Abstract
The link quality of wireless sensor networks is the basis for selecting communication links in routing protocols. Effective link quality estimation is helpful to select high-quality links for communication and to improve network stability. The correlation of link quality parameter and packet reception rate (PRR) is calculated by the Pearson correlation coefficient. According to Pearson coefficient values, the averages of the link quality indication, received signal strength indication, and signal-to-noise are selected as the parameters of the link quality. The link quality grade is taken as a metric of the link quality estimation. Particle Swarm Optimization (PSO) algorithm is used to optimize the parameters of the weighted extreme learning machine (WELM), including the number of hidden nodes, weights, and the normalization factor. A link quality estimator (LQE) based on the improved weighted extreme learning machine (LQE-IWELM) is constructed. In different scenarios, experiment results show that the improved weighted extreme learning machine (IWELM) is more effective than extreme learning machine (ELM) and WELM. Compared with the other three link quality estimation models, LQE-IWELM has better precision and G_mean.
Highlights
Wireless Sensor Networks (WSNs) are multi-hop selforganizing networks, which is formed by a large number of cheap micro-sensor nodes deployed in the monitoring area through wireless communication [1]
The Pearson correlation coefficient is used to calculate the correlation degree between link quality hardware parameters and packet reception rate (PRR), which is taken as the input of the link quality estimator; Secondly, we divide link quality grades according to the range of PRR and construct the link quality estimation model based on the weighted extreme learning machine (WELM); Aiming at the optimization of the parameters of WELM, the Particle Swarm Optimization (PSO) algorithm is used to optimize the number of hidden layer nodes and class weight, and the LQE-improved weighted extreme learning machine (IWELM) is proposed
LINK QUALITY ESTIMATION MODEL In this paper, WELM is used to build the link quality estimation model, received signal strength indicator (RSSI), link quality indication (LQI) and SNR are used as the input of the model, and the link quality grade is the output of the WELM
Summary
Wireless Sensor Networks (WSNs) are multi-hop selforganizing networks, which is formed by a large number of cheap micro-sensor nodes deployed in the monitoring area through wireless communication [1]. The improved weighted extreme learning machine (IWELM) algorithm is obtained by optimizing the parameters of WELM with particle swarm optimization. A simple review of related work on link quality estimation. The main contributions of this paper are as follows: (1) Considering the imbalance of the samples after dividing the link quality grades, the link quality estimator based on the weighted extreme learning machine (LQE-WELM) is proposed in this paper. The rest of this paper is structured as follows: In Section II, we briefly describe related studies, those that existing research in link quality estimation problems and extreme learning machine approaches.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
