Intelligent Beamforming for UAV Assisted IIoT Based on Hypergraph Inspired Explainable Deep Learning.

The development of industry is inseparable from the development of Industrial Internet of Things (IIoT), however, the limitations of previous wireless technologies can greatly affect the performance of all aspects of IIoT. With the continuous development of artificial intelligence (AI), deep learning enabled intelligent communication has been widely studied towards B5G and 6G. While, most existing neural networks for the current research often have poor performances on scalability and universality issues. In this paper, an unsupervised learning framework is introduced with mixed interference based graph neural network (MIGNN), to solve the distributed beamforming problem in unmanned aerial vehicle (UAV) assisted IIoT, including both machine to machine (M2M) and UAVs. Firstly, we constructed the system model of heterogeneous network and considered the beamforming of all the air-to-ground (A2G) and industrial M2M links with mutual co-channel interferences. Secondly, we formulated the resource allocation problem with a heterogeneous graph structure, to describe the relationships of different links and the interference among UAV assisted IIoT. Then, a MIGNN model is proposed and the hypergraph idea is integrated to compensate for the traditional graphs caused information loss. Furthermore, an unsupervised learning method is proposed to train the hyper-MIGNN for achieve the optimal beamforming results. At last, numerical simulations are conducted for both the proposed method and the traditional ones on beamforming performances and efficiencies. From the experimental results, it can be verified that comparisons with traditional deep learning methods, the proposed one has stronger scalability. Moreover, the hyper-MIGNN gets rid of the application limitations of traditional graph neural network (GNN) for just homogeneous networks.