Inspired by the recent advances in deep learning, we propose a novel iterative BP-CNN architecture for channel decoding under correlated noise. This architecture concatenates a trained convolutional neural network (CNN) with a standard belief-propagation (BP) decoder. The standard BP decoder is used to estimate the coded bits, followed by a CNN to remove the estimation errors of the BP decoder and obtain a more accurate estimation of the channel noise. Iterating between BP and CNN will gradually improve the decoding SNR and hence result in better decoding performance.
Seminars
Seminar Series
Recent Seminars
15 Feb 2018
This talk will describe several approaches to reducing energy consumption in internet-of-things applications and applications of data analytics to neuro-psychiatric disorders. Machine learning and information analytics are important components in all these things. Almost all things should have embedded classifiers to make decisions on data. Thus, reducing energy consumption of features and classifiers is important. First part of the talk will present energy reduction approaches from feature selection, classification and incremental multi-stage classification perspectives.
10 Nov 2017
We are witnessing an unprecedented growth in the amount of data that is being collected and made available for data mining. While the availability of large-scale datasets presents exciting opportunities for advancing sciences, healthcare, understanding of human behavior etc., mining the data set for useful information becomes a computationally challenging task. We are in an era where the volume of data is growing faster than the rate at which available computing power is growing, thereby creating a dire need for computationally efficient algorithms for data mining.
- 1 of 55
- next ›