Link Quality Estimator Based On Uncertain Multi-attribute Decision Making Theory

Nodes rely heavily on link quality to select communication links in the wireless sensor networks(WSNs).Because of the resource restriction of the sensor nodes and the complexity of the deployment environment,the communication links between nodes are unstable.The effective link quality estimation can reflect the real link state,and avoid the unnecessary energy consumption caused by routing reselection and data retransmission.Therefore,a good link quality estimator is crucial for improving data transmission efficiency,saving node energy consumption and enhancing network stability.The program is supported by the National Natural Science Foundation aiming at research on WSNs link quality estimation.After analyzed the wireless sensor networks,the link quality estimation is abstracted as a multiple attribute decision process.Link transmission efficiency,link volatility and received signal strength are selected as the estimation attributes of link quality.Packet Received Rate(PRR),mean square deviation of signal noise rate(SNR)and average of received signal strength indication(RSSI)are selected to quantify attributes respectively.Considering the complexity of deployment environment and inconsistency of link quality standards,link quality estimator based on uncertain multi attribute decision making(MADMLQE)is proposed.During the period of detection cycles the attribute is fuzzed by Horizontal approach.The improved entropy method is adopted to determine the weight of each attribute.The improved TOPSIS is employed to achieve the relative closeness degree,which is used to estimate link quality.PRR and SNR are chosen to evaluate the stability and agility of the proposed link quality estimator,MADMLQE.A wireless sensor networks is deployed in the indoor corridor.The links with different quality are obtained by setting the distance and transmitting power of nodes.The link quality parameters of each link are collected through the WSNs link quality testbed,which is applied to verify performance of the estimator,MADMLQE.Compared with the existing 7 link quality estimators,MADMLQE has better reliability,and it can estimate link quality more objectively.Furthermore,MADMLQE has a better trade-off between stability and agility.