An improved traffic safety information fusion algorithm in internet of vehicles

In Internet of Vehicles, information fusion can decrease the throughput of transmitted information by compressing the redundant information to improve the wireless resource's efficiency. In this paper, a D-S evidence theory based traffic safety information fusion algorithm is proposed based on the spatio-temporal correlation of traffic safety information and the advantages of the Bajraktarevic mean. Firstly, the spatio-temporal correlation status parameters of different vehicles are calculated by Cauchy's equation. Secondly, the basic probability assignments of status parameters are obtained by Bajraktarevic mean function to get the status parameters sets and their confidence sets. Finally, the time and location parameter sets are fused based on the Dempster fusion rule respectively. The fusion result with the maximum confidence will be chosen to disseminate. Simulation shows that the proposed algorithm can reduce the errors in space and time. The complexity of algorithm can be reduced effectively too.