Robust Resource Allocation In High-Dynamic UAV Multi-Hop Network

The UAV multi-hop network is a networking mode that applies the traditional mobile ad hoc network to the UAV field.The high-speed movement and three-dimensional spatial distribution of the UAVs cause the UAV multi-hop network to face the network topology frequently change,node-intensive,complex interference environment and unable to access complete information by nodes.On one hand,the high-speed movement of the nodes cause the network topology to change frequently,and the probability of collision and loss of data packets increases.However,the existing MAC protocol is rarely designed to take into account the impact of high mobility of the nodes on network performance.On the other hand,the UAV nodes are distributed in the three-dimensional space,hence the number of nodes around a single UAV is more than that of a node distributed in the two-dimensional space on the ground,which results in network densification and more complicated interference in the network.Though the complicated interference can be mitigated by using the power control technology,the traditional method will cause a lot of signaling overhead.Besides,the wireless channel fluctuations are more severe due to the high-speed movement of the node,and the node has estimation error and channel state information quantization noise,etc.,so that the ideal channel state information is hardly to be fed back,which lead to the information such as the channel state acquired by the node is uncertain.Therefore,it is necessary to design a low-cost and robust power control scheme to mitigate the interference in the network.For the above reasons,the resource allocation scheme in the traditional mobile ad hoc network cannot be directly applied in the UAV multi-hop network scenario.Therefore,it is urgent to study a resource allocation scheme that can be applied to the scenario.Firstly,this paper studies a robust joint physical layer power allocation and MAC layer time slot scheme based on cross-layer theory for the problem of high probability of packet collision,large signaling overhead and complex interference environment in UAV multi-hop network.The resource allocation scheme establishes an energy efficiency optimization model.Then,for the convenience of solving,the original decomposition method is used to decompose the optimization problem into MAC layer time slots assign sub-problems and physical layer power control sub-problems.Then,for the sub-problem of MAC layer time slot allocation,this paper analyzes the two scenarios that mainly cause packet collision and loss.By using the characteristics of GPS information,a Dynamic Location Prediction Time Division Multiple Address(DLP-TDMA)scheme is proposed to effectively reduce the probability of packet collision and loss.In the case of time slot allocation determination,by using the mean field game model which is good at modeling a large number of player interaction scenarios,a power control algorithm with low signaling overhead is proposed.Each node is able to use power control algorithms to effectively improve energy efficiency in its own assigned time slot.Finally,the function and basic implementation flow of the MATLAB UAV multi-hop network simulation platform built in this paper are introduced.Then the performance of the proposed algorithm is verified in the simulation platform.The simulation results show that The proposed slot allocation and power control algorithm can effectively reduce packet loss rate,collision probability and improve network energy efficiency.