In this repository we design and simulate some Deep Learning-Based End-to-End Wireless Communication Systems. The following methods have been proposed :
- End-to-End wireless communication system using conditional GAN as unknown channel over AWGN channel
- End-to-End communication system over AWGN channel while the channel is known(No GANs for modeling the channel)
- End-to-End communication system over Rayleigh channel while the channel is known(No GANs for modeling the channel)
- Hamming coding and decoding with Maximum Likelihood Decoder in Rayleigh Fading channel
- Hamming coding and decoding with Maximum Likelihood Decoder in AWGN channel
- ...
Simulation Results :
The following figures compares learning based methods with some of the SOTA traditional communication systems over 3 communication channels :
- AWGN
- Rayleigh Fading
- Frequency selective multipath
It can be understood that the conditional GAN has been able to comprehend the distribution of the corresponding communication channel and even surpasses some of traditional methods.
