Joint Source Channel Coding

Joint source-channel coding (JSCC) aims to optimize data compression and channel coding simultaneously, unlike traditional separate approaches. Current research heavily utilizes deep learning, employing architectures like variational autoencoders, transformers, and generative adversarial networks to achieve efficient and robust communication, particularly for image and video transmission and federated learning applications. This integrated approach shows promise in improving performance, especially in scenarios with limited bandwidth or high noise, leading to more efficient and reliable communication systems for various applications including autonomous driving and semantic communication. The focus is on achieving high fidelity with low latency, often prioritizing task-oriented performance over perfect signal reconstruction.