Abstract
There exist several applications of wireless sensor networks in which the reliable operation can be crucial. Fault prediction is a critical problem in reliability theory for ribbon wireless sensor networks (RWSNs). Accurate fault prediction can effectively improve the availability of the WSNs system. In this paper, we evaluated the network performance for RWSNs, studied the basic theory of kernel functions, proposed a new failure prediction method based on kernel function, and selected the radial basis function as kernel function failure prediction models from two aspects of node hardware failures and network failures for fault prediction. Theoretical evidence and experimental results have shown that the proposed algorithmic prediction method has higher accuracy of 12 and 15% than that of GRNN and PNN respectively. Finally, we provided extensive numerical results to demonstrate the usage and efficiency of the proposed algorithms and complement our theoretical analysis.
Published Version
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