Energy detection based spectrum sensing data mining for safety-message delivery in CR enabled VANET

A cognitive ratio (CR) is an intelligent radio transceiver designed to use the best wireless channels in its vicinity. It automatically detects available channels in wireless spectrum, and then accordingly changes its transmission or reception parameters. VANET is a special ad-hoc network that equips vehicles with wireless communication devices. Hence vehicles can talk with Road side unit (RSU) and the traffic information can be shared among them. This work improves algorithm is to improve the reliability of safety message transmission and to control the road accident. Prediction algorithm methods used are Spatial Correlation Mining, Temporal Correlation Mining, and Historical Spectrum Sensing Data Mining(HSS). All vehicles share hidden channel state transition model in statistic. Using Hidden Markov Model(HMM), the relationship between observations and hidden channel states are estimated, and characterized the channel state transition rule among sequential spectrum sensing decisions. The correlation among spectrum sensing data from different vehicles is called spatial correlation and Dirichlet Process(DP) is used to model this property. HSS prediction algorithm is used to pick out the channel with the greatest probability of availability. This can efficiently avoid harmful interference to PUs and reduce transmission delay of urgent safety message applications.