Leader selection in cooperative network based on MDL subspace algorithm for cognitive radio

This paper presents a simple and robust algorithm for determining a leader node in a cooperative network based on MDL (Minimum Description Length) subspace algorithm. The algorithm aims to improve the performance of the cooperating network in a spectrum sensing problem for cognitive radio. The outline of the communication and selection process is described and the SNR (signal to noise ratio) estimation algorithm is given. Simulation results show that the algorithm outperforms the non-cooperating network and a cooperating diffusion network with uniform combination weights.