Optimal fractional delay-IIR filter design using cuckoo search algorithm.

Area of Science:

• Digital Signal Processing
• Meta-heuristic Optimization Algorithms

Background:

• Fractional delay-infinite impulse response (FD-IIR) filter design is a complex, multi-modal optimization problem.
• Conventional gradient-based methods are inefficient for FD-IIR filter design.
• Meta-heuristic algorithms offer a promising alternative for optimizing such complex systems.

Purpose of the Study:

• To apply the cuckoo search algorithm (CSA), a novel global meta-heuristic optimization technique, for designing FD-IIR filters.
• To determine the optimal coefficients for FD-IIR filters to achieve ideal frequency response characteristics.
• To evaluate the performance of CSA in FD-IIR filter design compared to established evolutionary algorithms.

Main Methods:

• Utilized the cuckoo search algorithm (CSA) to determine optimal coefficients for FD-IIR filters.
• Employed a weighted least square (WLS) based fitness function to enhance design performance.
• Designed FD-IIR filters of various orders using the CSA approach.

Main Results:

• The CSA-based approach demonstrated superior performance compared to Genetic Algorithm (GA) and Particle Swarm Optimization (PSO).
• CSA exhibited faster convergence rates and achieved better optimal performance in FD-IIR filter design.
• Achieved very low absolute magnitude error (0.0037) and phase error (0.0046) for a 5th order FD-IIR filter.

Conclusions:

• The cuckoo search algorithm (CSA) is highly effective for designing FD-IIR filters, outperforming GA and PSO.
• CSA offers significant improvements in both convergence speed and the accuracy of designed FD-IIR filters.
• The proposed CSA-based method provides a robust and efficient solution for complex FD-IIR filter design challenges.