The Research On Indoor Location Algorithm Based On RSS Fingerprints

With the continuous development of the modern communication technology and the penetration of the smart phones or other smart terminals, people’s demands for the location-based services are increasing strongly, especially in indoor area. To get the needed information by using the Location-based services has already become an indispensable part of people’s daily life. In the indoor environment, the positioning accuracy by using the mature outdoor positioning technologies such as GPS, cannot meet the expectations. As a result, more and more people have begun to look for other location technologies. In recent years, the worldwide popularity of WLAN technology based on the IEEE 802.11, offers a new direction for the development of the indoor positioning.The RSS fingerprints algorithm is one of the most commonly used localization algorithm of WLAN. RSS fingerprints positioning method is first to be mapped the physical location to a fingerprint database, then matching the real-time RSS vector, and getting the location information. This paper firstly introduces the origin and development of indoor positioning and the research dynamic in domestic and foreign regions, then introduces the commonly used positioning technologies, and then focus on the classical localization algorithms based on the technology of WLAN, and the improved algorithms proposed by this paper. The complex indoor environment will make RSS instability and uncertainty, and lead to the phenomenon "many-to-many" between the RSS and the actual location. This brings great difficulty to the position. Therefore, before the orientation, therefore, the disturbance value of RSS vector should be removed before the position. To this point, by analyzing the distribution of D-RSS, the concept of RSS’s typicality is proposed firstly, and an indoor localization algorithm based on typicality judgment of RSS is also presented in this paper. According to the principle that the RSS values and the effective similar sample points, a typicality discrimination method for RSS values and a self-adapting K value are presented. When get the reasonable RSS values, we should obtain the actual position upon the similar sample points. In this paper, through the analysis on the characteristics of locating points, a localization algorithm with variable weights that adapt to changes of the different positioning scenes is proposed this position algorithm be identical with actual situation "the nearer distance, the more closely “By experimental verification, These two kinds of improved localization algorithm can not only greatly improve the positioning accuracy, But also have a certain environmental adaptability, and have the self-adaption ability to the changes of environmental factors.