Cyber-physical battlefield perception systems based on machine learning technology for data delivery

Abstract

Data delivery in Cyber-Physical Battlefield Perception Systems(CPBPS) is a challenging task due to the ubiquity locations and the high mobility of node. Due to the special geographical circumstances, communication networks based on fixed infrastructure are unlikely to be established. This paper presents an air-ground coordination communication transmission network, which consists of Unmanned Aerial Vehicle (UAV) subnets and ground vehicle subnets. The UAVs exploit air-to-air (A2A) and air-to-ground (A2G) communication links to assist vehicle communications. However, overreliance on satellite positioning may cause military information to leak. Therefore, we proposed a K-Nearest Neighbor (KNN )combined with genetic algorithms and based on machine learning system (MLS) for data delivery for battlefield environment to realize the privacy protection and guarantee the security with better prediction. The proposed KNN machine learning system can estimate the movement and path of vehicles based on the mobile information obtained. Furthermore, in order to transmit data of UAVs more efficiently, the genetic algorithms (GA) is utilized to determine the relative location of UAVs. Simulation results verify the performance of proposed algorithm.