This paper [1] introduces factor graphs and describes the sum-product algorithm, which is a generic message-passing algorithm operating on factor graphs. The algorithm computes various marginal functions associated with the global function.
This algorithm is very powerful, in fact, a surprisingly wide variety of algorithms developed in the artificial intelligence, signal processing, and digital communications communities can be seen as specific instances of the sum-product algorithm, operating in an appropriately chosen factor graph. Some examples are the forward/backward algorithm, the Viterbi algorithm, Pearl’s belief propagation algorithm, the iterative turbo decoding algorithm, the Kalman filter, and even certain FFT algorithms.
The task of the student is to learn and understand factor graphs and the sum-product algorithm. The student can then relate and analyze other known algorithm under the framework of the sum-product algorithm.
During the internship the student investigates spiking neural networks to perform equalization and demapping at the receiver side of an intensity-modulation / direct-detection communication system.
Diedolo, F.; Böcherer, G.; Schädler M.; Calabrò S.: Nonlinear Equalization for Optical Communications Based on Entropy-Regularized Mean Square Error. European Conference on Optical Communication (ECOC), 2022 more…BibTeX
Schädler, M.; Böcherer, G.; Diedolo, F.; Calabrò, S.: Nonlinear Component Equalization: A Comparison of Deep Neural Networks and Volterra Series. European Conference on Optical Communication (ECOC), 2022 more…BibTeX
2021
Böcherer, G.; Diedolo, F.; Pittala, F.: Label Extension for 32QAM: The Extra Bit for a Better FEC Performance-Complexity Tradeoff. 2020 European Conference on Optical Communications (ECOC), IEEE, 2021 more…BibTeX
Full text (
DOI
)