Abstract—By employing BS (Base Station) cooperation we can increase substantially the spectral efficiency and capacity of cellular systems. The signals received at each BS are sent to a central unit that performs the separation of the different MT (Mobile Terminal) using the same physical channel. However, we need accurate sampling and quantization of those signals so as to reduce the backhaul communication requirements. In this paper we consider the optimization of the quantizers for BS cooperation systems. Four different quantizer types are analyzed and optimized to allow better SQNR (Signal-to-Quantization Noise Ratio) and BER (Bit Error Rate) performance. Keywords—Base Stations cooperation scheme, Bit Error Rate (BER), Quantizer, Signal to Quantization Noise Ratio (SQNR), SC- FDE. I. INTRODUCTION ASE STATION (BS) cooperation techniques are under consideration for the LTE standard (Long Term Evolution) for high speed data transfer in wireless systems. Since the LTE technology is adopted in many countries of the world and is constantly developing, this technique can be used in future standard releases (see [1], [9] and references within). The application of the BS cooperation schemes for the uplink transmission is designated to decrease frequency reuse factor value up to 1. Frequency reuse factor is the rate at which frequencies can be used in cells of wireless network. Wireless system, operating with smaller frequency reuse factor, provides higher capacity and specter efficiency. Also the BS cooperation architecture provides macro-diversity effects and can compensate the high interference effects inherent to systems operating with universal frequency reuse factor (this interference is especially high at cell edges). The general idea of the base station cooperation technique is that signal, transmitted from MT (mobile terminal) is received by BS and then is sent to the CPU (central processing unit), that performs detection and separation of signals, corresponding to different MTs. The channel capacity between BS and CPU is limited, so signal samples, that are sent from BS and are processed in the CPU has to be quantized in the optimal way. It is K. Firsanov and S. Gritsutenko are with the Omsk State Transport University, Omsk, 644046 Russia (e-mail: Konstantin.firsanov@gmail.com st256@mail.ru). R. Dinis is with IT (Instituto de Telecomunicações) and FCT-UNL (Faculdade de Ciências e Técnologia da Universidade Nova de Lisboa), Lisbon, Portugal e-mail: rdinis@fct.unl.pt). This work was partially supported by the program Erasmus Mundus and Fundação para a Ciência e Tecnologia (FCT) (projects PEst- OE/EEI/LA0008/2013, MPSat PTDC/EEA-TEL/099074/2008 and ADCOD PTDC/EEA-TEL/099973/2008). necessary to represent signal samples with the number of bits large enough to avoid BER performance degradation. The efficient detection and quantization requirements have been determined in [1]. According to this work results, even coarse quantization 4-5 bits in in-phase and quadrature components of the complex envelope provides efficient separation of signals, transmitted by MTs and close to optimum macro diversity gain. But quantization as an operation can be done in a numerous ways. And for different signals at the quantizer input, different quantizer types, having the same number of quantization levels, provide different performance. So in order to select optimal quantizer, it’s necessary to estimate performance of every quantizer type. In this paper we consider the uplink of the broadband wireless communication system that consists of mobile terminals, base stations and CPU. MTs are regarded as employing SC-FDE schemes. Each BS receives signals, performs sampling and quantization and sends signal samples to the CPU, where user detection is done. The separation of signals is done with IB-DFE application, as shown in [9]. In this paper four quantizer types are analyzed: uniform quantizer, non-uniform floating point quantizer, non-uniform optimal Lloyd Max quantizer and non-uniform optimized quantizer, which preserves maximum information on its output. The signal is considered to be complex. The representation of complex signal is regarded as values of its in-phase and quadrature components. All quantizer types are investigated for SQNR and for the system performance in the terms of BER of the detector. The paper is organized as follows: Section II reviews every quantizer type investigated in the paper, Section III estimates each quantizer upon the SQNR term, Section IV contains evaluation of the performance of every quantizer type in the terms of the BER. Section V concludes the paper. II. REVIEW OF QUANTIZATION DEVICES Fig. 1 shows the structure of quantizer, that operates on the complex valued samples. This quantizer can be considered as two identical “I-Q’ memoryless nonlinearities, which operate separately on the real and imaginary parts of the complex samples x, applied to the quantizer input. The equation, that characterizes this class of quantizers, by the relation between input and output samples, can be recorded as: }) (Im{ }) (Re{ x jQ x Q x q + = (1) On the Quantizer Design for Base Station Cooperation Systems with SC-FDE Techniques K. Firsanov, S. Gritsutenko, and R. Dinis B World Academy of Science, Engineering and Technology International Journal of Electrical and Computer Engineering Vol:7, No:7, 2013 1028 International Scholarly and Scientific Research & Innovation 7(7) 2013 scholar.waset.org/1307-6892/16517 International Science Index, Electrical and Computer Engineering Vol:7, No:7, 2013 waset.org/Publication/16517