面向异构无人机中继网络的负载均衡:一种分层博弈方法

Load Balancing in Heterogeneous UAV Relay Network:A Stackelberg Game Method

研究了大规模无人机中继网络中的负载均衡问题。为了激励空闲无人机作为中继节点协助网络边缘无人机实现业务回传,提出一种基于价格激励机制的负载均衡方案,以提升系统容量,同时将该问题建模为斯坦伯格博弈,其中中继无人机为领导者,网络边缘无人机为跟随者。网络边缘无人机的效用函数中综合考虑了直接传输和中继传输获得的吞吐量和以及支付给中继节点的价格开销。中继无人机的效用函数为其所协助传输的无人机支付的总和。此外,理论证明了所提出的斯坦伯格博弈存在均衡解,并分析了中继无人机的定价策略,提出基于粒子群优化(PSO)的斯坦博格均衡搜索方案。最后,仿真验证了所提算法的有效性。所提方案能够显著提升网络容量,在给定参数设置下,全网吞吐量提升53%。

Exploration on the load balancing problem in large-scale UAV relay networks is done. In order to stimulate the idle drone as a relay node and assist the network edge drone to achieve service backhaul, a load balancing scheme based on price incentive mechanism is proposed, so as to improve system capacity. And the problem is modeled as the Stackelberg game, in which the relay drone is the leader and the network edge drone the follower. In the utility function of the network edge drone, the throughput obtained by direct transmission and relay transmission and the price overhead paid to the relay node are comprehensively considered. The utility function of the relay drone is the sum of the payments paid by the drones it assists in transmitting. The theory indicates that the proposed Stackelberg game has an equilibrium solution, and analysis is done on the pricing strategy for the relay on PSO （Particle Swarm Optimization） is proposed. drone. A Stackelberg equilibrium search scheme based Experiment simulation verifies the effectiveness of the proposed algorithm, and indicates that this scheme can significantly improve the network capacity. Under the given parameter settings, the overall network-throughput is increased by 53%.