The world and our societies are urbanizing. For the first time in 2009, more than half of world population lived in urban area, this ratio is thought to increase to 60% within 2030, and is over 75% in developed countries. Simultaneously, the environment of the citizens is progressively digitalizing with the instrumentation of the physical space, the digitalization of services and the new social networks. This is the rise of a continuum between physical and virtual spaces operated by communicating devices with one or several interfaces with the physical space (sensor, actuators, touch screens, movement detection, phones, connected vehicles…) that we call «capillary nodes », interconnected in «capillary networks ». Similarly to the web 2.0, the concept of «cities 2.0 » is emerging, exploring the possibilities of direct interaction between inhabitants and everyday management of their city. Beyond, the notion of «smart cities » is constantly progressing as the contribution of ICT in the organization of a partially automated city and build on its digital dimension. In particular, the «digital city » results of massive and thin sensing and interacting of the environment and the citizens activity, hence of the proliferation of capillary networks. UrbaNet (for « Urban Networks ») is an INRIA/INSA Lyon research team focusing on the networking problems fostered by capillary networks in urban environment.
Characterization of urban networks
The first research axis focuses on the characterization of the fundamental features of urban capillary networks. We will focus on three aspects of the network characterization, namely topology, user mobility and network traffic patterns. Our ultimate goal is to identify and calibrate relevant models, or to provide new models and traces when needed, so that they can be used for simulating urban networks scenarios.
Networks and protocols design
While the capillary networks concept covers a large panel of technologies, network architectures, applications and services, common challenges remain, regardless the particular choice of a technology or architecture. Our contribution will be twofold. The most traditional one will be to address the aforesaid issues on given technologies, wireless sensor networks, mesh or vehicular network, driven by the challenges of the urban environment. The second part will be to design the mechanisms that will integrate these technologies in capillary infrastructures. The challenges are both into bridging heterogeneous component of the network and into taking into account new routing opportunities into each component.
Evaluation of solutions
Three main methods of evaluation are considered in the team. Each of them has drawbacks, it is their complementarity which is relevant.
First is mean performance evaluation through simulations running the models and traces that we can obtain or develop.
Second is to exploit the combination of local network design rules or analytical models into formal verification or optimization algorithms. Working on the tractability of such methods applied to realistic models is a challenge per se. This kind of methods are very efficient to evaluate bounds on the performance a system can reach and to investigate on the tradeoffs between different metrics.
Experimental evaluation is another important part of our methodology. One challenge, among others, is to design small scale experiments that are significant anyway. In particular, we will investigate the possibility to evaluate vehicular networking protocols prototypes using the fleet of robots available at the CITI lab and the one deployment in the EquipEx FIT platform. We are not planing to do robotics, but to use them as a support for mobile networks. We will also deploy experiments within the EquipEx Sense City platform.