在廣泛的佈建無線設備下，藉由日漸普及手機或感測器促進了許多關於定位系統上的成功。本篇論文的核心是在802.11無線區域網路下，針對非攜帶式定位提出了一個特別的系統。關於非攜帶式系統主要是在強調如何去偵測人或物體的位置即使是身上沒有攜帶任何無線裝置[1]。提出的系統是藉由觀察從無線網路基地台（ＡＰ）發送訊號強度源（ＲＳＳ）的定位研究。ＲＳＳ訊號強度源將會被收集以及用於定位被追蹤物體的位置。為了判斷物體的位置，我們引用了整合性分類器，包含著名SVM和機率Bayesian分類器。除此之外，我們也提出了時間延遲器架構，來降低誤差跟錯誤藉此提高正確率。並且在平均約1.3 公尺距離內和10延遲預測步，我們利用四台ＡＰ提出的系統可以達到非常高的準確率：90.6%，而且當我們再拉高到20延遲預測步，分類準確率更進一次的達到94.3%。本篇提出的論文達到了令人訝異的成果: 高成功率、快速，便宜，可操作性應用在實際的無線區域網路環境中。而且在不需要額外的硬體設施也指出我們的系統能夠更有效率的利用於在現有廣大的設備之中。Wireless networks are ubiquitous now. It promotes the subject of location technology by the cell-phone or sensor. This paper proposes a special system for device-free localization over IEEE 802.11 wireless local area network (WLAN).This area of research emphasizes that a person does not need to carry a wireless device to be detected and located.Our system is to analyze the received signal strength (RSS) transmitted from access points (APs). RSS signals will be collected and used for locating the tracked subject. To identify the location of the subject, this work adopts an ensemble of classifiers, which contains the support vector machine (SVM) and the Bayesian classifier. Moreover, this work incorporates the classifiers with the tapped delay line (TDL) architecture to verify the decision made by the classifiers at different time unit. Within 1.3 meters distance error and 10 taps, the proposed framework achieves a high precision rate of 90.6% using four access points. The precision rate further increases of 94.3% by considering 20 taps.The characteristics of the proposed system are highly successful, fast, cheap, and can be applied to any established WLAN environment now. Therefore, by inexpensive equipment, it also indicates our system can be widely used for many real applications.
