It have stronger adaptability compared with the traditional fingerprint database,and has improved the accuracy and robustness of the location system.The VS-GMP algorithm is faster and more efficient compared with other greedy reconstruction algorithms,it has obvious advantages in the multi-target location problem with blind sparseness.