Experimental study of genetic algorithm based link adaptation for MIMO cognitive radio application

Cognitive radio is a potential candidate for resource management because of its capability to improve network efficiency and to satisfy the growing demand in the wireless cognitive radio technology applies the machine learning techniques for the betterment of the performance and resource management. Link adaptation is one of the application in Cognitive Radio (CR) resource management system. Link adaptation in cognitive radio is done by the help of sensing the electrometric environment and creating the knowledge base. From the knowledge base the operational parameters and protocols are adjusted to achieve predefined objectives. Many machine learning algorithms are used for cognitive radio performance improvement such as genetic algorithm, Rule Based Reasoning, Fuzzy logic, Artificial Neural networks. Among that genetic algorithm is used to optimize multi parameter simultaneously by iteratively. In this paper multiple parameter adjustment technique based on genetic algorithm are used to optimize bandwidth, band efficiency, transmission power, data rate and Bit error rate. In this work the real time experimental study of multi parameter optimization based 2X2 MIMO link adaption is carried out with the use of genetic algorithm. National instruments PXIe 5673 vector signal generator and 5663 vector signal analyser based SDR platform is used to implement the link adaptation scheme.