DB Kurka, D Gündüz - IEEE Journal on Selected Areas in …, 2020 - ieeexplore.ieee.org
… Inspired by the recent success of deeplearning … deep JSCC architecture that exploits the channel output feedback, called DeepJSCC-f, outperforms any separate source channelcoding …
… Inspired by the success of unsupervised deeplearning (DL) methods, in particular, the autoencoder architectures [4], [5], we design an end-to-end communication system, where the …
J Xu, B Ai, W Chen, A Yang, P Sun… - IEEE Transactions on …, 2021 - ieeexplore.ieee.org
… Abstract—Recent research on joint source channelcoding (… owing to the employment of deeplearning (DL). However, the … coding and the channelcoding rate according to the channel …
… To address these challenges we propose Neural Error Correcting and Source Trimming (NECST) codes, a deeplearning framework for jointly learning to compress and errorcorrect an …
DB Kurka, D Gündüz - 2019 IEEE 20th International Workshop …, 2019 - ieeexplore.ieee.org
… separate source and channelcoding (JPEG2000 followed by high performance channel codes) in single-layer transmission. Here we show that the advantages of deep JSCC extend to …
F Sohrabi, KM Attiah, W Yu - IEEE Transactions on Wireless …, 2021 - ieeexplore.ieee.org
… a communication system separately, the deeplearning framework can jointly design … coding design for each different scenario, the deeplearning framework implicitly learns the channel …
… Abstract—Recently, semantic communication has been brought to the forefront because deeplearning (DL)-based methods, such as Transformer, have achieved great success in …
Z Qin, H Ye, GY Li, BHF Juang - IEEE Wireless …, 2019 - ieeexplore.ieee.org
… -ChannelCoding: Typical source coding mainly utilizes the sparse property to remove the redundancy in source data while channelcoding … that source coding and channelcoding can …
M Honkala, D Korpi… - IEEE Transactions on …, 2021 - ieeexplore.ieee.org
… compatible with the channelcoding used in 5G systems. Using 3GPP-defined channel models, … We also show that the high performance can likely be attributed to DeepRx learning to …