From routine to network deployment for data offloading in metropolitan areas

Abstract

This paper tackles the WiFi hotspot deployment problem in a metropolitan area by leveraging mobile users' context and content, i.e., their trajectories, scenario interactions, and trac demands. The careful deployment of hotspots in such areas allow to maximize WiFi ooading, a viable solution to the recent boost up of mobile data consumption. Our proposed strategy considers the restrictions imposed by transportation modes to people trajectories and the space-time interaction between people and urban locations, key points for an ecient network planning. Using a real-life metropolitan trace, we show our routine-based strategy guarantees higher ooad ratio than the current approach in the literature while using a realistic trac model.

Quality of experience and access network traffic management of HTTP adaptive video streaming

“…Reference [298] scanned available governmental and private Wi-Fi hotspots in Paris, France, and found a high potential for Wi-Fi oloading. Reference [299] used user mobility, interaction, and tra c demands to nd a deployment of hotspots in metropolitan areas, which enables a higher oloading ratio. Similarly, [300,301] studied hotspot deployment, but additionally considered also battery levels.…”

On the Sampling Frequency of Human Mobility

“…• Active probing of subscribers' positions in mobile networks. Precise knowledge of subscribers' locations is a valuable information for mobile operators, for both network management and value-added service development [17,6]. Actively probing mobile devices for their position in a country-scale mobile network is a computationally expensive task, which has traditionally pushed operators to favor less controllable passive measurements [5].…”

Measurement-driven mobile data traffic modeling in a large metropolitan area

“…The studied dataset contains more than 1 billion sessions and each of them has the following information fields: (1) amount of upload and download volumes (in KiloBytes) during the session; (2) session duration in seconds; and (3) timestamp indicating when the session starts. Due to the routine behavior of people [1] and the large scale of the dataset, it suffices to study a subset of the whole dataset in order to capture the daily behavior of subscribers. Indeed, our analysis shows that there is low variability on subscribers' activity among the same hours on different days.…”

“…Emerging pervasive communication systems will thus face a number of challenges, including the need to operate in such extreme environments. In this context, understanding mobile data traffic demands per user is crucial for the evaluation of data offloading solutions designed to alleviate cellular networks [1], thus favoring the proliferation of pervasive communication. Moreover, the definition of a usage pattern can allow telecommunication operators to better timely plan network resources allocation and better set subscription plans.…”