On the channel tracking under uncertain state model for multiuser massive MIMO in high-rate Internet-of-Things

Abstract

The significant advancement of the internet-of-things has led to a dramatic increase in the number of moving or motionless devices that are connected to the cellular network. For handling such a great number of devices, 5G networks make use of massive MIMO technology. By considering a dynamic model for all channels variation of devices, joint channel estimation and tracking of all devices are studied regardless of the transmission or non-transmission of pilot by each device in the massive MIMO system. The channel predicting and updating is contingent upon the channel evolution model. This evolution does not necessarily follow a predetermined or fixed model. In this paper, the Recursive Least Squares (RLS) tracker and the Interacting Multiple-Mode (IMM) tracker are developed for tracking these channels. In addition, the autoregressive (AR) coefficients are obtained theoretically for all channels between devices and BS antennas by considering an AR model as an approximation of the channel model between a device and a BS antenna. Consequently, the optimal noise covariance of channel state is obtained adaptively online by means of these coefficients. Furthermore, exponential stability and error bound of the IMM tracker as well as the asymptotic stability of the RLS tracker are derived. Finally, the performance of the introduced trackers is assessed through simulations, and the reduced sum-rate of massive MIMO systems is shown under the effect of time-varying channel.