Prospective Research for Brussels (PRFB) Projects are one-man projects funded by the Institute for the encouragement of Scientific Research and Innovation of Brussels (IWOIB/IRSIB). These projects aim to support PhD students or graduated researchers working on projects contributing to the reflexion regarding the specific development of the Brussels-Capital Region. What follows is the list of PRFB projects funded in the context of Ambient-Oriented Programming.
Three hardware phenomena are currently radically affecting the way we deal with computers and the software that runs on them. First, there is the ubiquitous computing phenomenon which involves the presence of computing power in everyday life objects like wrist watches and cars. Second, there is the miniaturisation phenomenon that makes computers, phones and PDAs blend entirely. This allows users to move about freely. Third, there is the wireless networking phenomenon which allows all these computing devices to exchange information without hampering user mobility. These three phenomena lead to a near future where people are surrounded by countless networked computing devices that provide them with radically new ways of interacting with each other and with the world that surrounds them. Technically, the wirelessly connected mobile devices form mobile ad hoc networks.
Scalable software engineering models and programming language abstractions that can cope with the specific problems arising from these hardware phenomena inherent mobile ad hoc networks are lacking. For instance, using contemporary middleware technology, exchanging information between two devices (e.g. sending a message) requires about a page of Java code. Needless to say, this renders these techniques unscalable for the construction of complex applications. This research proposal is built on two main pillars: First, we aim to come up with a set of simple programming abstractions that allow developers to tackle the complexities inherent to writing software for mobile ad hoc networks. Second, we propose to validate our research by conducting a city-wide experiment in Brussels that shows the applicability of our results in order to come up with innovative applications for mobile ad-hoc networks. The result of our research will consist of a prototypical technological platform along with a number of experimental applications that illustrate how this technology can enhance the mobility of citizens in a metropolitan area such as the Brussels Region.
Fueled by technological innovation, wireless technology has boomed in the last few years. Usually the distinction is made between mobile ad-hoc networks, formed by spontaneous collocation of devices, and nomadic networks which are formed when devices migrate between several access points of a static infrastructure (like GSM towers). When these access points are also able to move, they are called mobile nomadic networks.
This kind of networks is common in cities like Brussels where public transport is equipped with modern network technology. Users move about with hand-held devices, connecting with several different access points of a “backbone infrastructure”, which is itself constantly in motion. Networks built around this infrastructure will undoubtedly lead to a new kind of distributed systems, which we will call urban applications or cityware. These applications can vary from simple (“I'm looking for a car”) to sophisticated collaborative applications that link shoppers to targeted mobile advertisements. Such urban applications can give rise to a new digital economy.
In distributed software, the notion of “references” between different components is very important. This is problematic when communication partners no longer share a stable connection, but have to communicate using movable, unreliable intermediaries. The limited communication range and mobility of these intermediaries make it very hard to maintain meaningful references. Nevertheless, these references form the base of all data structures, including the collaborative urban applications we envision.
The goal of this research is to invent a heterogeneous strategy that offers programmers the ability to work with familiar references, but which is still easily translatable to indirect communication. This strategy enables distributed components in a mobile nomadic network to refer to each other and supports the construction of urban-area datastructures, which are necessary to write urban applications.
I propose to develop an open middleware which programmers can use to build urban-area data structures, using the references they are familiar with. Sometimes the default strategy for distributing data structures will be sub-optimal: programmers will have a means to select the way data structures are disseminated. The data structures we can build with these references will be adapted versions of the well-studied distributed datastructures.