Design, implementation, and assessment of predictive cognitive channel access for industrial communication in the 2.4 GHz ISM band

We propose a predictive cognitive channel access mechanism that addresses issues experienced by current medium access schemes such as TDMA and CSMA such as collisions and spectrum inefficiencies. The proposed algorithm incorporates the prediction of whitespaces, i.e. spectral resources to eliminate the random exponential backoff used in CSMA/CA to increase the spectral efficiency. Prediction in the proposed algorithm is done using Artificial Neural Networks. The proposed method clearly outperforms the existing technologies in terms of probabilitity of collision and spectral efficiency pushing in the direction of wireless for industrial applications.