OTFS-based massive MIMO with fractional delay and Doppler shift: the URLLC case

This paper considers orthogonal time frequency space (OTFS)-based uplink multiuser multiple-input multiple-output (MU-MIMO) systems in the ultra-reliable-and-low-latency communication (URLLC) space. The maximum-ratio-combining (MRC) scheme is first applied at the base station (BS) to combine all the received symbols and then the maximum likelihood (ML) decoding metric is employed to decode the intended symbols. To better characterize the system performance, we consider the ideal pulse shape functions at both users and BS side while the effect of fractional delay and Doppler shift (FDDS) is also taken into account. Based on this, we first derive the theoretical refined achievable coding rate under short-packet transmission for both joint and individual decoding at the BS. Subsequently, the corresponding asymptotic results under massive MIMO are presented. Capitalizing on these results, we draw a series of interesting insights: i) when the sampling resolution along the delay domain is not high enough, the achievable coding rate is more accurate than the channel capacity as the performance metric; ii) almost no diversity gain is obtained for the FDDS under the case considered in this paper; iii) the achievable coding rate is agnostic to the users’ speed.