Abstract: Cognitive radio (CR) is an intelligent wireless communication system, and its core is to design cognitive engine. This paper presents a design scheme for cognitive engine based on least squares support vector Machine (LS-SVM) to adapt varied wireless environment according to experiential knowledge. A CR simulation cognitive engine is realized on 802.11a simulation platform by collecting data and training the classification and regression model of LS-SVM to learn channel characteristics. Under the premise of studying the sensaed information, a CR reconfiguration is realized according to the user' s needs. The simulation results show that the proposed cognitive engine can effectively realize the learning and reconfiguration function.