IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 8, NO. 6, JUNE 2009 2901 Optimized Multistandard RF Subsampling Receiver Architecture Rim Barrak, Adel Ghazel, Senior Member, IEEE, and Fadhel Ghannouchi, Fellow, IEEE Abstract—This paper presents a novel subsampling-based down-conversion topology for multistandard radio receiver de- sign. This receiver topology is based on two subsampling stages. The first stage has a fixed RF subsampling frequency; however, the IF sampling frequency of the second stage is variable and depends on the standard being considered. This approach overcomes various undesirable effects related to the sampling frequencies and noise aliasing. By optimizing the choice of the RF subsampling clock frequency, complete multistandard RF bands are down-converted to the same IF band. Quadrature baseband channel downconversion is achieved by a tunable IF sampling frequencies clock. A tunable band-pass RF filter and an IF band-pass filter are designed to perform anti-aliasing and limit wideband noise. A generic design methodology is proposed and validated through its application to a GSM, UMTS and IEEE-802.11g multistandard receiver. Simulation results of this receiver design example confirm the validation of proposed subsampling receiver topology and show the efficiency of the design methodology. Index Terms—Anti-aliasing, multistandard receiver, radio re- ceiver, RF subsampling, sampling noise. I. I NTRODUCTION F OR the last two decades, commercial success of wireless digital communication systems has enabled the continual development of radio communication standards, introducing new mobile services from high-quality voice to high bit rate data and multimedia communications [1-4]. Important academic and industrial research has been carried out, lead- ing to the design and manufacturing of low-power wireless equipment with high integration for each radio communica- tion standard. Today, the new commercial challenge in the enhancement of mobile technology is the development of multistandard software defined radio (SDR) based wireless mobile equipment that offer end-users multimode and multi- service facilities with the use of low-cost, low-power and highly integrated devices. The ultimate approach in achieving multimode operation is to design multistandard radio receiver hardware that can be reconfigured by software, with maximum hardware function- ality sharing between the various standards. High selective Manuscript received June 1, 2007; revised September 12, 2007; accepted July 3, 2008. The associate editor coordinating the review of this paper and approving it for publication was K. Sowerby. R. Barrak and A. Ghazel are with Cirta’Com Laboratory, Ecole Sup´ erieure des Communications de Tunis, 2088 Cit´ e technologique des Communications El Ghazala - Ariana - Tunisia (e-mail: {rim.barrak, adel.ghazel}@supcom.rnu.tn). F. Ghannouchi is with Intelligent RF Radio Laboratory, Electrical and Computer Engineering Department, University of Calgary, Calgary, Alberta, Canada T2N 1N4 (e-mail: fghannouchi@ucalgary.ca). Digital Object Identifier 10.1109/TWC.2009.070584 narrow-band superheterodyne radio frequency (RF) architec- tures, however, are difficult to realize in practice; and, since they are not considered as valuable solutions for new wideband or multistandard applications, three main wideband receiver architectures were recently developed for these applications [5]. The first one, wideband intermediate frequency (IF) ar- chitecture (WB-IF), converts the entire RF band to IF using a fixed local oscillator (LO) frequency and high-performance balanced down-converter, such as the six-mixer configuration based on the Weaver technique, in order to ensure high image rejection performance. It has the disadvantage of using a large number of analog components, which increases power consumption and reduces receiver integrability and flexibility. The second one, IF sampling architecture, down-converts RF channels to a fixed IF using a tunable LO and then per- forms a second frequency down-conversion using a band-pass analog-to-digital conversion (ADC). This architecture imposes a more stringent dynamic range for the ADC stage due to IF sampling. The third one, direct conversion architecture, down- converts RF signals directly to baseband, increasing receiver integrability and flexibility. However, this analog architecture suffers from a DC offset problem and requires highly accurate phase and gain matching between quadrature LOs and signal paths, which imposes large constraints on LO specifications in terms of tunable range and phase noise. To overcome analog down-conversion constraints, a new design approach based on a radio frequency (RF) subsampling technique has recently been proposed so that integrable RF receiver front-end architectures for Global System for Mo- bile communication (GSM) [6], wireless local area network (WLAN) [7] and Bluetooth [8] can be designed. These ar- chitectures are based on switched-capacitor (SC) circuits to perform discrete-time processing after sampling. This leads to excellent improvement in radio receiver performance, integra- tion and low-power dissipation, but the designed RF front- end receivers are optimized for only a single standard [6- 8]. Multistandard operation and adaptability need a tunable sampling frequency for each RF band and for each carrier frequency in the same band. This implies difficult constraints on sampling clock specifications, in terms of tunable range and phase noise. A low sampling frequency is required to reduce power consumption of the ADC; however, it increases track- and-hold (T&H) circuit aperture jitter noise and subsampled thermal noise. All of these constraints provided the motivation for the subject of this paper, in that definitions of appropriate design solutions are needed to take advantage of this new RF architectural concept. In this paper, a novel RF subsampling-based down- 1536-1276/09$25.00 c  2009 IEEE