Abstract: An automatic word-length determination tool (SATRANS) based on the simulated annealing algorithm was developed. SATRANS can  automatically transform the system from floatingpoint model to fixed-point model and provide a series of word-length solutions that form a tradeoff curve for hardware complexity vs. signal quality. SATRANS was demonstrated to find word-length for an infinite impulse response filter (IIR). The results show that SATRANS can provide better word-length solution in comparison to the traditional search method based on  greedy strategy. The word-length optimized IIR targeting Xilinx Virtex-6 FPGA device was implemented, which improves the performance  by 12.4% and 62.8% while saves almost 93.9% and 97.9% of area in comparison to the IEEE single and double floating-point generators.  

Key words: word-length optimization, simulated annealing algorithm, floating-point to fixed-point, infinite impulse response filter