Joint Channel Estimation and Data Detection for Time-Varying MIMO Channels in UAV Networks

In the context of wireless communications, channel estimation and data detection are two pivotal tasks that exert considerable influence on system performance, such as enhancing spectral efficiency and improving quality of service (QoS). Nevertheless, these tasks become more challenging as user mobility increases, e.g., in unmanned aerial vehicle (UAV) networks, leading to channel variations over time. Consequently, our focus in this study is on an uplink massive multi-input multi-output (MIMO) system in a UAV network with time-varying channels. We propose an online processing approach for joint estimation and detection (JED) in these channels. Specifically, we introduce a method that utilizes variational Bayes (VB) inference to approximate the true posterior distributions. Our assessment of the VB method encompasses its performance in terms of symbol error rate (SER) and computational complexity. Additionally, we conduct an analysis of how time correlation and communication time impact its effectiveness.