NANO: Network Access Neutrality Observatory

This paper tackles a technical problem that is of growing interest in light of the ongoing network neutrality debate: We aim to develop a system that can reliably determine whether particular ISP is discriminating against a service using only passive measurements from end-hosts. This problem presents significant challenges because many types of discrimination can often resemble commonplace performance degradations (e.g., resulting from failure or misconfiguration). To distinguish discrimination from degradation, we propose a statistical method to estimate causal effect and develop a system, NANO, based on this method. NANO aggregates passive measurements from end-hosts, stratifies the measurements to account for possible confounding factors, and distinguishes when an ISP is discriminating against a particular service or group of clients. Preliminary simulation results demonstrate the promise of NANO for both detecting various types of discrimination and absolving an ISP when it is not discriminating.