Intelligent, Energy Efficient and Secure Tactile Communication using Federated Learning in 6G Network

About the Project

In 5G, ultra-reliable low-latency communication has been a backbone for various applications that need an extremely low delay. However, further research will be required to minimize the round trip delay and improve energy efficiency and security of user data in the 6G network [1]. In the proposed haptic/tactile communication, we may need “Network Edge Intelligence” where we build a digital twin in the edge server near the controller. A digital twin is a simulation environment that mimics the behaviour of a real-world system. Moreover, the proposed system should be intelligent enough which can autonomously and jointly predict and optimize the 6G network parameters. Furthermore, network edge intelligence needs to be integrated where intelligent remote sensors or IoT devices may make decisions on the local machines and send the data to a gateway for further screening before sending them to the vertical processing unit. Due to the stochastic nature of wireless channels, algorithms that enable autonomous prediction of network parameters will need to be designed to achieve ultra-low latency in end-to-end applications.

The federated learning is a suitable AI method for delay-sensitive applications like tactile Internet [2]. Federated learning does not send raw data to the machine learning model, but instead brings the model to the data. The model is trained locally on each device, and the data never leaves its original location. In terms of implementation, the data of the original haptic sensors is initially collected by the respective sink node and then provided to the virtual machine associated with it. The model on this virtual machine is trained using the distributed dataset. After the models’ training, the trained models are sent to the edge server. The edge server then fuses all the models and sends the fused model back to each virtual machine. From a security point of view, this is an optimal method to train AI models because no one can see, touch or process the user data in transit.

The Tactile Internet has huge potential in the future to realise the human digital twin, remote surgery and Metaverse applications, amongst others. Such applications entail strong interactions and extremely immersive quality of service, ultra-high reliability and ultra-low latency. The objectives of this project are to investigate a unique design of an energy-efficient wireless network, AI-enabled wireless channel model and network edge intelligence to realise Tactile Internet in a 6G framework.

Applicants should have, at least, an Honours Degree at 2.1 or above (or equivalent) in Computer Science or related disciplines. In addition, they should have a good mathematics background, excellent programming skills in Python/MATLAB and an interest in machine learning and AI. Qualified applicants are strongly encouraged to contact informally the lead academic, Dr Deepak GC (email: d.gc@kingston.ac.uk), to discuss their application.

Funding Notes

there is no funding for this project