Paul Patras

The number of Internet-connected devices is projected to reach one trillion by 2035 and new applications with huge societal impact potential are emerging. These include home automation, assisted-living robotics, self-driving cars, and countless more.

Such applications are placing unparalleled demands on the mobile networks that underpin them. Although wireless technology advances facilitate increased data rates, these are not enough. Extreme throughput, latency, and reliability requirements cannot be fulfilled solely through denser network deployments. New tools for precision traffic engineering and optimal configuration of the infrastructure are mandatory, in order to guarantee pervasive high-speed connectivity. These are challenging tasks, because they require to elucidate wide-scale mobile user behaviour and how to manage resources in real-time, in ever more complex settings.

Many of the Internet of Things (IoT) devices will be in future personal living spaces, improving their users’ comfort and quality of life, and reducing the effort placed on care providers. As millions of new devices go online monthly, detecting security hazards and protecting user safety and privacy become increasingly difficult. There is an urgent need to design defensive methods for threat detection, counteraction, and privacy preservation, which can advance state-of-the-art techniques for cyber resilience.

To address these challenges and to overcome the looming network performance and cyber security crisis, requires expert knowledge at the intersection of networked systems, artificial intelligence (AI), and security, which I can provide. My team and I have been pioneering applications of AI to the mobile networking domain, to improve the analysis, resilience, and management of hyper-connected systems. My research is one slice of the big pie of expertise available at the University of Edinburgh, which combined can fuel the development of future digital technologies capable of supporting pervasive services with a trillion-pound economic output.