Vehicle-mounted ad-hoc network intersection prediction routing method based on CP (Counter Propagation) neural network

Abstract

The invention discloses a vehicle-mounted ad-hoc network intersection prediction routing method based on a CP (Counter Propagation) neural network, and mainly aims to solve the problems of frequent link fracture among nodes in a data packet forwarding section, large data packet transmission delay and low transmission success rate due to high-speed moving of vehicles, frequent change of network topology and a plurality of vehicle distribution densities in a traffic environment. According to the implementation scheme, a next section priority at which data packets are forwarded on an intersection is classified by use of the CP neural network; when a node arrives at the intersection, an average link communication probability of adjacent sections, average node densities of the adjacent sections and a distance ratio of a distance from a next intersection to a destination to a distance from the current intersection to the destination are taken as inputs of the CP neural network in one Hello message period; the priorities of adjacent data packet forwarding sections are taken as outputs; and a section with a highest priority is selected to be an optimal data packet forwarding section. Through adoption of the vehicle-mounted ad-hoc network intersection prediction routing method, the data packet forwarding success rate is increased, and the transmission delay of the data packets is shortened. The method can be applied to a vehicle-mounted ad-hoc network.

Description

technical field [0001] The invention belongs to the technical field of communication, and mainly relates to a routing method in a vehicle ad hoc network, which can be used for routing decisions at road intersections in urban scenes. Background technique [0002] In the traffic environment, the high-speed movement of vehicles, the frequently changing network topology and the various distribution of vehicle densities make the node links frequently broken, which has become a problem that VANET routing protocols must face and solve. The existing predictive routing algorithms at intersections have the following problems: First, some protocols do not consider vehicle density, and frequent link breaks occur in selected routes, resulting in data loss or increased delay; second, through statistical measurement methods or historical The average vehicle density is obtained as the optimal route based on statistical data. The vehicle density will change dynamically with the high-speed mo...

AI Technical Summary

Problems solved by technology

The existing predictive routing algorithms at intersections have the following problems: First, some protocols do not consider vehicle density, and frequent link breaks occur in selected routes, resulting in data loss or increased delay; second, through statistical measurement methods or historical The average vehicle density is obtained from statistical data as the optimal route. The vehicle density will change dynamically with the high-speed movement of the vehicle. Sometimes the selected path is poor, which has a great impact on the real-time performance of the routing protocol.

Third, from the perspective of the MAC layer, the greater the node density, the more packets the sending node will send, and the longer the time for the node to access the channel, this will increase the transmission delay of the packet, which is not conducive to the packet transmission. send quickly

Fourth, the high vehicle density can improve the connectivity between nodes to a certain extent, but the uneven distribution of vehicle density cannot guarantee the certain connectivity between nodes, and the reliability of routing based on vehicle density information is poor

If you set weights for the three parameters of link connectivity probability, node density, and distance ratio, and select road segments by calculating the scores of adjacent road segments, the weight selection of each parameter will have a lot of blindness

Images