In The Wireless Sensor Network (wsn) Study Of The Problem Of Passive Location Based On Rss

Received signal strength (RSS) based device-free localization is a novel technology in wireless sensor network (WSN). The basic idea of this technol-ogy is to deploy WSN nodes around an area of interest. Every pair of nodes communicate with each other, forming a large number of wireless links. When people or objects (referred to as "objects" in the following paragraphs) enter this area, radio waves of the links nearby would be absorbed, reflected, or d-iffracted, and the RSS values of these links would be changed. Based on these variations of RSS measurements and the positions of wireless links, as well as appropriate models and algorithms, we can then localize the locations of ob-jects in real time. The advantages of RSS-based device-free localization lie in the fact that the objects to be localized do not need carrying any device (so we call it "device-free"), and that radio waves can not be affect by illumination or obstruction. These advantages bring broad application prospects and great development potential to RSS-based device-free localization technology, es-pecially in areas like security-and-monitoring, search-and-rescue, military and etc.First, this paper proposes a background learning and motion imaging based method for RSS-based device-free localization. This method can rapidly adapt to areas with complex and time-varying multipath environments, achiev-ing real-time imaging of multiple moving objects. We first incorporate two basic background learning methods from image processing area to RSS-based device-free localization, so as to calculate the probability of each link being affected by objects from the RSS value of that link. Then, based on the af-fecting probability of all links, we apply Tikhonov regularization algorithm to obtain a2D distribution image of the moving objects. From the value distribu-tion of pixels in the image, the number and locations of objects can be easily estimated. Experiments show that, compared to existing methods, the proposed method has higher localization accuracy and stronger environmental adaptive ability.Second, in this paper, we present a histogram feature and detection win-dow based method for RSS-based device-free localization. This method can real-time estimate the number and locations of objects, in areas with very dif-ferent multipath environments. First, through single-link experiments, we find that different links show the same histogram feature when obstructed by object-s. Then, based on this feature, detection window is further induced to localize the coordinates of multiple objects. At last, we propose to connect detection windows with different histogram precisions in a cascade, so as to effectively reduce the computation complexity. Experiments show that, compared to exist-ing methods, the proposed method has higher localization accuracy and lower computation complexity.Finally, we analyze and compare the two proposed methods in such per-spectives as localization accuracy, environmental adaptive ability, and compu-tation complexity. Moreover, we conclude the contents of the whole paper, and give an outlook of the RSS-based device-free localization technology for future research.