WSNs Link Quality Evaluation Method Based on Multi-Hidden Layer Multi-Channel Extreme Learning Machine

In this paper, we analyze the link characteristics of WSNs (wireless sensor network) and the existing link quality evaluation methods. Considering the link quality assessment affected by susceptibility to multi-path fading and low energy of the link, and evaluates the link quality based on hardware parameters. The received signal strength indicator, link quality indicator and signal-to-noise ratio are selected as the comprehensive evaluation parameters of the link quality, the link quality level is divided according to the packet reception rate, and the link quality level is used as the link quality evaluation index. On this basis, a multi-hidden-layer and multichannel extreme learning machine is used to evaluate the link quality of WSNs, and a link quality evaluation model is constructed.