Impact of prefix hijacking on payments of providers

Abstract

Abstract-Whereas prefix hijacking is usually examined from security perspectives, this paper looks at it from a novel economic angle. Our study stems from an observation that a transit AS (Autonomous System) has a financial interest in attracting extra traffic to the links with its customers. Based on real data about the actual hijacking incident in the Internet, we conduct simulations in the real AS-level Internet topology with synthetic demands for the hijacked traffic. Then, we measure traffic on all inter-AS links and compute the payments of all providers. The analysis of our results from technical, business, and legal viewpoints suggests that hijacking-based traffic attraction is a viable strategy that can create a fertile ground for tussles between providers. In particular, giant top-tier providers appear to have the strongest financial incentives to hijack popular prefixes and then deliver the intercepted traffic to the proper destinations. We also discuss directions for future research in the area of hijacking-based traffic attraction.

Collaborative Verification of Forward and Reverse Reachability in the Internet Data Plane

“…5 We found COVE to be scalable to very large ASes in the Internet. COVE was designed to improve a network's views of forward and reverse reachability in the Internet data plane even if COVE is partially deployed by as few as two networks connected by a link.…”

“…Once there is a set of interconnected networks that deploy COVE, there is strong incentive for another network to join and for the set to accept a new member to increase every network's views of data plane reachability. We also illustrated how network operators can use COVE to 5 Our dataset is avalable from www.cs.utexas.edu/users/lam/Internet dataset/ perform very useful network management tasks which cannot be done effectively using existing methods and tools.…”

“…Numerous papers have been published on detecting prefix hijacking using active probes in the data plane as well as control plane data. The reader is referred to two recent papers on this topic [5], [24] and references therein.…”

T4P: Hybrid interconnection for cost reduction

“…Although tussles affected the Internet from its early days [22], this growing diversity together with the ubiquitous peering resulted in frequent disagreements over peering cost allocations. Traffic imbalances [23] in peering relationships led to demands of monetary compensations [2] and to a large set of techniques for minimizing costs and maximizing revenues [24]- [26]. The tensions were specially acute between networks that primarily connect residential users and those networks that connect content providers.…”

Economics of traffic attraction by transit providers

“…In contrast, our work evaluates traffic attraction by intermediary ASes that seek to increase their transit revenues. While Bangera and Gorinsky [14] simulate the YouTube prefix-hijacking incident to assess its economic implications, our investigation considers multiple attractors from different transit tiers and evaluates novel countermeasures, such as multi-stage filtering and disconnection. Goldberg et al [10], Kalogiros et al [11], and Levin et al [13] present game-theoretic investigations of economic incentives in inter-domain routing.…”