Channel Propagation Model Identification for Spectrum Database: A Spark Based PVOS-ELM

Spectrum database plays an increasingly important role in spectrum measurements and monitoring, which lays foundations for accurate and real-time spectrum sensing in future cognitive radio networks. A successful identification of the channel propagation model is of great importance to construct spectrum database so as to give an accurate picture of spectrum use in real-world environments. In this paper, based on the principle behind the voting-based online sequential extreme learning machine (VOS-ELM) and the Spark cloud computing platform, a novel parallel VOS-ELM (PVOS-ELM) algorithm will be proposed and implemented for real-time channel model identification in real-world propagation environment. The power measurement, kurtosis and skewness etc. will be used as features which are extracted from the received signal. Furthermore, the novel data parallel and task parallel processing schemes will be proposed to improve the computation efficiency of the proposed algorithm on Spark cloud computing platform. Extensive simulations and experiments with real-world data samples will be carried out. The experimental results illustrate that the proposed Spark based PVOS-ELM algorithm enjoys a significant accuracy performance improvement in channel identification and a much higher computation efficiency.