The direct conversion receiver (DCR) architecture has received much attention in the last few years as an effective means to obtain user terminals with reduced cost, size, and power consumption. A major drawback of a DCR device is the possible insertion of I/Q imbalances in the demodulated signal, which can seriously degrade the performance of conventional synchronization algorithms. In this paper, we investigate the problem of carrier frequency offset (CFO) recovery in an OFDM receiver equipped with a DCR front-end. Our approach is based on maximum likelihood (ML) arguments and aims at jointly estimating the CFO, the useful signal component, and its mirror image. In doing so, we exploit knowledge of the pilot symbols transmitted within a conventional repeated training preamble appended in front of each data packet. Since the exact ML solution turns out to be too complex for practical purposes, we propose two alternative schemes which can provide nearly optimal performance with substantial computational saving. One of them provides the CFO in closed-form, thereby avoiding any grid-search procedure. The accuracy of the proposed methods is assessed in a scenario compliant with the 802.11a WLAN standard. Compared with existing solutions, the novel schemes achieve improved performance at the price of a tolerable increase of the processing load.
Periodic Preamble-Based Frequency Recovery in OFDM Receivers Plagued by I/Q Imbalance
D'Amico, Antonio;Morelli, Michele;Moretti, Marco
2017-01-01
Abstract
The direct conversion receiver (DCR) architecture has received much attention in the last few years as an effective means to obtain user terminals with reduced cost, size, and power consumption. A major drawback of a DCR device is the possible insertion of I/Q imbalances in the demodulated signal, which can seriously degrade the performance of conventional synchronization algorithms. In this paper, we investigate the problem of carrier frequency offset (CFO) recovery in an OFDM receiver equipped with a DCR front-end. Our approach is based on maximum likelihood (ML) arguments and aims at jointly estimating the CFO, the useful signal component, and its mirror image. In doing so, we exploit knowledge of the pilot symbols transmitted within a conventional repeated training preamble appended in front of each data packet. Since the exact ML solution turns out to be too complex for practical purposes, we propose two alternative schemes which can provide nearly optimal performance with substantial computational saving. One of them provides the CFO in closed-form, thereby avoiding any grid-search procedure. The accuracy of the proposed methods is assessed in a scenario compliant with the 802.11a WLAN standard. Compared with existing solutions, the novel schemes achieve improved performance at the price of a tolerable increase of the processing load.File | Dimensione | Formato | |
---|---|---|---|
printedVersion.pdf
solo utenti autorizzati
Tipologia:
Versione finale editoriale
Licenza:
NON PUBBLICO - Accesso privato/ristretto
Dimensione
693.31 kB
Formato
Adobe PDF
|
693.31 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
manuscript.pdf
accesso aperto
Tipologia:
Documento in Post-print
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
447.67 kB
Formato
Adobe PDF
|
447.67 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.