Faculty: Business and Law
Supervisors: Prof Emanuele Giovannetti; Prof Yonghong Peng
Location: Cambridge
Apply online by 16 March 2025This PhD project explores the role of Internet Exchange Points (IXPs) and their evolving dynamics in a world increasingly shaped by hyperscalers like Google, Amazon Web Services (AWS), and Meta.
IXPs are critical nodes in the global internet ecosystem, providing local traffic exchange and reducing dependency on costly international bandwidth. Their role is particularly significant in developing countries, where efficient IXPs connectivity can foster local content delivery, improve latency, and reduce costs, addressing aspects of the digital divide.
Despite their potential, IXPs in developing regions face challenges such as underutilisation, poor interconnection practices, and limited participation from local networks. These barriers are further complicated by increasing market concentration from global “hyperscalers”, which dominate peering, reshape interconnectivity economics, and operate global infrastructure to provide computing, storage, and networking services.
These companies achieve operational efficiency and scalability by designing and managing their hardware, software, and data centres to handle enormous workloads and serve billions of users while benefiting from their personal data, for example, by better training their algorithms, achieving higher profiling of their services, further increasing their attraction to a larger customer base.
By operating a vast network of data centres interconnected worldwide, hyperscalers provide services with minimal latency and high reliability. They usually own and manage their infrastructure, including servers, networking equipment, and software platforms. Their operations rely heavily on automation, machine learning, and custom-built solutions to optimise performance and minimise operational costs. This allows them to provide personalised cloud computing services , AI, big data analytics, and content delivery networks (CDNs).
Disparities in access exacerbate the difficulties faced by smaller networks. Hence, while hyperscalers enhance global connectivity and service accessibility, their dominance may lead to tipping processes of increasing market power, often reinforced by access and usage of personal data, as a key input for training their AI-driven offerings, affecting competition and inclusivity in IXP ecosystems and risking to reverse the global processes of digital inclusion and development by hampering non-hierarchical peering structures at IXPs and reducing their potential to enable Internet access in underserved regions, fostering digital equality.
Emerging innovations, such as edge computing, alter traffic flows, creating new economic dynamics for local networks and communities. These, in turn, may provide a better way to decentralise traffic exchanges, increasing their resilience and inclusivity. However, if these technologies become dominated by existing hyperscalers, they might be used to further leverage market power across different segments.
These emerging trends in market power dynamics, fed by network externalities and technological developments, have the potential to change the interconnectivity landscape, especially in developing countries. Moreover, hyperscalers participation at IXPs can reshape the economics and topology of interconnectivity by driving higher traffic volumes at IXPs, influencing pricing models and peering terms for smaller networks, and shifting traffic flows through private interconnections or proprietary CDNs.
This research will analyse, from both economic and data measurement perspectives:
The study will develop strategies to optimise IXP performance, foster equitable peering ecosystems, and mitigate the risks of market concentration. Insights from this research will provide policy recommendations for global and regional Internet governance, helping to ensure that the Internet remains a force for inclusion and equality.
The research will benefit from resources provided by the Internet Society (ISOC), including access to datasets, collaborations with IXP operators, and policy forums. Specialised software for graph-theoretic and econometric analysis will also be available, with expert guidance from the supervisory team.
The project will contribute novel metrics for evaluating IXP connectivity, provide insights into market power dynamics, and propose sustainable peering models. These findings will be disseminated through high-impact journals, international conferences, and partnerships with international organizations like ISOC and the International Telecommunication Union.
If you would like to discuss this research project, please contact Prof Emanuele Giovannetti: [email protected]
Apply online by 16 March 2025The successful applicant for this project will receive a Vice Chancellor’s PhD Scholarship which covers the tuition fees and provides a UKRI equivalent minimum annual stipend for three years. For 2024/5 this was £19,237 per year. The award is subject to the successful candidate meeting the scholarship terms and conditions. Please note that the University asserts the right to claim any intellectual property generated by research it funds.
Download the 2024/5 terms and conditions (2025/6 terms and conditions TBC)