Abstract—This paper investigates a novel modification to the Finite Radon Transform (FRAT) for Orthogonal Frequency Division Multiplexing (OFDM) systems, aiming to enhance bit error rate (BER) and peak-to-average power ratio (PAPR) performance while reducing computational complexity. Conventional FRAT, constructed with (p×p) matrices where p is prime, has exhibited superior performance at lower matrix dimensions. To capitalize on this observation and streamline the processing, we introduce a centralized signal constellation and replace the conventional one-dimensional and two-dimensional Fourier transforms, along with the requisite reordering stage, with a computationally efficient row-column-diagonal summation of the input (p×p) data matrix. Simulation results demonstrate significant improvements in BER performance and PAPR. Specifically, the lowest order FRAT matrix (p=3) achieves a 6 dB signal-to-noise ratio (SNR) gain at a BER of 10−5 compared to the original FRAT. Furthermore, its BER performance approaches that of Quadrature Phase Shift Keying (QPSK) and surpasses that of 16-Quadrature Amplitude Modulation (16-QAM) and 64-QAM. Notably, the proposed modification yields a substantial 14 dB PAPR reduction relative to the original FRAT, exhibiting only a marginal 1 dB difference compared to standard M-QAM mappers. The efficacy of this modified FRAT within an OFDM-based WiMAX system is validated through MATLAB/Simulink simulations employing the realistic time-variant Code Division Testbed (CODIT) channel model.