Optimization Method And Application Of Vehicle Network Node Based On 5G

The traditional vehicle network data acquisition obtains information by using the geomagnetic coils which set at important intersections and sections,as well as the traffic monitoring means such as cameras on the road surface.This method is limited by coverage areas and instantaneity.Mobile crowd-sensing technology makes the data acquisition device have more mobility,and the participating vehicles are more regularly driven,and the perceived coverage is wider.However,research on the data acquisition of the car network mobile crowd-sensing is still insufficient at present,which has no reasonable node selection mechanism especially.This paper firstly designs a reasonable service node selection method,and then applies the selected service node to the routing decision,and designs an effective data forwarding mechanism.First of all,this paper proposes an improved Markov location prediction method based on Gaussian analysis,which is used to calculate the probability that the participating vehicles reach the target area.The Markov position prediction algorithm based on Gaussian analysis proposed in this paper no longer adopts this method of equally spaced time segments to find time transition points.Instead,according to the Gaussian mixture model,the transition probability between each position is calculated over a whole period of continuous time.According to the obtained relationship between the obtained time and the transition probability,we can find the most likely time transition points.Finally,these time points found can be regarded as the state transition point of the Markov model,then the Markov model could be established to calculate the probability of the vehicle reaching a certain position.Secondly,this paper designs a method based on genetic algorithm to select the optimal vehicle service node set by useing the probability of the service vehicle reaching the target area and the absolute distance of the service vehicle to the target area,which calculated by improved Markov location prediction algorithm and the road network structure.The method can find the vehicle service node that meet the condition and complete the sensing service at a relatively fast speedwhich can not only improve the sensing service quality,but also effectively save resources.Finally,this paper proposes an improved location-based opportunistic network routing protocol,which applies the service nodes selected above to routing decisions.The data transmission between the vehicle nodes can adopt the data forwarding mechanism in the opportunistic network,which can better adapt to the urban and dynamic vehicle networking environment,improve the delivery rate,reduce the possibility of network congestion,and reduce the network spending.