Three-Phase Pulse Train Assisted Sampling for Wideband Signal Reconstruction: A MATLAB-Based Comparative Study

Nandani Choudhary

Abstract: This study presents a MATLAB-based framework for wideband signal reconstruction using three pulse trains shifted by 120?. Random binary sequences are converted into pseudo-random binary sequences and interpolated to generate continuous wideband waveforms. Three phase-shifted pulse trains perform multi-branch sampling through peak detection, and the extracted samples are reconstructed and combined to estimate the original signal. The approach is evaluated for pulse density parameters N = 15, 90, and 150 using mean squared error, root mean squared error, maximum error, and correlation coefficient. The results demonstrate that increasing pulse density improves reconstruction fidelity, reduces average reconstruction error, and increases similarity with the original waveform to approximately 98.6%, confirming the effectiveness of the proposed multi-branch sampling architecture.

Keywords: Wideband Signal Reconstruction, Multi-Branch Sampling, Pulse Train Sampling, Pseudo-Random Binary Sequence, Peak Detection, MATLAB Simulation, Signal Processing