Cognitive radio (CR) engines often contain multiple system parameters that require careful tuning to obtain\r\nfavorable overall performance. This aspect is a crucial element in the design cycle yet is often addressed with ad\r\nhoc methods. Efficient methodologies are required in order to make the best use of limited manpower, resources,\r\nand time. Statistical methods for approaching parameter tuning exist that provide formalized processes to avoid\r\ninefficient ad hoc methods. These methods also apply toward overall system performance testing. This article\r\nexplores the use of the Taguchi method and orthogonal testing arrays as a tool for identifying favorable genetic\r\nalgorithm (GA) parameter settings utilized within a hybrid case base reasoning/genetic algorithm CR engine\r\nrealized in simulation. This method utilizes a small number of test cases compared to traditional design of\r\nexperiments that rely on full factorial combinations of system parameters. Background on the Taguchi method, its\r\ndrawbacks and limitations, past efforts in GA parameter tuning, and the use of GA within CR are overviewed.\r\nMultiple CR metrics are aggregated into a single figure-of-merit for quantification of performance. Desirability\r\nfunctions are utilized as a tool for identifying ideal settings from multiple responses. Kiviat graphs visualize overall\r\nCR performance. The Taguchi method analysis yields a predicted best combination of GA parameters from nine\r\ntest cases. A confirmation experiment utilizing the predicted best settings is compared against the predicted mean,\r\nand desirability. Results show that the predicted performance falls within 1.5% of the confirmation experiment\r\nbased on 9 test cases as opposed to the 81 test cases required for a full factorial design of experiments analysis.
Loading....