【摘要】：近年來,基于Wi-Fi接收信號強(qiáng)度的室內(nèi)定位技術(shù)一直是研究領(lǐng)域的熱點(diǎn)問題.隨著智能家居和可穿戴計(jì)算的高速發(fā)展,大量新型智能設(shè)備的出現(xiàn)進(jìn)一步推動了室內(nèi)定位技術(shù)的發(fā)展,同時(shí)也帶來了新的挑戰(zhàn).可穿戴設(shè)備與傳統(tǒng)智能設(shè)備相比,有著與用戶更加緊密的位置綁定關(guān)系,是一類更加適合的室內(nèi)定位平臺.但另一方面,類似智能手表、眼鏡、手環(huán)以及戒指等可穿戴設(shè)備,由于其自身資源受限的特性,迫切需要一種低功耗的新型室內(nèi)定位算法.通過本文在Moto 360二代智能手表上進(jìn)行的實(shí)驗(yàn)可以發(fā)現(xiàn),基于Wi-Fi RSS的室內(nèi)定位服務(wù)會使智能手表的使用時(shí)間縮短82%以上,其中99%的定位能耗,是為了保證定位精度而大量進(jìn)行射頻信號采集所造成的.簡單的減少信號采集量將帶來顯著的定位精度下降,如何在保障定位精度的前提下,盡可能地減少信號采集量是低功耗定位技術(shù)面臨的核心挑戰(zhàn).該文提出了一種面向能耗受限的可穿戴設(shè)備的室內(nèi)定位技術(shù)TinyLoc.TinyLoc在實(shí)時(shí)定位階段僅需要一次信號采集,同時(shí)運(yùn)用用戶運(yùn)動特性彌補(bǔ)信號采集量減少而帶來的精度缺失.實(shí)驗(yàn)結(jié)果表明,在90%的情況下,TinyLoc對于完整路徑上的點(diǎn)平均誤差可以達(dá)到3m以內(nèi),另一方面,在相同實(shí)驗(yàn)環(huán)境下,TinyLoc能耗為傳統(tǒng)Wi-Fi定位算法的1/6,是MoLoc的64%.相比傳統(tǒng)的基于Wi-Fi信號的定位算法,TinyLoc可以延長Moto 360二代智能手表約3倍的定位工作時(shí)間.
[Abstract]:In recent years, indoor positioning technology based on Wi-Fi received signal strength has been a hot issue in the field of research. With the rapid development of smart home and wearable computing, the emergence of a large number of new intelligent devices has further promoted the development of indoor positioning technology, but also brought new challenges. Compared with traditional intelligent devices, wearable devices have closer location binding relationship with users, and it is a more suitable indoor positioning platform. On the other hand, wearable devices such as smartwatches, glasses, bracelets and rings are in urgent need of a new low-power indoor positioning algorithm because of their limited resources. Through the experiment on the second generation smartwatch of Moto 360, it can be found that the indoor positioning service based on Wi-Fi RSS can shorten the service time of smartwatch by more than 82%, of which 99% of the positioning energy consumption is caused by a large number of RF signal acquisition in order to ensure the positioning accuracy. A simple reduction of signal acquisition will lead to a significant decline in positioning accuracy. How to reduce signal acquisition as much as possible under the premise of ensuring positioning accuracy is the core challenge of low-power positioning technology. In this paper, an indoor positioning technology for wearable devices with limited energy consumption is proposed. TinyLoc.TinyLoc needs only one signal acquisition in the real-time positioning phase, and makes use of user motion characteristics to make up for the lack of accuracy caused by the reduction of signal acquisition. The experimental results show that the average error of TinyLoc for the point on the complete path can reach less than 3m in 90% case. On the other hand, in the same experimental environment, the energy consumption of TinyLoc is 1 鈮，

本文編號：2509495