<p>This article suggests an improved technique for the design of a two-channel linear phase filter bank also known as quadrature mirror filter (QMF) bank. In the present work, a new hybrid sparse particle swarm optimization (HSPSO) algorithm to design a QMF bank is proposed. An updated fitness function is also presented which enhances the performance of QMF bank. The fitness function is based on weighted error minimization from the transition band, constrained ripples in the passband and stopband, measure of ripples, and deviation from perfect reconstruction condition. The QMF bank is also designed using particle swarm optimization (PSO), hybrid of particle swarm and gravitational search optimization (PSOGSA), and sparse particle swarm optimization (SPSO) algorithms. Each optimization algorithm's performance is contrasted. The proposed HSPSO provides the best results for different order filter banks with minimal execution time. The presented work is also compared to previously published QMF bank design methodologies and a substantial improvement has been made in several performance metrics. An efficient architecture for QMF bank using linear phase symmetric prototype FIR filter with coefficients represented in CSD number system utilizing adder tree structure is proposed. The proposed architecture for the FIR prototype filter is compared with existing architectures and provides an average reduction in slice, power, and delay are 39.19%, 58.41%, 51.51%, and average improvement of 8.35% in throughput, respectively. Further, the proposed architecture of QMF banks is implemented on Basys 3 FPGA Board and slice utilization, delay, and power results are reported.</p>