Localization Algorithm Of Wireless Sensor Network Under Complex Environment

Wireless sensor networks are providing with the characteristics of low-power, low cost, multi-purpose, so it has been widely used in many areas, such as automatic control, construction industry, environmental monitoring, and dangerous area, especially for the environment area where is dangerous and people can’t reach. In practical applications, sometimes the network system which deploys a large number of sensor nodes in the monitoring field by aircraft scattering randomly. The raw sensor data must be combined with spatial information to be meaningful. Therefore, it is significant to realize the self-localization of wireless sensor nodes, so how to improve the accuracy of positioning and provide more high-quality positioning service is a hot research topic in recent years.In a complex environment, because of the radio interference arises by many factors, so the existing algorithms, such as ToA(Time of Arrival), TDoA(Time Difference of Arrival), Ao A(Angle of Arrival), RSS(Received Signal Strength) and other location algorithms will produce large error.Under National Science & Technology Pillar Program during the 12 th Five-year Plan Period under Grant No. 2012BAD10B01, the paper mainly focus on the localization algorithms for the complex environment, and have a deep research on the weighted centroid algorithm for fuzzy reasoning and RSS algorithm, the main contributions are as follows:(1)Establishing a mechanism for the RSS reliability of beacon nodes. When the beacon nodes are interfered by obstacles, the unknown node receives the RSS is unstable, so there will produce a big change within a range, in order to overcome the effects of the environment on the RSS, We often adopt to filter the RSS, to make the RSS value is more close to the real value. But sometimes RSS filtering with the signal blocked by obstacle cannot get into the real value, and the unknown node receives the RSS is small, so it will cause larger error. In this paper, we established the RSS reliability mechanisms, and using the reliability for the credible degree of beacon nodes emit radio waves. According to the RSS values received to identify the accuracy of the unknown nodes, If the RSS value of this beacon node receives is in a reasonable range, then the unknown node receives RSS values are accurate; if the signal of beacon nodes are disturbed, then the RSS value of the received node are not accurate, so we will consider the unknown node receives the RSS value is not accurate too.(2)We have proposed the centroid localization algorithm based on the Mamdani fuzzy probability of reliability, In wireless sensor networks, the weighted centroid localization algorithm is simple, so it is convenient to use in practical applications. However, if the beacon nodes are small and disturbed by obstacles in complex environments, the weighted centroid localization algorithm has a poor precision. And the accuracy of weights obtained by the fuzzy centroid positioning is not ideal. This paper proposes a centroid localization algorithm with probabilistic fuzzy logic using Mamdani type based on the accuracy of reliability, we can get the value through the Mamdani fuzzy reasoning. Then through the real environment to generate RSS membership function and membership function, reduced the interference about RSS by the complicated environmental factors.(3)We have proposed the Sugeno type fuzzy centroid localization algorithm based on the reliability of RSS, so as to improve the accuracy of the algorithm by through the Sugeno fuzzy reasoning. Through the adaptive fuzzy neural network in a complex environment to study the parameters of the fuzzy system, we achieved a better effect than the centroid localization algorithm based on the Mamdani fuzzy probability of reliability.(4)We have proposed the bat algorithm with inertia factor and Lévy flight strategy, it can be applied in the positioning system in WSN. So the global optimization to solve the positioning problems has become a hot research. Some scholars have put forward using the localization of genetic algorithm and swarm algorithm to solve the problem, but there exists the defects of low accuracy when in solving high-dimensional problems. This paper improved the traditional bat algorithm and proposed the bat Algorithm with Inertia Weight Factor and Lévy Flight, ILBA), the flying bat formula is improved in the two forms. The first strategy, we use the inertia factor similarly to the inertia factor in the particle swarm optimization(PSO) to keep the fly speed of the bat, and the inertia factor can make the ILBA to adjust adaptively. The other strategy is to adopt the guidance to guide the flying way, the strategy can expand search space and avoid falling into local optimum search. Finally the simulation test shows ILBA has a better solution accuracy and faster convergence on positioning problems than basic arithmetic BA and LBA algorithm.(5)We implements these two localization algorithm in an real mobile positioning system. There are different localization algorithms in different areas, and we can achieve a good positioning accuracy.