An approach was proposed to apply a self-organizing map neural network to organize vehicular ad hoc net-works

which were used for car driving-aid systems and automatic driverless systems in intelligent transportation area.Based on the location and driving direction information of the periodic message of each vehicle

the vehicles were organ-ized into vehicular ad hoc networks according to the similarity of destination and driving direction.The organized vehi-cles only communicate mainly with the vehicles in the same vehicular ad hoc network and deal with messages from the same network.Theoretic analysis and simulation results show that the system transmission delay of this proposed algo-rithm is lower than one of the previous method

and the system throughput is remarkably improved.