一個不需帶裝置的室內人員偵測方法

Abstract

以無線網路為基礎的定位方法中，大多數的定位法都是讓欲定位的目標上帶著訊號傳送器或是訊號接受器來進行定位。然而，目前依然無令人滿意的方法使目標免持裝置且能達到定位效果。在本論文中，我們提出了一個以統計學的角度的方法來定位無持裝置的目標在哪一個房間，其方法是透過不斷的監測無線網路的訊號變化性來判斷目標在哪一個房間。這個方法分為訓練階段與定位階段，我們假設在環境中存在若干的無線網路基地台與一無線網路訊號接收器，在訓練階段我們將收集目標在各房間時各無線網路基地台對無線網路訊號接收器的訊號分佈狀態；而在定位階段時，我們將利用即時的各無線網路基地台對無線網路訊號接收器的訊號分佈狀態與訓練階段中的狀態做比較，進而找出定位目標所在的位置。

Among all the wireless localization techniques, most works require users to carry a device, such as a transmitter or a receiver, on the target object to be localized. The need is to develop a device-free localization system, This paper proposes a statistics-based scheme to locate people, where we monitor the received signal strength (RSS) of the Radio Frequency (RF) system in the background environment to recognize that a person may be located in a certain room. This scheme consists of two phases: training phase and online phase. In the training phase, we assume that the building has been deployed with some Wi-Fi APs and a fix receiver; our scheme will measure the RSS distribution while a person is in a room, or not in a room, as the training patterns. In the online phase, we compare the current RSS distribution against the training patterns to detect which room a person is now in. We believe that our framework can provide a valuable solution for device-free localization. A prototype system is developed to verify the practicability of our framework with real data.