Wireless Communications and Networks  Lab (WHYNET) 

Research Topics

The basic idea of semantic communications is to extract the “meanings” or “features” of sent information from a source, and “interpret” the semantic information at a destination. Conversly, traditional communication systems aim to offer a high data transmission rate and a low symbol (bit) error rate.

Three-Level Architecture of Semantic Communication System.

2. Privacy-preserving Machine Learning (PPML)

PPML is designed to protect and preserve the privacy of data while training and deploying machine learning models. These methods allow the extraction of insights and patterns from data without exposing sensitive information, using perceptual encryption. PPML is particularly valuable in scenarios where data privacy is crucial, such as in healthcare, finance, and personal data analysis.

A taxonomy of image encryption methods based on their levels of security. From left to right, the encryption algorithms computational complexity decreases and security is traded for usability, i.e., to enable other multimedia applications such as format compliant storage and even processing the encryption domain.

3. Deep learning-based communication system design

DL-based End-to-End learning is a promising technology for future wireless communication systems. Its key idea is to implement the transmitter, channel, and receiver as a single neural network (NN). Various DL algrotihms can be leveraged to achieve end-to-end optimization over practical wireless systems such as OFDM, MIMO-OFDM, NOMA, etc.

ML implementation in Physical layer for data-driven  wireless communication.

4. Wireless Communication Systems in IoT environment

AI is revolutionizing the Internet of Things (IoT) ecosystem, particularly in the realm of wireless communication systems. By integrating ML algorithms, these advanced systems can autonomously analyze and optimize data transmission, enhancing efficiency and reliability. This enables smarter, more responsive IoT networks, where devices seamlessly adapt to changing environments and user demands.

An overview of used case scenario utilizing Edge–Cloud collaboration for deep learning-based CV applications in (a) generic IoT ecosystem and (b) Intelligent Transport System within a Smart City.

5. Network Security and Privacy Techniques

Source Location Privacy (SLP) is a security measure in wireless sensor networks that aims to protect the location of a source node from being disclosed to unauthorized entities. It is crucial in scenarios where the source's location is sensitive or confidential, such as in military or surveillance applications, preventing adversaries from tracking or targeting the source based on its transmitted signals.

Classification of Privacy in wireless sensor network (WSN)

Projects