Cognitive radio-enabled distributed cross-layer optimization via genetic algorithms

In this paper, we propose two optimization approaches based on genetic algorithms in addition to several assisting mechanisms for making these approaches more efficient and robust in actual network implementations. Using the fact that a cognitive radio is capable of sensing the prevailing environmental conditions and automatically adapting its operating parameters in order to enhance system (and potentially network) performance, the proposed approaches will optimize individual wireless devices using partial operating parameter and environmental information from adjacent devices within the network. Assuming stationary wireless nodes, with all wireless links employing non-contiguous orthogonal frequency division multiplexing (NCOFDM) in order to enable dynamic spectrum access (DSA), the proposed approach will attempt to simultaneously minimize the bit error rate, minimize out-of-band (OOB) interference, and maximize overall throughput using a multi-objective fitness function. Simulation results show that the proposed procedure is able to reduce BER by one order of magnitude.