接收訊號到達時間為符合比對項目以增進微軟室內定位方法之模擬研究

Abstract

在室內定位中，最常被使用的特性項目有兩項，分別為接收訊號強度以及接收訊號所到達時間。以現有的無線架構及設備，接收訊號強度可以輕易被量測得到，但也非常容易受周遭環境變動的影響，改變量測所得之值，因此接收訊號強度僅能提供低解析度的定位方法。相對來說，接收訊號到達時間為所有特性項目中最準確的，但是假若直線訊號接收路徑上有阻擋物，則以接收訊號到達時間量測距離會造成嚴重的錯誤。在本論文中，我們討論在同時使用接收訊號強度及接收訊號到達時間的情況下，如何增進兩者的優點。我們認為，在小範圍區域中，僅以接收訊號到達時間定位可以取得不錯的準確度，而僅以接收訊號強度定位，在相對小的信任區間內可達到相當的顯著性，因此我們建議先以接收訊號強度定位，縮小可能的區域範圍，再使用最多符合數目的技巧，以接收訊號到達時間為比對項目，搜尋此縮小後區域。模擬結果顯示，此方法可以明顯提高準確度，此外，我們也將討論各種相關參數對準確度的影響。

In indoor geolocation, the received signal strength (RSS) and Time of arrival (TOA) are two most common metrics to be used. RSS is easily obtained from the existing wireless infrastructure, but it is strongly influenced by the surrounding changes and thus results in a low resolution. TOA, on the other hand, is more accurate than other metrics but exhibits larger error than simply using RSS if the direct line of sight (DLOS) is blocked. In this thesis, we study how to boost the advantages of RSS and TOA when using them together. In viewing of that simply using TOA in a small area usually results in good accuracy and simply using RSS with relatively small confidence area will gain a fairly high level of significance, we propose to use the RSS first to locate the small area and then apply a technique called the longest matching to search the area with the TOA patterns. The simulation results provide an apparent improvement and we discuss the effect of tuning the design parameters.