Hybrid Supervised-Unsupervised Channel Estimation Scheme with Dynamic Transmission of Pilots

Bibliographic citation

Dapena, A., Castro, P.M. & García-Naya, J.A. 2011, "Hybrid supervised-unsupervised channel estimation scheme with dynamic transmission of pilots", Neural Processing Letters, vol. 33, no. 1, pp. 1-15.

Type of academic work

Academic degree

Abstract

[Abstract] Multiple-Input Multiple-Output (MIMO) digital communications standards typically include pilot symbols in the definition of the transmit signals with the purpose of acquiring the Channel State Information (CSI) using supervised algorithms at the receiver side. Such pilot symbols convey no information and, therefore, system throughput, spectral efficiency and transmit energy consumption are all penalized. In this article, we propose to acquire the CSI combining supervised and unsupervised algorithms. Our strategy avoids the periodical transmission of unnecessary pilots by using a simple decision criterion to determine the time instants when the performance obtained with an unsupervised algorithm degrades or, equivalently, the time instants when pilots are required. We show the performance of this scheme for MIMO systems with Decision-feedback equalizers at the receiver.

Description

The final publication is available at Springer via http://dx.doi.org/10.1007/s11063-010-9161-x

Rights