
doi: 10.21236/ada559183
handle: 11299/215866
Tor's network congestion and performance problems stem from a small percentage of users that consume a large fraction of available relay bandwidth. These users continuously drain relays of excess bandwidth, creating new network bottlenecks and exacerbating the effects of existing ones. Attacking the problem at its source, we present the design of three new algorithms that throttle clients to reduce network congestion and increase interactive client performance. Unlike existing techniques, our algorithms adaptively adjust throttling parameters given only information local to a relay. We implement our algorithms in Tor and compare significant client performance benefits using network-wide deployments of our algorithms under a variety of network loads. We also analyze the effects of throttling on anonymity and compare the security of our algorithms under adversarial attack. Software patches for our algorithms will be submitted to Tor.
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 6 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Average | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Average | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Average |
