Design and Development of Adaptive Blind Beamforming Algorithm and its implementation on FPGA

Abstract

The intense and ever growing number of the wireless communication users newlineand its applications has demanded for the design of powerful and efficient system. newlineThese demands are related to spectrum efficiency, channel capacity, improved newlineperformance and low system cost. To meet the demands of wireless communication newlinesystem design, smart antenna technology has proven to be one of the ultimate newlinesolutions. Design of smart antenna based systems involves design and development of complex signal processing algorithm, which are called as Beamforming algorithm . newlineBeamforming algorithms will be executed on processors (DSP or FPGA) for the newlineimplementation of smart antenna systems. With the advancements in the newlinesemiconductor integration technology in particular FPGAs, it has become possible for newlinethe fabrication of algorithm based computing system on a single chip. A simplified approach for the faster and effective hardware implementation of Blind Adaptive beamformer and a novel method towards the development of single chip solution for newlinebeamformer is presented in this work. newline The traditional approach of algorithm based system design or design of any wireless application such as beamformer is laborious and time consuming due to newlineseparate hardware and software design paradigm and involved manual design newlineapproach. With the usage of modern simulation and synthesis tools, the design and newlinedevelopment time of system or module will be minimized and design process can be newlineautomated. Blind adaptive beamforming offers hardware implementation flexibility newlineand well suited for digital modulated data communication applications. Modified newlineNormalized Least Mean Square Normalized Constant Modulus Algorithm newline(MNLMSNCMA) adaptive beamforming algorithm is proposed and implemented. newline Performance of MNLMSNCMA was verified using MATLAB tool and it newlineproduced array gain of 8.54 dB and side lobe gain of -5 dB which was better compare newlineto CMLMS algorithm. Proposed algorithm is convergence ready because of newlinemodification of initial weight values by normalization approach. Synthesis of newlineMNLMSNCMA was carried out using Precision synthesis RTL plus tool of Mentor newlineGraphics in conjunction with HDL coder. Synthesis results in terms of resource newlineutilization and compilation time, with comparison of devices are presented. Such an newlineapproach will be very much beneficial in development of single chip solution for newlinebeamformers effectively with minimum possible time of design newline