適用於室內環境之射線追蹤法輔助無線定位

Ray Tracing Assisted Location Estimation for Indoor Environment

王才茂(Wang-Tsai-Mao)

指導教授 ： 曾柏軒

國立臺北科技大學/電資學院/電腦與通訊研究所/碩士(2014年)

https://doi.org/10.6841/NTUT.2014.00241

摘要

無線通訊技術一直都是通訊研究很重要的一個課題，其中的無線定位技術更
是近幾年的熱門研究內容。在無線定位技術中以訊號強度(RSS) 為主流定位方
式，其原因為訊號強度資訊容易獲得。近年科技發展迅速，在OFDM 系統與寬
頻系統的進步，我們可以獲得準確的CIR 與RSS，所以有很多課題在研究利用
CIR 來作無線定位。然而，大多數的研究課題謹研究CIR 定位效果，沒有提供
良好的CIR 模型，或實際量測數據內容。本研究利用射線追蹤法來建立模擬的
CIR 模型系統，並且利用儀器進行一系列的通道測量，以獲得實際數據。在研究
中發現以射線追蹤法搭配實際量測數據所進行的無線定位，稱為基於射線追蹤法
的資料庫特徵比對定法(ray-tracing based fingerprinting，RBFP)，比單一使用量
測數據，稱為特徵定位法(Fingerprinting，FP) 的準確度還要提升20%。並且本研究還發現當二者結合再配合特定權重後，我們稱為射線追蹤法輔助的特徵比對
定位法(ray-tracing assisted fingerprinting，RAFP) 可以提高定位效果46%，可以使定位在平均67% 位置誤差範圍為1.8 公尺。

關鍵字

訊號強度 ； 時域通道脈衝響應 ； 頻域通道頻率響應

並列摘要

Wireless communication technique has always been an important theme in communication research field. Moreover, wireless positioning technique is the most popular research topic in recent years. Received signal strength (RSS) is the mainstream in wireless positioning technique since the information of signal strength can be received easily. With the rapidly development of OFDM and wireless broadband system, more accurate channel impulse response (CIR) and RSS can be obtained. Therefore, there are many topics researching how to use CIR for wireless indoor positioning. However, most of the researches only concentrate on the effects of CIR positioning without providing good CIR mathematical model or real CIR measurement data. This research utilizes ray-tracing technique (RT) to create a simulated CIR model and performs a series of channel measurement experiments using OFDM-based sounding system to obtain measurement data. In this study,we propose a ray-tracing based fingerprinting (RBFP) based on the comparison
between the online measurement and the simulated CIR database using RT.The error of the proposed RBFP is validated in the experiment that is 20% less than that of the traditional fingerprinting (FP). In addition, the ray-tracing assisted fingerprinting (RAFP), which is proposed by combining the RBFP and the FP in specific weights, can increase 46% of positioning accuracy and the root mean square error of the RAFP is 1.8 meters less than the FP.

並列關鍵字

received signal strength ； channel impulse response ； channel frequency response, indoor location estimation

參考文獻

[15] 莊凱棋“時間平滑的轉換子空間旋轉法應用於移動中同調訊號方向及信號估測,” in 中原大

[10] 葉大千“數種室內環境中多輸入多輸出型態之通道特性與傳輸效能,” in 台灣大學電信工程

of Indoor MIMO-UWB Transmission for Different Material Partitions”Tamkang Journal

[24] 許正德“IEEE 802.16 系統之通道估測及等化器設計與實現”, 碩士論文. 中央大學通訊工

[11] 詹曜嘉“相關性多輸入輸出通道之解構與分析,” in 台灣大學電信工程研究所, 2011.