A longitudinal approach to measuring the impact of mobility on low-latency anonymity networks

The increasing mobility of Internet users is becoming an emerging issue for low-latency anonymity networks such as Tor. The increase in network churn, generated by a growing mobile client base recycling connections, could impact maintaining the critical balance between anonymity and performance. New combinatorial approaches for measuring both anonymity and performance need to be developed in order to identify critical changes to the network dynamics, and trigger intervention if and when required. We present q-factor, a novel longitudinal approach to measuring anonymity and performance within highly dynamic environments. By modelling q-factor, we show that the impact of mobility, over time, on anonymity is significant. However, by using q-factor, we are able to anticipate and significantly reduce the number of these critical events occurring. In order to make more effective strategic design and/or real-time network decisions in the future, low-latency anonymity networks will be required to adopt an even more proactive approach to network management. The potential impact from increasing mobile usage needs to be considered, as what may initially be perceived as a good solution, may in fact degrade, or in the worst case could destroy the anonymity of users over time.