Channel tracking of uncoordinated access under uncertain evolution for massive connectivity with massive MIMO

Abstract

Massive MIMO technologies are a major candidate to serve high-rate Internet of Things (IoT) with moving or motionless devices if all channels of devices are known to BS. In this paper, an unknown dynamical model for channel evolution is assumed over time, and regardless of whether devices send pilots or not, joint channel estimation and tracking are investigated. At first, a new suboptimal tracker, Uncertain Interacting Multiple Model (UIMM), is derived for the coordinated access. We define a new setup, referred to as an uncoordinated access, in which many devices with an unknown probabilities access BS without coordination. Then, three novel suboptimal trackers are designed based on UIMM for this uncoordinated access. To decrease computational complexity of the designed trackers, two schemes are applied. The Mean Square Error (MSE) gap between the mentioned scenario and the previous uncoordinated access is computed theoretically and represented by simulation results. Also, a new fundamental limit is defined which is referred to as Cost of Uncoordinated Access under Uncertain Channels (CUAUC). The CUAUC is used to evaluate the MSE gap between the optimal coordinated tracker and the designed suboptimal uncoordinated trackers. Finally, the CUAUC and the efficiency of the designed trackers are evaluated via simulations.