Please use this identifier to cite or link to this item: http://hdl.handle.net/10773/18537
Title: Analytical characterization of CSI quantization effects on broadband multipath channels
Author: Teodoro, Sara
Silva, Adão
Dinis, Rui
Gameiro, Atílio
Keywords: Channel state information
Nonlinear distortion
Quantization errors
Limited feedback
Issue Date: 2017
Publisher: Elsevier
Abstract: The knowledge of channel state information (CSI) at transmitter is crucial in precoding techniques to efficiently explore wireless multiuser multi-antenna capabilities namely to mitigate the interference and/or to separate spatial streams. For frequency division duplex systems, the CSI available is a quantized version fed back from the receivers, and the nonlinear effects associated to CSI quantization can lead to significant performance degradation. The knowledge of the statistical characteristics of those nonlinear distortion effects is important for performance evaluation purposes, as well as to design robust precoding techniques. In this paper we consider a general broadband multipath channel sampled in the frequency domain and obtain the analytical statistical characterization of the nonlinear distortion effects associated to CSI quantization. Our analytical approach is valid for non-uniform quantizers, and it also includes quantizer saturation and channel oversampling effects. It provides a detailed statistical characterization in time and frequency domains, associated to the channel impulse and frequency responses, respectively. The derived theoretical statistical characterization of quantization distortion is validated by simulation, showing its high accuracy. As an application, we consider a multicarrier transmission employing a precoding technique with reduced feedback techniques employing a small number of quantization bits per channel sample.
Peer review: yes
URI: http://hdl.handle.net/10773/18537
DOI: 10.1016/j.dsp.2016.09.015
ISSN: 1051-2004
Appears in Collections:DETI - Artigos

Files in This Item:
File Description SizeFormat 
sara_dsp.pdf751.55 kBAdobe PDFrestrictedAccess


FacebookTwitterLinkedIn
Formato BibTex MendeleyEndnote Degois 

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.