Amplify and Forward Relaying Optimization for Uplink CDMA Communications Subject to Constant Multiple Access Interference Cost Naoufel Debbabi † , Sami Chtourou ‡ , Ines Kammoun * , Mohamed Siala † † Sup’Com, El Ghazala, Tunisia ‡ Military Academy, Fondouk Ejjedid, Tunisia * ENIS, Sfax, Tunisia naoufel.debbabi@tunisiana.com, sami.chtourou@isecs.rnu.tn, ines.kammoun@ieee.org, mohamed.siala@supcom.rnu.tn Abstract— In radio mobile cellular networks, the users located at the cell border are suffering from a bad quality of service (QoS). Cooperative communications is a promising technique that can overcome this weakness by allowing the users to profit from powerful benefits of the spatial diversity. Nevertheless, abstraction is usually made on the impact of the additional multiple access interference inherent to relaying. In this paper, we propose a novel approach that optimizes the system performance of an amplify-and-forward (AF) relaying scheme subject to a constant multiple access interference cost for the uplink CDMA (Code Division Multiple Access) communications. We present a variety of simulation results illustrating the cellular system performance enhancement in term of BER (Bit Error Rate) over direct transmission. Keywords— Cooperative Relaying, CDMA, uplink, multiple access interference cost, performance optimization. I. INTRODUCTION Medium- and small-scale propagation fading in radio-mobile cellular environments is a serious drawback that can highly degrade the performance of a transmission scheme, mainly for users in the cell border. Cooperative relaying, through virtual antenna arrays, is a promising technique that can overcome this weakness by reducing the fading effects when combining multiple replicas of the transmitted signal. Therefore, relaying is able to increase the cell coverage and guarantee a suitable QoS for users located at cell border. Recently, cooperative relaying has been receiving a growing interest. Two cooperative relaying schemes are typically used: Amplify-and-Forward (AF) and Decode-and-Forward (DF). In both, the source can transmit the signal to the destination by the aid of one relay (one-hop relaying) or many relays (multi-hop relaying). Many research activities have proposed new relaying techniques that outperform the conventional AF and DF relaying [1]. Other research works have focused on increasing the system capacity [2]-[4]. Moreover, [5] and [6] have tackled system performance optimization, in terms of BER, for one-hop cooperative relaying. A common drawback of previous works is the omission of the additional multiple access interference inherent to relaying which translates into a decrease in cellular system capacity, especially in case of CDMA communications. In this paper, we quantify the global interference in the system by defining a global multiple access interference cost, which reliably reflects the capacity loss incurred by one-hop AF relaying use. Then, we optimize the cellular system performance, in terms of BER, for uplink CDMA communications, subject to a constant multiple access interference cost. This paper is organized as follows. Section II introduces the system model. In section III, our novel optimization approach of the performance of an AF-based relaying scheme is depicted. Then, section IV is dedicated to some simulation and numerical results high-lighting the performance enhancement of the optimized relaying scheme over direct transmission. Finally, some conclusions and perspectives are provided. II. SYTEM MODEL This work focuses on the CDMA uplink communication. The hexagonal cellular network layout pattern, encompassing one- hop AF relaying, is shown in Figure 1. Based on two cooperation phases, the source (S) first sends the transmitted signal to the relay (R) and the destination (D). Then, the relay simply amplifies the received signal from the source before relaying it to the destination. Within a given cell, SD h , SR h and RD h quantify respectively Source-Destination (S-D), Source-Relay (S-R) and Relay- Destination (R-D) propagation channel losses, which take into account Rayleigh fading. The path loss coefficients follow a propagation model of 1/ d α , where α denotes the path loss factor. Shadowing is omitted here. 978-1-4244-5091-6/09/$25.00 ©2009 IEEE 1012