植基於雙指向天線與Zigbee 之室內定位技術研究

A Study Based on Bidirectional Antenna and Zigbee for Indoor Localization

杜昀儒(Yun-Ru Tu)

指導教授 ： 陳同孝 ； 陳民枝

國立臺中科技大學/資訊學群/資訊工程系碩士班/碩士(2011年)

摘要

無線室內定位技術目前應用廣泛，最常應用於人員與物品的管理，如：大樓保全，醫療院所的病人位置、狀態監控，和學校學生的活動範圍管制及倉儲的貨物位置管理等，為了得知目標物的位置，定位準確度是室內定位技術重要的議題；目前最常見的無線室內定位技術方法為信號接收強度(Received Signal Strength, RSS)，利用發送端與感測端無線訊號的強度搭配演算法進行目標物位置的計算，但此方法容易受到環境與障礙物的干擾引導致準確度較低與，環境如風、電子設備的電波訊號及溫度引響較大，風和電子設備會造成無線訊號強度不穩定或訊號干擾(Noise Interference)，溫度的高低會使RSS 變弱或增強，雖然無線訊號可穿透障礙物，但遇到障礙物時或使訊號衰弱，或產生多重路徑(Multi-Path)，這兩種情況會導致收集到的RSS 為非視線(Non-Line of Sight, NLOS)訊號，NLOS 訊號必須經過演算法的修正才能降低與目標物的誤差。本論文提出三種方式降低環境與障礙物的干擾，一、使用指向性天線(八木型)，具有訊號強不易受干擾和方向性的訊號場型，讓目標位置的訊號更容易判斷；二、最佳訊號功率預測演算法，利用Zigbee 可調整發射功率的優點，遇到不同大小的室內環境可以進行訊號的調整，降低訊號過強產生多重路徑或過弱無法接收訊號的問題；三、定位演算法，利用回歸函數再次修正訊號雜訊，並以簡易的三角函數達成定位的目標。為了能夠證明本研究提出之功率預測法能在不同環境下定位，本研究選擇走廊與教是兩
種環境進行實驗。
實驗結果分兩部分，第一部分為功率選擇於不同環境的比較，結果顯示所提出之預測功率的定位準確度較應用其它功率高，證明本研究提出之最佳訊號功率預測演算法能提供不同環境最佳的功率選擇；第二部分為本研究實驗結果誤差平均為51.1cm，實驗總共使用兩個感測點，與近年學者之無線室內定位結果較佳。因此本研究所提出之無線室內定位技術有提升定位的準確度、減少設備成本及應用於不同室內環境中的優點。

關鍵字

信號接收強度 ； 非視線 ； 多重路徑 ； 指向性天線(八木型) ； Zigbee

並列摘要

Wireless indoor localization technology has many applications. The indoor localization technique has been used in the management of personnel and goods, such as, building security, patient location and status monitoring, student activities control, storage of cargo locations and management, and more. In order to know the target location, positioning accuracy is an important issue of indoor localization technology. The most method wireless indoor localization used is received signal strength (RSS) where the RSS is between the sender and sensor side. In the case of environmental interference and obstructions, there will be a low accuracy problem. Also, environmental interferences like wind, electronic equipment, radio signals and high temperature, will cause instability in the wireless signal strength or signal noise interference. The temperature will weaken or strengthen RSS. Although the wireless signal can penetrate obstacles, the encountered obstacles or weak signals or Multi-Paths collected in both cases can lead to a non-line of sight (NLOS) to the RSS. The NLOS signal correction algorithm must be used to reduce target error. This paper presents three ways to reduce interferences of the environment and obstacles, namely, (1) using the directional antenna (Yagi type), which has strong signal less susceptible to interference and directional signal field type and the target location of the signal is easier to judge; (2) the best signal power prediction algorithm by using Zigbee which has advantage of the adjustable emissive power where the emissive power can be adjusted for different indoor space to reduce the weak/strong signal problem; (3) positioning algorithms by applying the regression analysis function correction signal noise, and simple trigonometric functions to achieve targeted goals.

The experiments are conducted for two indoor locations, a classroom and a corridor. The results were divided into two parts. In the first part, adjustable emissive power is applied in two places. Results showed that the forecast made by the power of positioning has higher accuracy than other high power applications. Two signal power prediction algorithms were used to provide the optimal power options in different environment. In the second part, experimental results showed that the average error of 51.1cm. Two sensor points were used and the results compared with works of other researchers in the same field. Results showed that the proposed method has higher accuracy and lower cost of equipment. In this study, the proposed wireless indoor location technique improved the overall positioning accuracy, reduce sensor points requirements and the emissive power is adjustable for used in different indoor environments.

並列關鍵字

RSS ； NLOS ； Multi-Path ； Yagi ； Zigbee

參考文獻

[2] 劉明山，植基於ZigBee與SVM之室內絕對定位演算法，國立台中技術學院資訊科技與應用研究所碩士論文，2009。

[5] A. Narzullaev, Y. Park, K. Yoo and K. Yu, “A fast and accurate calibration algorithm for real-time locating systems based on the received signal strength indication,” International Journal of Electronics and Communications(AEU), Vol. 65, 2011, pp.305-311.

[6] A.N. Khan, X.J. Sha and X.Y. Zhang, “Improving handoff call handling in future mobile networks with location tracking,” Information Technology Journal, Vol. 7, 2008, pp. 936-941.

[8] E.S. Nadimi, H.T. Søgaard, and T. Bak, “ZigBee-Based Wireless Sensor Networks for Classifying the Behaviour of a Herd of Animals Using Classification Trees,” Biosystems Engineering, Vol. 100, No. 2, June 2008, pp. 167-176.

[9] F. Fang,, M. Xudong and D. Xianzhong, “An Improved Particle Localization Algorithm for Mobile Robot in Indoor Environment,” Information Technology Journal, Vol.10, 2011, pp. 1017-1023.