Optimization of transmitter and receiver filters for OQAM-OFDM systems using nonlinear programming

This paper proposes a family of optimized transmitter and receiver FIR filters for orthogonal frequency division multiplex (OFDM) systems with offset QAM modulation using nonlinear-programming. Two objective functions in the frequency domain (considering both OFDM orthogonal condition and Nyquist condition), least square error (LSE) and minimizing maximal spectral side lobe (Mini-max), are used. The nonlinear programming is implemented with a modified sequential quadratic programming (SQP) algorithm, which guarantees a super-linear convergence. Resultant optimized FIR filters are given with their coefficients and spectra.