Based on Deep Learning CSI Recovery for Uplink Massive Device Dynamic Internet of Thing

Abstract

This paper investigates uplink massive MIMO communication scenarios in dynamic Internet of things (IoT) networks. In this paper, dynamic IoT mainly consists of the Internet of vehicles (IoV) and the original IoT network. Because the speed of vehicle is very fast, the number of users is constantly changing in the IoT network, which leads the structure of the IoT network to change. We mainly consider how to obtain the channel state information (CSI) of active users. Due to active users and inactive users, the system model is considered a sparse structure. This structure inspired us to give an algorithm suitable for the sparse structure and obtain more accurate channel state information of dynamic IoT networks, though these numerical results, under the premise of guaranteeing performance, can greatly reduce the complexity of the algorithm.