Section: New Results

Network data exploitation

Participants: Panagiota Katsikouli, Elli Zavou, Stéphane D'Alu, Hervé Rivano, Razvan Stanica.

Spatio-temporal Characterization of Mobile Data Traffic

Mobile traffic data collected by network operators is a rich source of information about human habits, and its analysis provides insights relevant to many fields, including urbanism, transportation, sociology and networking. Urban landscapes present a variety of socio-topological environments that are associated to diverse human activities. As the latter affect the way individuals connect with each other, a bound exists between the urban tissue and the mobile communication demand. In [3], we investigate the heterogeneous patterns emerging in the mobile communication activity recorded within metropolitan regions. To that end, we introduce an original technique to identify classes of mobile traffic signatures that are distinctive of different urban fabrics. Our proposed technique outperforms previous approaches when confronted to ground-truth information, and allows characterizing the mobile demand in greater detail than that attained in the literature to date. We apply our technique to extensive real-world data collected by major mobile operators in ten cities. Results unveil the diversity of baseline communication activities across countries, but also evidence the existence of a number of mobile traffic signatures that are common to all studied areas and specific to particular land uses.

Similarly to mobile phone data, GPS traces of vehicles convey information on transportation demand and human activities that can be related to the land use of the neighborhood where they take place. In [10], we investigate the land use patterns that emerge when studying simultaneously GPS traces of probe vehicles and mobile phone data collected by network providers. To this end, we extend previous definitions of mobile phone traffic signatures for land use detection, so as to incorporate additional information on human presence and mobility conveyed by GPS traces of vehicles. Leveraging these extended signatures, we exploit an unsupervised learning technique to identify classes of signatures that are distinctive of different land use. We apply our technique to real-world data collected in French and Italian cities. Results unveil the existence of signatures that are common to all studied areas and specific to particular land uses. The combined use of mobile phone data and GPS traces outperforms previous approaches when confronted to ground-truth information, and allows characterizing land use in greater detail than in the literature to date.

The spatial and temporal profiles of mobile phone traffic can be studied simultaneously. In [11], we present an original approach to infer both spatial and temporal structures hidden in the mobile demand, via a first-time tailoring of Exploratory Factor Analysis (EFA) techniques to the context of mobile traffic datasets. Casting our approach to the time or space dimensions of such datasets allows solving different problems in mobile traffic analysis, i.e., network activity profiling and land use detection, respectively. Tests with real-world mobile traffic datasets show that, in both its variants above, the proposed approach (i) yields results whose quality matches or exceeds that of state-of-the-art solutions, and (ii) provides additional joint spatiotemporal knowledge that is critical to result interpretation.

Using Mobile Phone Data in Cognitive Networking

In the next few years, mobile networks will undergo significant evolutions in order to accommodate the ever-growing load generated by increasingly pervasive smartphones and connected objects. Among those evolutions, cognitive networking upholds a more dynamic management of network resources that adapts to the significant spatiotemporal fluctuations of the mobile demand. Cognitive networking techniques root in the capability of mining large amounts of mobile traffic data collected in the network, so as to understand the current resource utilization in an automated manner. In [4], we take a first step towards cellular cognitive networks by proposing a framework that analyzes mobile operator data, builds profiles of the typical demand, and identifies unusual situations in network-wide usages. We evaluate our framework on two real-world mobile traffic datasets, and show how it extracts from these a limited number of meaningful mobile demand profiles. In addition, the proposed framework singles out a large number of outlying behaviors in both case studies, which are mapped to social events or technical issues in the network.

Study of Wi-Fi Localization from Crowdsourced Datasets

The wide adoption of mobile devices has created unprecedented opportunities to collect mobility traces and make them available for the research community to conduct interdisciplinary research. However, mobility traces available in the public domain are usually restricted to traces resulting from a single sensor (e.g., either GPS, GSM or WiFi). In [26], we present the PRIVA'MOV dataset, a novel dataset collected in the city of Lyon, France on which user mobility has been collected using multiple sensors. More precisely, this dataset contains mobility traces of about 100 persons including university students, staff and their family members over 15 months collected through the GPS, WiFi, GSM, and accelerometer sensors. We provide both a quantitative and a preliminary qualitative analysis of this dataset. Specifically, we report the number of visited points of interests, GSM antennas and WiFi hotspots and their distribution across the various users. We finally analyse the uniqueness of human mobility by considering the various sensors.

Thanks to this collected data, it is possible to combine information from several probes. A very common use case is the collection of network scans with location to help the localisation feature of these devices. Nevertheless, most users are not aware of this spying. The collected data might represent infringements of privacy. One possible solution to keep gathering these data while maintaining privacy would consist in device-to-device communications in order to break the links between data and users. In [22], we propose an approach to test the feasibility of such a system. We collected data from mobile users to combine location and network scans data. With this data, we test the accuracy level we can reach while using Wi-Fi localisation. We analyse how a new measure should be pushed and how many scans should be realised to provide location-based Wi-Fi. We analyse the minimal dataset to cover the set of locations covered by users and prove that a multiuser gathering system can benefit the users.