| Summary: | 碩士 === 國立中央大學 === 通訊工程研究所 === 100 === During the last few years, Wireless LAN (WLAN) has been rapidly growing and becoming more and more popular. In particular, the techniques of indoor location estimation are receiving a lot of attentions due to wide variety of service such as building the medical health care, tracking people for security issue, sustaining emergency supplement system, providing directional guidance, and so on. With the matured indoor location technology, the new demand on wireless applications is location-based service (LBS). It is toward the usage for the business, public securities and safety requirements. Therefore, increasing the accuracy of positioning is a significant issue for indoor location technologies.
In indoor environments, the shadowing of positioning environment, the multi-path and reflection phenomenon make the position estimation inaccurate. In traditional location fingerprint method shows that more Access Point (AP) deployments can obtain enough Receive Signal Strength (RSS) information to describe signal characteristics of the target space. However, it does not make sense and increase waste costs. On the other hand, insufficient AP deployments would significantly downgrade the accuracy of position estimation, because RSS information which is measured for the target object could be not enough.
This thesis provides a novel indoor positioning solution to increase the accuracy of location estimation by gathering abundant RSS from the AP which has multi-antennas architecture. Since the AP equipped multi-antennas, the AP gathers RSS from all of antenna sets periodically. This system obtains enough RSS to compute the user position.
Experimental results show that the proposed indoor location estimation demonstrates a high accuracy of positioning and outstanding performance while RSS is collected by smart antenna technique. Our proposed method not only achieves the high-precision positioning but also provides a rational deployments of cost in indoor location environments. The numerical results show that the accuracy is increased from 1.34% to 92.4% and the error variance is reduced by 96% as compared to the traditional location fingerprint method.
|
|---|
