Research On Obstacle-aware Link Quality Estimation In Wireless Sensor Networks

Link quality estimation is a fundamental building block for communication protocols in Wireless Sensor Networks(WSNs). However, the link quality of low-power communications is susceptible to environmental changes, such as the temperature, humidity, noise, and especially mobile obstacles. As a result, it is difficult for the existing link quality estimation methods to yield accurate prediction at a low cost.In order to improve the estimation performance of link quality while the wireless connection is blocked by obstacles, we try to use sensor data to quickly detect environmental changes and assist link quality estimation in this thesis. First, we studied the impact of obstacles on the link quality, and discussed the temporal and spatial correlation between the obstacles detection sensor data and link quality changes. After that, an obstacle-aware link quality estimation algorithm is proposed, in which we designed an online-learning algorithm for establishing a mapping function from sensor data to link quality, and used the mapping function to predict link quality changes according to current sensor readings. At last, in order to alleviate the limited sensor detection range issue, we designed a cooperation algorithm among nodes to let one sensor node use sensor data from neighboring nodes in link quality estimation. As such, the sensor detection range can be extended.Experimental results show that the proposed algorithm performs better than the existing Four-bit algorithm in response time, accuracy and communication overhead. At the same time, the proposed method incurs modest storage overhead, and is applicable to WSN platforms.