植基於ZigBee與SVM之室內絕對定位演算法

A Study of Indoor Localization Based on ZigBee Wireless Sensor Networks and Support Vector Machine

劉明山(Ming-Shian Liou)

指導教授 ： 陳同孝

共同指導教授 ： 陳民枝

國立臺中科技大學/資訊學群/資訊科技與應用研究所/碩士(2009年)

https://doi.org/10.6826/NUTC.2009.00056

摘要

摘 要
近年來，在室外定位上主要是以全球定位系統(GPS, Global Positioning System)為主，但是GPS會受到建築物的遮蔽(Shadowing Effect)及複雜地形的影響（例如：叢林或洞窟），並不適用於室內定位。以紅外線、超音波、超寬頻作為室內定位時，則易受到範圍的限制，在強光及多人的環境下定位準確度不高，且設備成本昂貴，造成應用上的瓶頸；以802.11a/b/g、RFID及ZigBee等無線區域網路技術進行室內定位時，則易受無線電信號反射、折射、繞射、散射及多重路徑的特性影響，造成信號不穩定的現象，若要克服這種特性，其演算過程就較為繁複，因此截至目前為止，並沒有一個完善的室內定位系統存在。本文提出離線、連線兩階段之定位演算流程，僅使用少量的節點、計算過程簡單、減少資料分析時的耗用時間、不需要設定太多的參數、兼具高速度與準確度、又易實作的室內定位系統。在離線階段，利用SVM進行樣本資料的分析，以評估網格區塊的大小，再以高斯機率分佈，分析RSS信號出現之機率及信號強度的分佈情況，作為定位時的依據；在連線階段，則以離線階段之機率密度函數，輔以無線電地圖，與追蹤目標進行比對，進而判定追蹤目標的所在位置。由實證結果顯示，本研究與802.11b及RFID之定位系統比較，精確度可達1.2 m；與其他的ZigBee定位系統比較，於2.4 m × 2.4m的網格內，其準確率達到100.00%；前述結果均較其他的定位系統為優。

關鍵字

接收信號強度 ； ZigBee ； RADAR ； 定位 ； SVM

並列摘要

ABSTRACT
In recent years, outdoor localization is based mainly on the Global Positioning System (GPS). However, GPS is affected by shadowing effects and complicated terrains (such as, jungles or tunnels); and, is not appropriate for indoor localization. The infrared system, ultrasound and ultra width bands have limited range for indoor localization. Positioning accuracy is poor where there are strong lights and crowdedness. Due to expensive setup costs, radio applications have bottlenecked. The 802.11a/b/g, RFID, and ZigBee wireless local network technology during indoor positioning are affected by wireless signal reflection, refraction, diffraction, scattering and multipath interferences which results in signal instability. Complicated algorithms are needed to resolve this problem. To date, there has not been a comprehensive indoor localization system available yet. In this thesis, two stage localization processes were proposed which required fewer positioning nodes. The computational process simple; computation time was significantly decreased during analysis. Few parameters were needed in the algorithm since speed and accuracy were the ultimate goals. The proposed positioning system was easy to setup. When off-line, the support vector machine (SVM) was used to conduct positioning assessment. The Gaussian probability distribution was used to analyze the probability of RSS signals. When online, the off-line stage of the probability density functions was used to assist radio map and to match the location of the tracing goal to decide on the location of points. Comparisons were made with 802.11b and RFID localization system accuracies could be achieved up to 1.2m. Comparisons with others ZigBee localization system showed that accuracies could be achieved up to 100.00%. The results significantly outperformed other positioning systems.

並列關鍵字

receive signal strength ； ZigBee ； radar ； positioning ； SVM

參考文獻

[1] A. Chehri, P. Fortier, and P. M. Tardif, “UWB-Based Sensor Networks for Localization in Mining Environments,” Ad Hoc Networks, Vol. 7, No.5, July 2009, pp.987-1000.

[2] A. H. Sayed, A. Tarighat, and N. Khajehnouri, “Network-Based Wireless Location: Challenges Faced in Developing Techniques for Accurate Wireless Location Information,” IEEE Signal Processing Magazine, Vol. 22, No. 4, July 2005, pp. 24-40.

[4] A. Ward, A. Jones, and A. Hopper, “A New Location Technique for the Active Office,” IEEE Personal Communication, Vol. 4, No. 5, October 1997, pp.42-47.

[5] C. C. Chang, and C. J. Lin, “LIBSVM: a Library for Support Vector Machines,” Software Available at http://www.csie.ntu.edu.tw/~cjlin/libsvm, 2008.

[7] D. Fox, J. Hightower, L. Liao, D. Schulz, and G. Borriello, “Bayesian Filtering for Location Estimation,” IEEE Pervasive Computing, Vol. 2, No. 3, July 2003, pp. 24-33.