Particle Filtering Based LDPC Decoding Algorithm for Hierarchical Modulation Schemes

In this project, we present a neural network based particle filtering scheme to enhance the LDPC scheme. This is a decoding algorithm preferred for the hierarchical broadcasting systems. Unlike the traditional hierarchical demodulation approaches that suffer from serious interlayer interference (ILI), the proposed method exploits the iterative particle filtering on the decoding section of the secondary layer to mitigate the ILI impairment, there by offering significant gains in reception performance. The algorithm can be applied to different channel coding and mapping schemes under various channel profiles. In addition, the new algorithm has been implemented on a broadband in-band–on-channel digital radio broadcasting system. Numerical results from both simulations and laboratory experiments are provided to quantify the performance advantages of the new algorithm. The performance improvement is shown at least by 2db improvement with respect to a BER of 10e-5. The simulation is performed using Matlab ver 7.12.