Please note: This presentation will take place in DC 2310 and online.
Bohan Zhang, Master’s candidate
David R. Cheriton School of Computer Science
Supervisor: Professor Raouf Boutaba
Over the past 45 years, cellular networks have evolved from a luxury service for the few to an indispensable necessity for everyone, giving rise to countless new applications. This remarkable technological advancement has fundamentally transformed everyone’s life. The latest generation, 5G, promises unprecedented network performance in terms of latency, throughput, and subscriber density, as well as groundbreaking new services like autonomous driving, remote surgery, smart cities, and ubiquitous coverage. However, numerous challenges exist in achieving these promises, with one of the most significant being the signalling storm.
A signalling storm occurs when the volume of control signals exceeds the network’s processing capacity, leading to service disruptions. During these disruptions, retries from the users and network entities amplify the problem, creating a snowball effect. This persistent threat, present since the era of 3G, remains a major concern in 5G networks. Over the past decade, signalling storms have caused severe outages worldwide, affecting millions of lives, resulting in financial losses amounting to millions of dollars, and millions of subscriber hours lost.
With the transition from 4G to 5G, network elements have been virtualized and assigned more granular tasks, significantly increasing their numbers and the number of signals required to complete system procedures. From the subscriber’s perspective, the enriched diversity of application scenarios and the increase in subscriber numbers have exponentially raised the complexity of network resource management and orchestration. Notably, the massive globally roaming cellular IoT devices are susceptible to being compromised to execute low-cost Distributed Denial-of-Service (DDoS) attacks on operators worldwide. Due to the significant losses in the past, governments, operators, and academia consider signalling storms a major threat to 5G.
This thesis investigates past cases and numerous industry white papers, research articles, and 3GPP standards, summarizing the causes, scenarios, and mitigation solutions for signalling storms in 5G. Following these investigations, we propose a blockchain-assisted 5G Authentication and Key Agreement protocol to mitigate registration signalling storms. We also introduce a secure and efficient group handover protocol to address handover signalling storms in 5G Low Earth Orbit satellite non-terrestrial networks (NTN). Furthermore, we design a signalling load-aware conditional handover to reduce signalling peak in 5G NTN. Network capacity has always been a driving force behind the evolution of cellular networks. We hope this work will shed light on the future design of 5G and beyond cellular networks.
To attend this presentation in person, please go to DC 2310. You can also attend virtually through Zoom.