架構於無線人體感測網路下心跳同步方法之時間誤差改良機制

機構典藏 > College of Engineering > Graduate Institute & Department of Electrical Engineering > Thesis > Item 987654321/111440

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/111440

Title:	架構於無線人體感測網路下心跳同步方法之時間誤差改良機制
Other Titles:	A novel time synchronization scheme to improve the heart-beat driven MAC in wireless body sensor network
Authors:	王瑋;Wang, Wei
Contributors:	淡江大學電機工程學系碩士在職專班吳庭育;Wu, Tin-Yu
Keywords:	分時多重存取協定;氾濫式時間同步協定;TDMA;Heartbeat-Driven MAC;FTSP
Date:	2016
Issue Date:	2017-08-24 23:52:55 (UTC+8)
Abstract:	本研究是針對人體無線區域網路的通訊協定中，基於分時多重存取(TDMA)協定架構下，利用心跳來作為時間同步的Heartbeat-Driven MAC(H-MAC)來進行改良。H-MAC的設計應用於星狀拓樸的人體無線感測網路，利用人體本身固有的心跳脈搏之頻率用以取代原本使用外部無線電定期接收時間同步訊號來實現時間同步，如此可以節省放置在人體身上感測器的能源消耗，並降低感測器發送封包所產生電磁輻射對人體造成傷害。雖然利用心跳頻率的信息嵌入在不同的生物信號，透過檢測波形峰值感知數據可以實現人體感測網路(Body Sensor Networks, BSN)時間同步，不過由於脈搏都是利用心臟收縮與舒張擠壓血液所產生的跳動，雖然身體各處所測得脈搏跳動次數皆相同，但因為血管擴於身體各處，心臟收縮傳送到身體各處血管的距離也會不同，因此也會造成身體各處所接收到同一次心臟收縮的時間有些微的時間差，而造成同步的問題。在分析完H-MAC，我們對於利用心跳同步可能產生的問題，加入了無線感測網路時間同步協定中，發展最為成熟的氾濫式時間同步協定(Flooding Time Synchronization Protocol, FTSP)用以修正時間同步造成誤差的問題，藉由將時間郵戳的記錄提早到MAC層來減低計算誤差，並且利用線性迴歸(Linear Regression)作為預測補償時間，用以降低感測器之間的同步誤差，計算出感測器與資料收集器之間的時間差用以估計感測器放置人體身上的距離位置，透過我們的改良H-MAC+，可以使得人體各部位間的感測器在開始進行時間同步時，可以先透過FTSP協定來讓感測器準確的達到時間同步，接著再交給藉由心跳及脈搏進行時間同步，用以減低感測器在同步時的能源消耗，當感測器之間因為心跳同步發生問題而造成誤差過大時，此研究會透過本研究所提出的平均誤差判斷機制自動進行時間誤差修正，減少資料傳輸時所發生封包遺失的問題，由於大部分工作時間還是藉由心跳及脈搏進行時間同步，所以我們提出的研究方法更可以降低電磁波對人體的傷害。 The purpose of this paper is to improve the TDMA-based Heartbeat-Driven MAC (H-MAC) protocol for WBAN. Designed for the star topology in WBAN, H-MAC uses heartbeat rhythm information, instead of periodic synchronization beacons, to attain time synchronization. In this way, H-MAC can reduce the power consumption of WBAN sensor nodes as well as the electromagnetic radiation that could be harmful to human health when sensors send packages. By embedding heartbeat rhythm information into different biosignals, BSN synchronization can be achieved according to the peak values of wave forms. Pulse means the number of times one''s heart beat per minute because of systole and diastole. Although the number of pulses measured at different body parts will be the same, blood vessels run over the whole body and their distances away from the heart could greatly vary during the same systole, resulting in a slight time difference and the problem of synchronization. According to our analysis of H-MAC, our proposed approach therefore includes the Flooding Time Synchronization Protocol (FTSP), currently the most mature method for time synchronization, to handle time differences. Our proposed approach advances the timestamp to the MAC layer for error reduction, uses the linear regression to predict the time compensation for synchronization error reduction among sensors, and computes the time differences between each sensor and the coordinator to estimate the sensor placement on human body. With our improved H-MAC+, sensors on different body parts can be synchronized first via FTSP and next accurately synchronized with heartbeat and pulse to reduce the power consumption duringsynchronization among sensors. Our proposed approach automatically performs time error adjustment to reduce data loss during transmission when the error among sensors greatly increases because of heartbeat. Since most of the time we use heartbeat rhythm for time synchronization, so it can be reduce the SAR for body parts to decrease the negative effects of electromagnetic radiation on human body.
Appears in Collections:	[Graduate Institute & Department of Electrical Engineering] Thesis

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	137	View/Open

Loading...