Incentives for Prefix Deaggregation in the Internet

2 Mar

Andra Lutu, Research Assistant, Institute IMDEA Networks

In-house Presentation

The scalability issues the global routing system has been experienc- ing over the past years have raised serious concerns in the Internet community. One of the main culprits for the rapidly growing BGP routing table is the defragmentation process of the address blocks allocated to ASes, also known as prefix deaggreation. The Internet is a complex system and understanding its behavioural evolution is certainly a challenging task. For this reason, the use of eco- nomic models can provide intuitive explanations of the complex interactions between networks that result in the aforementioned phenomenon of prefix deaggreation. In this paper, we propose a game theoretic model to analyze the incentives behind the deag- gregating strategies of the networks. Announcing more-specific prefixes in the Internet impacts the size of the global routing table and increases the network operators’ capital expenditure for rout- ing equipment capable of sustaining the growing Internet. How- ever, we prove that the ASes are driven by fundamental economic reasons towards this type of apparently harming behaviour. We find that by announcing more prefixes the originating network achieves a more predictable traffic pattern and reduces the peak levels of bandwidth consumption. We show that this happens because, both in the equilibrium point and the social welfare point, the cost re- duction achieved by smoothing the traffic distribution outweighs the additional cost incurred by the routing table expansion.

The conference will be conducted in English


  • Location: Aula 4. 1F03, Departamento de Telemática, Edificio Torres Quevedo, Universidad Carlos III de Madrid, Avda. Universidad, 30, 28911 Leganés – Madrid 

  • Organization: NETCOM Research Group (Telematics Department, University Carlos III of Madrid, Spain); IMDEA Networks (Madrid, Spain)
  • Time: 01:00 pm
  • Add to Calendar: iCalendar Outlook Google