The impact and implications of the growth in residential user-to-user traffic

Abstract

It has been reported worldwide that peer-to-peer traffic is taking up a significant portion of backbone networks. In particular, it is prominent in Japan because of the high penetration rate of fiber-based broadband access. In this paper, we first report aggregated traffic measurements collected over 21 months from seven ISPs covering 42% of the Japanese backbone traffic. The backbone is dominated by symmetric residential traffic which increased 37%in 2005. We further investigate residential per-customer trafficc in one of the ISPs by comparing DSL and fiber users, heavy-hitters and normal users, and geographic traffic matrices. The results reveal that a small segment of users dictate the overall behavior; 4% of heavy-hitters account for 75% of the inbound volume, and the fiber users account for 86%of the inbound volume. About 63%of the total residential volume is user-to-user traffic. The dominant applications exhibit poor locality and communicate with a wide range and number of peers. The distribution of heavy-hitters is heavy-tailed without a clear boundary between heavy-hitters and normal users, which suggests that users start playing with peer-to-peer applications, become heavy-hitters, and eventually shift from DSL to fiber. We provide conclusive empirical evidence from a large and diverse set of commercial backbone data that the emergence of new attractive applications has drastically affected traffic usage and capacity engineering requirements.

Can ISPs be profitable without violating “network neutrality”?

“…To model the user traffic demand, we refer to a study of residential broadband access networks in Japan [4]. That study found that the distribution of the amount downloaded by a user is heavy tailed.…”

Optimally designing caches to reduce P2P traffic

“…We assume that the ratio of P2P users among all users in each city is almost constant, and the number of users accommodated in lower network i is proportional to the population p i in the area for which the lower network i covers. In this case, we can assume that the ratio of requests generated by users accommodated in lower network i against the total requests in N lower networks is proportional to r i , which is the population covered by the lower network i divided by the total population among N lower networks [8]. In other words, r i is obtained by…”

Parallel video streaming optimizing network throughput

“…Before evaluating each video streaming system, we investigate the effect of link weights on the network productivity in this section. Assuming the gravity model on r sd [6], i.e., setting r sd = p s p d , we compare h gv , the average hop length of flows, and V max,gv , the network productivity, when using min.hop and inv.cap. Fig.…”