Dejan Samardzhiev

• BSc (Hons) Computer Science (Software Engineering)
• Autonomous Vehicle Navigation with Collaborative Diffusion for City Congestion Control

In recent years traffic control is becoming a world-wide problem. Many old fashion solutions include building smarter roads, and bigger parking spaces, but in recent years the growing field of autonomous vehicles presents many new opportunities for improving traffic. This paper focuses on method to control traffic and decrease the average journey times by implementing collaborative diffusion.

Collaborative diffusion is method in which agents, autonomous vehicles, communicate between each other their next goal destinations and make decisions so as to minimise their average journey time for reaching their end locations.

The technique is implemented and explored in a simulated environment which is based on the densely populated traffic network of the Newcastle upon Tyne city network. The tests are compared to other implementations of algorithms for traffic control and the results are shown to be substantially more efficient.

Funding source: Newcastle University

Project supervisor: Dr Gary Ushaw