An Agent-Based Network Resource Management Concept for Smart City Services

AutorDoruk Sahinel, Cem Akpolat, Fikret Sivrikaya, Sahin Albayrak
QuelleIEEE/IFIP Wireless On-demand Network systems and Services Conference (WONS) 

Massive data generation with the introduction of IoT devices and new technologies in smart cities require a flexible network management architecture to meet the dynamic demands of smart city services. In Intelligent Framework for Service Discovery and Composition (ISCO) project, we envision a smart city network operator that is tenant to multiple smart city services, and it can provide network resources and network functions on-demand to satisfy the connectivity requirements of these services. To explain this structure, we created an augmented sightseeing scenario that reflects various smart city service requirements. Then, we give an insight into the technological enablers of ISCO network management platform and explain how these technologies can satisfy the needs of the smart city scenario. Finally, we briefly present our ISCO network architecture and network resource management approach, which uses a distributed game-theory based learning algorithm to solve a multi-agent decision problem, and discuss how autonomous network management concept can be used to optimize resource allocation in our smart city communication platform.