Joint Channel Estimation and Symbol Detection for UAV-Assisted Systems Using Tensor Framework

In this paper, we focus on the unmanned aerial vehicle (UAV)-assisted systems and present three joint channel estimation and symbol detection algorithms using tensor framework. Specifically, the UAVs amplify and forward the received signals from the vehicle to the base station (BS) transmitted through time-varying channels, and the signals received at the BS are constructed to the PARAFAC model or the PARATUCK2 model. Then, three different algorithms are proposed to fit the tensor models, corresponding to different communication scenarios. In particular, a tensor decomposition algorithm that can convert the PARATUCK2 model to the parallel factor (PARAFAC) model is proposed. Our simulation results show that the proposed algorithm can effectively estimate channels and detect symbols.