Leveraging on the Impact of Imperfect Channel Estimation for MIMO Relaying Systems

Nikolaos I Miridakis,Theodoros A Tsiftsis,Guanghua Yang

doi:10.1109/access.2019.2939501

Nikolaos I Miridakis, Theodoros A Tsiftsis + Show 1 more

Open Access

https://doi.org/10.1109/access.2019.2939501

Copy DOI

Abstract

In this paper, we study a multi-user multiple-input multiple-output (MIMO) dual-hop wireless relaying system, where the decode-and-forward protocol is applied at the intermediate relay node. The well-known MIMO spatial multiplexing technique is employed, where the linear yet efficient zero-forcing detection is utilized at the receiver side of each hop. The practical scenario of channel estimation errors and/or delayed channel response is embraced when the signals experience mutually independent Rayleigh fading. In oppose to the approach followed in the vast majority of the published works, in this manuscript the channel estimation error term is reckoned as a signal rather than interference or noise, which may significantly enhance the overall system performance. Analytical results involving outage probability and average symbol error rate are extracted in closed-form. Capitalizing on the above performance metrics, new engineering insights are manifested highlighting clearly the superiority of the proposed approach against the conventional one when a delayed/imperfect channel response is present.

Highlights

One of the most challenging issues that modern wireless communication systems face is the densely occupied spectrum bandwidth that urges the need for spectrally efficient methods
Spatial multiplexing independent data streams that combined with both linear and non-linear detection schemes have been adopted in MU-Multiple-input-multiple output (MIMO) systems
Most literature works conventionally handle the impact of channel estimation error as a noise-like segment upon the detection of the received signal. It was shown in [12], [13] that the channel estimation error can be processed in a more efficient basis, by treating it as a signal rather than noise, since the error term can be demodulated via envelope detection; thereby, it can further enhance the overall system performance. Motivated by the latter observation, we study spatial multiplexed dual-hop relaying multiuser MIMO (MU-MIMO) systems operating over multipath Rayleigh channels and ZF detection at the receiver end

Summary

INTRODUCTION

One of the most challenging issues that modern wireless communication systems face is the densely occupied spectrum bandwidth that urges the need for spectrally efficient methods. It was shown in [12], [13] that the channel estimation error can be processed in a more efficient basis, by treating it as a signal rather than noise, since the error term (along with its associated transmitted symbol counterpart) can be demodulated via envelope detection; thereby, it can further enhance the overall system performance Motivated by the latter observation, we study spatial multiplexed dual-hop relaying MU-MIMO systems operating over multipath Rayleigh channels and ZF detection at the receiver end. At each consecutive frame and before the actual data transmission phase, a training phase is executed at the receiver side for CSI acquisition This is implemented through pilot symbols whereas the classical minimum mean squared error (MMSE) channel estimation is being implemented at the receiver side (i.e., at the relay and destination), while full-blind transmitters are assumed. It is noteworthy to state that the relay uses M out of Nr antennas for transmission during the second hop (whenever Nr > M )

OUTAGE PROBABILITY

ASYMPTOTICALLY HIGH SNR REGIME

CONCLUSION