In the near future, it is reasonable to expect that new types of systems will appear, designed or emerged, of massive scale, expansive and permeating their environment, of very heterogeneous nature, and operating in a constantly changing networked environment. Such systems are expected to operate even beyond the complete understanding and control of their designers, developers, and users. Although they will be perpetually adapting to a constantly changing environment, they will have to meet their clearly-defined objectives and provide guarantees about certain aspects of their own behavior. We expect that most such systems will have the form of a large society of networked artefacts. Each such artefact will be unimpressive: small, with limited sensing, signal processing, and communication capabilities, and usually of limited energy. Yet by cooperation, they will be organized in large societies to accomplish tasks that are difficult or beyond the capabilities of todays conventional centralized systems. These systems or societies should have particular ways to achieve an appropriate level of organization and integration. This organization should be achieved seamlessly and with appropriate levels of flexibility, in order to be able to achieve their global goals and objectives. And they should do this in a sensitive and proactive way to meet the current or anticipated needs of their "users". For this reason, they definitely need to adapt to the changes in their environment and change their internal organization by communicating, cooperating, and forming goal-driven sub-organizations. Our envisioned systems have an identified purpose (which depends on the application). Adaptation should continue to serve this purpose. This means that sudden variations of external service requests or environmental physical conditions or of motion of network nodes should not stop the system from serving its goal. Instead the system must continue to operate in a set of desired states with maintained, or gracefully degraded or even improved quality of service.
Main Objective
The aim of this project is to establish the foundations of adaptive networked societies of small or tiny heterogeneous artefacts. We intend to develop an understanding of such societies that will enable us to establish their fundamental properties and laws, as well as their inherent trade-offs. We will approach our goal by working on a usable quantitative theory of networked adaptation based on rigorous and measurable gains. We also intend to apply our models, methods, and results to the scrutiny of large-scale simulations and experiments, from which we expect to obtain valuable feedback. The foundational results and the feedback from simulations will form a unifying framework for adaptive networks of artefacts that hopefully will enable us to come up with a coherent working set of design rules for such systems. In a nutshell, we will work towards a science of adaptive organization of large nets of small or tiny artefacts.
  • We will provide constructive (algorithmic) distributed adaptation techniques.
  • We will provide laws on the effect of adaptation on the system performance, cost of distributed coordination of adaptation, incurred overhead (in terms of communication, energy) and possible trade-offs. We will elaborate schemes to qualify and measure adaptation.
  • We will investigate limits of adaptation (how much to adapt, how long to adapt) and cases where adaptation is impossible.
  • We will test our theoretical insights in practical scenarios by means of simulations and experiments.
The scale and nature of these systems requires naturally that they are pervasive. We will assume that the systems of our study have this property. This inherent characteristic is both a benefit and a constraint for the study of such systems.

The scope of the project is mainly of a generic level. We intend our foundational/modeling effort to cover a very wide range of possible scenarios that include systems made of medium to large numbers of tiny heterogeneous and communicating artifacts. Such scenarios include, for example, monitoring of earthquake regions, forests for fire protection, fluids, robot swarm organization in unknown terrains, and nodes in traffic. However, we foresee two possible use scenarios for applying the results of FRONTS to real world applications: one that uses RFID artefacts for monitoring systems for industry and a second one that uses wireless sensor devices for monitoring traffic. These forseed scenarios still include a wide range of different technical situations where (a) devices use point-to-point or broadcast communication primitives in order to interact, (b) devices interact with base-stations or organize by avoiding dependence on any centralized party, (c) there is sparse geographical distribution (with only few nodes being within the communication range of each other node) or dense geographical distributions (with many nodes communicating and interacting with other nodes) and (d) some of the devices could be mobile. These conditions will naturally change throughout the system evolution and it is imperative that the nodes will sense the changes to the physical environment and accordingly adapt the operation in order to keep the performance of the system at acceptable levels. The devices will have to adapt the communication infrastructure, the energy spending, the internal structures and roles in order to react to the dynamically changing physical environments.

Our foundational approach to adaptation includes effort to devise schemes to measure the quality of adaptation and the degree of optimality of adaptation. Some already forseen measures are (a) how fast the network adapts to the environmental changes (response time) and (b) how much the system pays in terms of energy and communication overhead for a quick adaptation. Less obvious but important measures include: (a) how much to adapt (and the limits of adaptability), (b) how much the system pays in overhead in order to stay prepared for possible future adaptations (cost of maintaining global structures, economy of enegy, cost of continuous readyness and awareness).
Scientific and technological objectives
The ability of networked societies of small artefacts to adapt is composed of two almost orthogonal dimensions, each with its own issues and objectives:
  • The ability for internal continual self-organizational of the network.
    • We will characterize the network awareness of components and adaptability to the needs and to changes in the environment and in the operating conditions.
    • We will investigate the necessary technical requirements for the network to be always able to adapt (i.e. be ready).
    • We will examine how fast it responds (in real time) to track variations in the operation of the network.
    • We will investigate the influence on the performance of the network as the individual entities are adapting (how long does it take to reach a "steady state").
  • The ability to adapt to environmental changes in a dynamic way. In particular, for systems deployed to achieve particular goals, this adaptability should also address the needs, constraints, and commands of its users.
    • We will investigate the ability to adapt in cases of alerts.
    • We will provide rules to prioritize the environmental changes (characterization of changes as major/critical where adaptation is needed, provide some thresholds).
Of course, an adaptive society needs to be composed of individual artefacts that have certain capabilities. We do not plan to consider the capability of each individual artefact to alter and adapt its own hardware by reassembling. Instead, our focus is on their capability of soft adaption, which affects their position and role in their society and their interaction with the other individuals and the environment. Such capability is to some extent technologically feasible for individual artefacts even today; what we really lack is the knowledge on how to combine the artefacts in useful adaptive nets.

To achieve the two main research goals described above, we need to solve several scientific and technological problems

The internal self-organization requires to address at least two problems: (a) how to continually adapt the communication infrastructure and (b) how to achieve "self-stability", which allows effective recovery from transient unexpected faults. We believe that the second problem is of central importance because self-stabilization is an indispensable property of the systems under examination. The adaptation to the environment and to the needs of users requires to address the following problems: (a) how to achieve distributed cooperation, (b) how the system "tribes" discover and track resources, (c) how the net reacts to imposed, uncontrolled dynamicity (such as externally imposed movements of the artefacts because, e.g., they follow ocean currents or are attached to humans), and (d) the extremely important objective of how trust develops or emerges in the whole net or its parts.

Both kinds of adaptational ability require to be able to cope, on one hand, with all kinds of threats, faults, and attacks, and on the other hand, to be able to establish and maintain trust to the humans and to the other parts of the net. Adaptive security and trust in dynamic settings are tasks we need to address in both lines of objectives of our research.

All MapsMaps
All MapsMaps
All MapsMaps
All MapsMaps