Section: New Results

Capillary Network Solutions

Participants: Patrice Raveneau, Trista Lin, Marco Fiore, Hervé Rivano, Razvan Stanica.

Managing user mobility is historically one of the most critical issues in cellular radio access networks (RANs). That task will become an even greater challenge due to cellular users on-board vehicles and networked cars that autonomously access Internet-based services, whose number is expected to grow dramatically in the next few years. There is thus a need to characterize RAN access from/by vehicles in a similar way to what has been done for traditional pedestrian access. In [11] , we proposed a first study of the macroscopic and microscopic features of pervasive vehicular access in a case-study large-scale urban environment, in presence of realistic datasets of the road traffic and RAN deployment. We found that pervasive vehicular access is characterized by unique temporal and spatial variability in the urban region, such that it may require a dedicated RAN capacity planning: the presence of stable vehicular access load patterns and mobility flows can help to that end. Also, we identified the theoretical distributions that best fit key metrics for RAN planning, i.e., the vehicular users’ inter-arrival and residence times at cells, and discuss how their parameters vary over time and space.

Smart parking, allowing drivers to access parking information through their smart-phone, is another important service for vehicular users, which can be provided not only through cellular networks, but also by using metropolitan wireless networks, whose deployment strategy needs to be guided by efficiency and functionality. In [8] , we introduced and studied a deployment strategy for wireless on-street parking sensor networks. We defined a multiple-objective problem in our analysis, and solved it with two real-world street parking maps. We presented the results on the tradeoff among minimum energy consumption, sensing information delay and the amount of deployed mesh routers and Internet gateways, i.e., the cost of city infrastructure. We also analyzed these tradeoffs to see how different urban layouts affect the optimal solutions. The overall smart parking architecture and services made the object of the PhD thesis of Trista Lin [1] , where the analysis of the entire system can be found, including results on the wireless sensor networks used to collect data from parking places and the Publish-Subscribe service used to disseminate this information to users.

Offloading is a promising technique for alleviating the ever-growing traffic load from infrastructure-based networks such as the Internet. Offloading consists in using alternative methods of transmission as a cost-effective solution for network operators to extend their transport capacity. Wi-Fi offloading is one of the most effective approaches to relieve the cellular radio access from part of the burgeoning mobile demand. To date, Wi-Fi offloading has been mainly leveraged in limited contexts, such as home, office or campus environments. In [18] , we investigated the scaling properties of Wi-Fi offloading, by studying how it would perform on a much larger scope than those considered today. To that end, we considered a real-world citywide scenario, built on data about actual infrastructure deployments and mobile traffic demand, and observed which amount of traffic could be accommodated by the existing pervasive Wi-Fi access infrastructure, were it opened to mobile users. We found that more than 80% of the mobile traffic demand in a large urban area may be easily served by Wi-Fi access points, under a wide range of system settings.

A new offloading technique was introduced in [20] and further detailed in [4] , where we advocate the use of conventional vehicles equipped with storage devices as data carriers whilst being driven for daily routine journeys. The road network can be turned into a large-capacity transmission system to offload bulk transfers of delay-tolerant data from the Internet. The challenges we addressed include how to assign data to flows of vehicles and while coping with the complexity of the road network. We proposed an embedding algorithm that computes an offloading overlay where each logical link spans over multiple stretches of road from the underlying road infrastructure. We then formulated the data transfer assignment problem as a novel linear programming model we solve to determine the optimal logical paths matching the performance requirements of a data transfer. We evaluated our road traffic allocation scheme using actual road traffic counts in France. The numerical results show that 20% of vehicles in circulation in France equipped with only one Terabyte of storage can offload Petabyte transfers in a week.