Hardware efficient FIR filter design using Global Best Steered Quantum Inspired Cuckoo Search Algorithm

In this paper, a new algorithm namely Global Best Steered Quantum Inspired Cuckoo Search Algorithm (GQICSA) is proposed for obtaining optimized set of coefficients to implement Finite Impulse Response (FIR) Filter. Adder cost of a filter is estimated after quantizing the filter coefficients followed by Common Sub-expression Elimination (CSE). We found from the simulation results that reduction in word length of coefficients does not make the filters fail to achieve the ideal frequency response. Moreover, filters developed using GQICSA outperform the benchmark filters designed by Parks McClellan Algorithm in terms of stopband attenuation. Analysis of the results reveal that GQICSA not only improves over various conventional algorithms including Cuckoo Search Algorithm (CSA), it also surpasses modified version of CSA, Quantum Inspired CSA (QICSA) updated using quantum principles, for optimizing filter coefficients to design lower hardware costing filter without compromising the filter responses and efficiency. GQICSA also provides significant improvements compared to CSA and QICSA in terms of stopband attenuation and execution time for higher order filter design. Efficiency of GQICSA over QICSA and conventional CSA is also proved with 16 benchmark functions.