| 摘要: | 過去幾年，物聯網(Internet of Things, IoT)吸引了來自各個領域的關注。在物聯網設備不斷監控環境屬性並提供可遠程控制功能的同時，物聯網設備的地理空間位置對許多提供空間分析的應用是個必要的資訊。然而，儘管全球定位系統(Global Navigation Satellite System, GNSS)是常用的定位技術，在每個物聯網設備中增加功耗和GNSS接收器的成本並不符合多數物聯網設備的需求。因此，為了不大幅增加功耗及硬體成本，本研究的想法是利用物聯網裝置的無線通訊協定進行定位。確切而言，本文提出了一種使用LoRa(Long Range)的本地無線通信協議的定位系統的設計和實現。LoRa是近年提出的一種低功耗無線通信技術，主要用於物聯網設備之間傳輸信息。使用者可以通過以下幾種位置相關參數(location dependent parameter, LDP)來估計設備的位置，包括接收信號的強度(received signal strength indicator, RSSI)、信號傳播的持續時間（time of arrival, TOA）、訊號到達閘道器之時間差(time difference of arrival, TDOA)、和到達角度(angle of arrival, AOA)。由於近期研究顯示TDOA比RSSI可達更好的精度，且TOA需要較高的硬體需求以提供信號傳輸的開始時間，AOA需要高敏天線等硬體需求，所以這項研究旨在檢驗TDOA位置算法在LoRa上的適用性，但此套算法不侷限於LoRa。此外，為了進一步提高定位精度，本研究通過設置參考節點，應用差分定位概念補償定位誤差，此方法命名為輕量化的差分TDOA (Light Differential TDOA, Light-DTDOA)。在設有九個未知節點和三個閘道器其場域大小為四平方公里的實驗中，結果顯示在使用差分修正前的誤差達到1.35264E+14公尺且提出的Light-DTDOA可提升精度至定位均方根誤差約44米。;Internet of Things (IoT) has been attracting attention from various domains over the last few years. While IoT devices continuously monitor environmental properties and provide remotely controllable functions, the geospatial location of IoT devices is an expensive information to obtain. Whereas Global Navigation Satellite System (GNSS) is a commonly-used positioning technology, large power-consumption and excessive cost of adding a GNSS receiver in every IoT device is not realistic. Therefore, this paper presents a design and implementation of a positioning system using a local wireless communication protocol called LoRaWAN. LoRaWAN is a low-power, wireless communication technology recently proposed to transmit information between IoT devices. A device’s position can be estimated by monitoring a location dependent parameter (LDP) such as received signal strength indicator (RSSI) measuring received signal strength, time of arrival (TOA) measuring the time duration of signal travel, angle of arrival (AOA) measuring the angle of receive signal, and time difference of arrival (TDOA) measuring the TOA differences between gateways. While recent studies show that TDOA can achieve a better accuracy than RSSI and TOA requires a signal’s start time information, AOA need high sensitivity antenna, this research aims at examining the applicability of TDOA position algorithm on LoRaWAN. To further improve positioning accuracy, we apply the concept of differential correction by setting up a reference node to compensate local positioning error. This method is named light differential TDOA (Light-DTDOA). In the experiment with nine nodes and three gateways, the results show that the positioning error is 1.35264E+14 meters without applying differential corrections and the proposed Light-DTDOA can improve accuracy to around 44 meters. |
