CASCADAS

The technical approach is based on four key scientific principles as key enablers for the CASCADAS vision, and around which the future communication services infrastructures should be designed and built:

• Situation awareness: the capability of services to autonomously adapt to the context from which they are requested and in which they execute demands the technologies to capture situational data and at the same time the ability of the system to effectively exploit it. What is still missing is the investigation of the principles and the algorithms with which such growing amount of distributed information can be organized in proper, strongly distributed “networks of knowledge”, and exploited for the purpose of situated and adaptive service provisioning.
• Semantic Self-organization: Self-organization and the algorithms underlying the emergence of global patterns in complex systems have been (and still are) extensively studied in communications, e.g., in P2P computing, ant-based optimization, social networks. There is the need to explore their potential as enablers for service composition and aggregation, drawing inspiration from biological models, and employing proven techniques to abstract from their “organic” implementation and derive design principles adapted to the requirements of artificial systems.
• Self-similarity: to realize the vision and make its embodiment manageable, the communication infrastructure and services must be fully scalable. One promising option is to explore the potential of self-similarity, whereby individual components self-organize and self-aggregate so as to reproduce nearly identical structures over multiple scales. Self-similarity may indeed be a key enabler also for the composition of complex communication-intensive services and for the structuring of the possibly enormous and multi-faceted networks of knowledge items they will have to exploit.
• Autonomic Component-ware: All the above principles are to be federated by a sound “autonomic component” model, to provide both a general model and a robust framework for building autonomic, self-organizing, semantic services. This component model will supply the basic mechanisms and interfaces to support self-similarity, self-organization and situation awareness. Therefore, our autonomic service components will be explicitly conceived as situated in a knowledge network, fitted with mechanisms for self-aggregation and composition, and designed so as to promote the emergence of high-level ensembles that exhibit self-similarity independently of scale.

The project is structured into 5 work packages (Figure 1), each dealing with specific research thrusts recognized to be critical elements for the situation-aware and autonomic communication services of the future.

Guiding and Validation Activities are the scope of WP6 which provide the means to drive the technology research thrusts, keep them focused around a common perspective and goal, and, later, experiment and validate the research results. Socio-economic analysis will complement the technical requirements by helping in identifying the best directions for optimal penetration of the emerging technologies and results of the project within the European Research Area. In the second phase of the project WP6 will develop a demonstrator of a complete application scenario by integrating all the software and contributions from the Investigation Activities.

The Dissemination Activities will implement a comprehensive outreach and dissemination strategy through 3 pillars, each mapped to a WP: training, dissemination & exploitation, demonstration.

Key Issues

CASCADAS considers a scenario in which dynamic and heterogeneous networks, possibly enriched with sensors and devices connecting with the physical world, have to host the dynamic deployment and execution of applications and services. Such applications and services have to serve users according to both their social situation and the current network and physical situations.

At the application level, CASCADAS considers developing and deploying application and services (by individuals users as well as by software companies and system managers) in terms of ACE components or of ACE aggregates. These components dynamically self-organize as needed with each other and with the already deployed ones, and will start interacting so as to provide the desired functionality in a situation-aware way without (or with very limited) configuration efforts. Below the application level, a sort of “middle-level” hosts knowledge (properly organized in knowledge networks) and ACE-based tools to enforce specific properties such as situation-awareness via knowledge networks, semantic self-organization, adaptive QoS, and security. This middle level is fed both by application-level and social-level knowledge (coming from the upper levels) and by network-level and physical-level knowledge (coming from the lower levels), and continuously interact with these levels, in a sort of continuous tuning feedback that ensures adaptability and, thanks to the connection with the lower-levels, also cross-layer tuning. The power of dynamically influencing and controlling the behaviour of the network and of the application is guaranteed by the possibility of dynamically injecting in the middle-level proper ACEs components to exert such influence. The lower levels, i.e., those concerned with actual network architectures and with physical sensing and embedded systems, are not directly within the CASCADAS scope. Still, CASCADAS will take into account the network-level and the physical-level in terms of the information that, from such level, can reach the higher levels and can be exploited to enforce situation-awareness.

i) Autonomic Communication Elements

The key idea is to identify and rely on a new model of distributed components (ACEs), able to autonomously self-organize with each other towards the provisioning of specific user communication services, and able to self-adapt such provisioning to social and network contexts. These features are likely to dramatically reduce the costs associated to the development and configuration of complex communication services, to leverage the exploitation of distributed computing and communication resources, and to make services more usable and more fitted to user needs. The most important result of the project will be an Open Source toolkit with a set of well-integrated abstractions, algorithms, tools, and application demonstrations.

ii) Pervasive Supervision

Pervasive supervision addresses the runtime construction of an ad hoc and dynamic runtime structure that encompasses a set of cooperating ACEs, and exerts a fully automated and de-centralized control of the communication-intensive service provisioned collectively by those ACEs. This research thrust is primarily relevant to the principles of self-organization and self-similarity, but clearly also relates to situation-awareness.

iii) Component aggregation

The development of algorithms and techniques to achieve dynamic adaptation and enforce given service properties through self-organized component aggregation of ACEs. That kind of aggregation will be the basis for identifying and exploring opportunities for co-operation within ensembles of ACEs, which would allow the collective system to exhibit certain desired properties, for example hit situation dependent QoS targets. This research thrust is primarily relevant to the principles of self-organization and situation awareness.

iv) Trust, security and self-preservation

The development of trust, security and self-preservation techniques, which are of are paramount importance because of the very assumptions upon which the idea of ACEs relies: the heterogeneous nature of the network, the varied capabilities of ACEs, their ability to self-organize and cooperatively supervise each other, which implies the lack of centralized administrative control. Since an ensemble of ACEs possesses those highly dynamic adaptation characteristics, we intend to exploit them to make sure that the resulting system is highly robust and secure, and trust-worth. This research thrust is primarily relevant to the principles of self-organization and situation-awareness

v) knowledge networks

The identification of models and tools for the organization, correlation and composition of knowledge networks, according to which ACEs can exploit all the available information about their situation, however sparse and diverse. Situation is intended here as context, considered in the broadest sense, relating to both (i) the social-organizational context from which services are invoked; (ii) the technological and physical environment in which ACEs live and execute [Est02, Dey00]. This research thrust is obviously primarily relevant to the principle of situation awareness, but also represents a common substrate upon which all the other activities will rely.