Swarms of small devices that bridge the physical with the digital domain and communicate with each other have become essential parts of our daily activity. They form networks to support myriads of new and exciting applications. Users would like such systems to be dependable and adaptive to their needs, as well as to sudden changes of the environment and specific application characteristics.
FRONTS focuses on the algorithmic foundations of such systems and is inspired from diverse fields such as Economics and Markets, Biological Systems and even Thermodynamics and Physics.
We will showcase some key points of our work:
- design rules, best practices and coherent foundations of adaptive networks of tiny artefacts
- mechanisms for the internal continuous self-organization of the network, and for adapting and responding to external changes dynamically
- a unifying software for developing applications that benefit from the FRONTS results
The Visitors Experience
We will demonstrate the main functionalities of an adaptive networked society of tiny artefacts. The networked society will be comprised of 40 wireless sensor nodes and 3 mobile robots. All 43 nodes are resource-limited (16 MHz RISC processor, 96 KB of memory, 802.15.4 radio) and will execute the FRONTS software. They will form a distributed self-organizing society, a notion similar to a distributed operating system, but more suitable for populations of tiny artefacts.
We will showcase the main functionalities and the fundamental abstractions in such a "society" design. These fundamental abstractions actually determine the "spirit" and "style" of the system. The FRONTS society is capable of adapting to a range of external events such as dynamic hardware failures, introduction of new nodes, removal of nodes, appearance of obstacles, security threats, etc. The adaptation phases of the FRONTS society are done in such a way that they always guarantee that the system operates within given performance criteria. Via this demonstrator we will showcase the findings of FRONTS towards providing a unifying scientific framework and a coherent set of design rules, for global systems resulting from the integration of autonomous interacting entities, dynamic multi-agent environments and ad-hoc mobile networks.
The visitor will be able to interact with this adaptive society via the wireless controllers, smart phones (that support 802.15.4 wireless communications) and a tablet (that supports wi-fi) in the following ways:
- The visitor will be able to use the smart phones that communicate directly with the society (over 802.15.4) or the tablet that communicates with the society over the Web, to issue commands to retrieve sensor readings from specific devices or request for aggregated values (e.g., average, min, max etc.) over a given area.
- The visitor may use the smart phone or the tablet to send commands to the wireless controllers to switch on and off their light or the fan.
- The wireless controllers have 3 buttons (red, blue, green) to transmit messages between them. The overhead display will visualize how messages are routed through the society in order to reach their destination.
- The tablet can be used to cause node failures (by switching them off, by resetting them, or by setting random values to their internal memory). Then the visitor will observe how the distributed society will react to these external challenges and adapt the network infrastructure to maintain certain levels of performance. Then the nodes may be switched on again; and the society will adapt once again.
- The smart phone can be used to overhear messages (i.e., as a packet sniffer) that are transmitted from specific sensors and even spoof messages in the network. The society will adapt to such attacks and secure the society against them.
- Special nodes will be available that jam the wireless frequencies over a very limited area. The visitor can position them at any place over the plexiglass box. The society will react to the disruption of communications and adapt the network infrastructure to maintain certain levels of performance. Once the jamming node is switched off, the society will adapt once again.
- The visitor may use the smart phones or the tablet to enable or disable the tracking services of the robots that move at the higher level of the plexiglass box. The society will use the overhead display to point the whereabouts of the robot.
We expect that Theoreticians, Software Designers, Distributed Systems and Pervasive Computing people will get a coherent view of our vision for an adaptive, self-organizing architecture and how systems built based on the FRONTS approach will react to externalities.
Relating to long term, multi-disciplinary, foundational research in future and emerging information technologies
After three years of work we have made decisive steps towards the target of FRONTS. We have produced formal models for adaptive networks of locally interacting devices. Our models capture asynchronous dynamic local interactions, the use of local memories of devices, as well as, continuous interactions with the environment. Our formal models are verifiable and programmable. We have also produced two sets of complete schemes: one set of schemes for the internal continuous self-organization of the network, and another set of schemes for adapting and responding to external changes dynamically. Our schemes for the internal continuous network reorganization include virtual coordinates of devices, smart tags and recommendation methods. Our schemes for reacting to the environment include energy-awareness and assignment of roles to devices. Adaptive security and trust of pervasive networks of tiny devices is a strong contribution of FRONTS: We now have implemented defences to Sybil attacks and have designed a novel use of RFID tags (what we call “security helpers”). Our systems can cooperatively protect themselves from subsystem takeovers.
Establishing a science of an adaptive organization of large networks of small or tiny artefacts will allow researchers to use our models, laws and fundamental properties in order to further investigate the operation of such systems in particular cases, specific to real application scenarios. System designers and solution developers will be able to apply the fundamental principles during the design process in order to investigate the performance of the systems under development and better understand the inherent trade-offs of the resulting network.
Apart from the theories and schemes we have also produced a unifying software that can be executed on a wide range of hardware and software platforms of the Internet of Things, Web of Things and Wireless Sensor Networks. The FRONTS software system establishes distributed and self-organized societies that offer a well-focused set of services. Such a system will be extremely valuable to application developers of the Future Internet. It provides a network infrastructure that is inherently adaptive and operates within certain performance bounds that are known a-priori. Thus the developer can focus on the aspects that are important for the application that are under development. Essentially this significantly reduces the effort required for development of robust and inter-operable applications in the Future Internet.
Pervasive Adaptation is still a growing and not yet completely explored area of future and emerging information technologies. We believe that FRONTS has contributed much towards the realization of Pervasive Adaptation. Our exhibit relates to the following areas of future long term, multi-disciplinary, foundational research:
- Incentives based cooperation for adaptation: Such schemes (resembling markets and nations in economic changes) have not yet thoroughly studied. Societies are very adaptive, and interaction with sociology and even mass psychology should be investigated. We have made a start on such methods in FRONTS but further thorough investigation is needed.
- Adaptation in unexpected crisis situations: This is a very demanded area, since almost all current adaptation models assume some knowledge about "changes" and may not tolerate "extreme" situations. There are issues here with respect to limited resources, hard deadlines, faults and recovery related to FRONTS research results.
- Logics for adaptive pervasive systems of large scale. This may need a paradigm breaking in logical rules. Such logics may be strange, and deduction may mean different things from what we know. Dynamicity, variety and evolution should be captured. FRONTS results can help towards understanding Dynamic Complex Systems and the process of Verifying such logics and components.