: Synthetic aperture radar is an imaging radar which provides high resolution images from long distances during day and night and in all weather conditions. Large amount of data, high data rate and complex signal processing are inherent characteristics of this radar that in practical implementation, makes real-time image formation difficult. Thus, in hardware implementation, there is a tendency to reduce hardware complexity and simultaneously increase speed and accuracy of computations. Fixed point computations in hardware implementation is preferred to floating point computations because it reduces required hardware resources. However, the accuracy of fixed point representation is less than floating point representation and one of the challenges that engineers face in hardware implementation is to find the proper word length in a way that the least hardware complexity and acceptable image quality are achieved. In this thesis, firstly, three main synthetic aperture radar imaging algorithms, namely range Doppler algorithm, Chirp Scaling algorithm and ?K algorithm are implemented on FPGA. Then, the amount of used hardware resources and output image quality for different word lengths are extracted. Based on the results, the proper word lengths are derived. For the output image, point target with data matrix of the size 64*128 and area target of the size 1024*1024 are formed with the three algorithms and the results are evaluated with Synthetic Aperture Radar image quality assessment parameters. The implementation is performed with pipelined architecture to increase processing speed. In order to design the hardware model, system generator software made by Xilinx is used and the image formation systems are implemented on Zynq platform. Results show that the proper word length for RDA, CSA and ?K is 14, 16 and 16 bits respectively. It is concluded from the results that CSA has the best accuracy and the most resource consumption and RDA has the least accuracy and resource consumption. This is because of the use of accurate phase functions of CSA. Keywords: fixed point hardware implementation, imaging algorithms, proper word length, synthetic aperture radar, system generator
