PeopleNET aims at leveraging human mobility and contact opportunities to provide a flexible support to extemporaneous and location-aware people-to-people and people-to-things interactions. Opportunistic Switched Networks (OSNs) are the main focus of the proposal and are considered as a component of a wider networked scenario characterized by increasing heterogeneity of end-points (devices, people, things) and networks (Internet, MANETS, DTN). In PeopleNET the University of Milano, UniMI, will contribute to (1) the analysis and modelling of human mobility, (2) the design of the forwarding algorithms, and (3) the experimentation of an Android-based opportunistic platform in workplace. In such a opportunistic scenario, consensus exists on the fact that human mobility reflects the human attitude to spontaneously and temporary aggregate and socialize into both stable and extemporaneous communities to share common services/content with some spatial or functional locality. The activity of UniMI will start from this research framework and attempt to achieve original contributions as the consequence of the following intuitions: I. Forwarding decisions are likely to benefit of combining social behavior with more stable on-line relationships, i.e. those that people maintain in either their private (e.g., social network communities) or working (e.g., first ranked e-mail destinations) life. II. The scheduled mobility of public transportation systems can profitably contribute to offer urban-wide connectivity for transporting a large class of data between remote places. III. Workplaces are the social environment where human relationships are easy to verify, almost stable and where mobility is often repetitive. It represents a good context where collect contact traces and experiment the Android-based opportunistic platform. Nevertheless, it also represents a hostile context for opportunistic networks (densely populated of both people and technology). Contrasting hints that however create a challenging mix to understand the real potential of OSN technology.
CONTACT ANALYSIS AND MOBILITY MODELS
Most part of the PeopleNET activity follows this basic phase of data gathering and analysis. UniMI will be active in tracing contacts in workplaces and contribute with other URs to analyze human mobility on both small and large scale. Small scale analysis aims to characterize and estimate the influence of social relationships, communities and locality on contacts. This is beneficial to design forwarding policies inside localized communities, such as workplaces or public aggregation places (intra-community analysis). Large scale analysis is helpful to investigate and model the quasi-deterministic mobility of commuters that use the public transportation system. This is beneficial to characterize human mobility between two remote aggregation places (inter-community analysis). As a further outcome of this activity, a thorough definition of the notion of community within an opportunistic scenario is expected to emerge.
UniMI will be active in the design of novel forwarding algorithms that perform forwarding decisions through an autonomic process that extracts the people social attitude from contacts and detects communities while moving. This will lead to a new generation of algorithms that account both for context and community-specific information and leverage the on-board geo-located information to steer the message forwarding. UniMI will focus on both unicast and multicast communications over OSNs. The latter paradigm requires a thorough analysis of the multicast semantics before initiating the phase of protocol design. UniMI is finally interested to the quasi-deterministic mobility scenario offered by the public transportation system. In this direction, the challenge is the design of line-to-line and bus-to-bus routing protocols able to provide a path with minimal QoS guarantees between remote places of a city. All algorithms will be designed to improve the forwarding performances in terms of reduced delivery latency and resource consumption, and maximized delivery likelihood.
UniMI will be active in the design and deployement of an opportunistic test bed. The experimentation will take place in workplace and will be deployed on a set of Android smart phones. UniMI will operate inside the PeopleNET developing team. All activities start from an opportunistic platform prototype available in the Lab.
The project is funded by the Italian Ministry for Instruction, University and Research (MIUR) under Protocol N. 2009BZM837, from October 17, 2011 to October 17, 2013.
Deliverables: (available under "Publications")
|Last Updated on Thursday, 17 October 2013 10:21|