Font Size: a A A

Research On Wireless Sensing Technology Of Wireless Sensor Networks

Posted on:2016-11-01Degree:MasterType:Thesis
Country:ChinaCandidate:L YangFull Text:PDF
GTID:2348330488974184Subject:Engineering
Abstract/Summary:PDF Full Text Request
Wireless sensing technology that is context-aware technology and one of the key technologies of wireless sensor networks, operate the contextual information captured by numbers of sensors and then enable the system to be self-decisions based on the information. However, due to the interference from deployment, limited energy, environmental noise, internal manufacturing defects or other factors, wireless sensor network node prone to failure or abnormal readings.Wireless body area networks collect physiological signals or signals of human motion through numbers of biological sensors attached on the human body or embedded inside the body, sometimes errors may occur during the data collected by the sensor and transmission because of the interference, in order to ensure the accuracy of the data,we should detect the abnormal data prior to application of wireless sensing technology.In this thesis, we analyzed the characteristics of the physiological data collected by the wireless body area network, such as the data of heart rate, and verified the Markov property of the heart rate data, and introduced the hidden Markov model to detect abnormal data in medical data and achieved a high accuracy. Using hidden Markov model to detect abnormal data in medical data, it not only avoids the heavy task of collecting historical data, but also optimizes the parameters in hidden Markov model when the training data contains errors, and then let us to take advantage of Viterbi algorithm detection of abnormal data.In this thesis, we also analyze the traditional outlier detection algorithms, and propose a method combining Mahalanobis Distance and kernel density estimation to detect abnormal number of heart rate data. Results comparing two experiments show hidden Markov model used in medical data anomaly data with high accuracy. Finally, we conducted a distance aware experiment using the smart phones Bluetooth communication module for wireless sensing technology. In the absence of circumstances outside the network, Bluetooth communication module can be self-organizing network via radio technology into Piconet and communicate with each other. The distance-aware technology of Bluetooth devices can provide its products with wider application range.
Keywords/Search Tags:wireless sensor network, wireless body area network, Outlier data, wireless sensing technology, hidden Markov model, Bluetooth communications, distance-aware technology
PDF Full Text Request
Related items