This is a non-IMPACT record, meaning that access to the data is not controlled by IMPACT. For access, see the directions below.

This Resource is offered and provided outside of the IMPACT mediation framework. IMPACT and the IMPACT Coordination Council/Blackfire Technology, Inc. expressly disclaim all conditions, representations and warranties including but not limited to Resource availability, quality, accuracy, non-infringement, and non-interference. All Resource information and access is controlled by entities and under terms that are external to the IMPACT legal framework.


The kyutech/throughput dataset
External Dataset
External Data Source
Crawdad (Dartmouth)
56 (lowest rank is 56)

Category & Restrictions



Measurements on real HTTP throughput via several 3G/LTE providers in Japan.

While recent LTE provider promise very high rates, end users often complain that HTTP throughput is not sufficient for rich multimedia, even as non-demanding as YouTube. It is interesting that this artifact can be found in both 3G and LTE providers. For example, it is not uncommon for a YouTube video to freeze on a terminal with LTE connection. The main cause of such artifacts is the congestion due to high and further increasing number of subscribers. Congestion can happen at several points inside a given 3G/LTE provider, but it is not important to end user because content is commonly found outside of 3G/LTE providers, guaranteeing that all subscribers have to do through the bottleneck. This dataset contains measurements conducted on 3 (anonymized) 3G/LTE providers in Japan in April-May 2013. The measurement design was very simple -- a webapp running on 3 terminals (that were tweaked not to go to sleep and not used for anything else) would regularly send web requests to a web server placed within a university campus and measure the round-trip time of the request. Bulk size was randomly selected from a number of preset values and was always downloaded (HTTP get request with bulk returning in reply). It was verified that the wired part of the e2e connection was never the bottleneck, implying the throughput was always impaired within a given 3G/LTE provider.


Additional Details

throughput, dataset, the kyutech/throughput dataset, kyutech, 996, inferlink, external data source, source, external, corporation, inferlink corporation, lte, 3g, providers, http, measurements, japan, real, provider, bulk, connection, youtube, request, web, congestion, subscribers, bottleneck, design, guaranteeing, simple, april, artifact, regularly, freeze, trip, uncommon, sleep, running, measurement, rates, verified, main, user, anonymized, promise, randomly, time, 2013, increasing, returning, multimedia, terminal, demanding, rich, content, preset, requests, users, maratishe, downloaded, terminals, complain, tweaked, reply, conducted, selected, artifacts, campus, happen, webapp, measure, video, values, other, implying, wired, impaired, server, gmail, sufficient, send, e2e, size, commonly