The project activities are divided in three working groups:
WG1: Autonomous management and real-time control (leader: Phillip Leitner)
We will develop algorithms for static, semi-dynamic and fully dynamic service compositions, which are subject to end-to-end user-perceived QoS constraints, and where third-party and cloud services may be subject to negotiated service-level agreements. This will include methods for determining cost-optimal service compositions, and optimal decision tables for dynamic service (re)composition aiming to achieve the optimal tradeoff between static and dynamic service composition, both having their pros and cons in terms of simplicity and effectiveness. Starting with small-scale examples with a single composite service provider and a single priority, WG1 will gradually increase complexity, and consider complex settings with multiple composite service providers and priority mechanisms, where the intricate interaction between control techniques will be subject to study.
WG2: Methods and tools for monitoring and service prediction (leader: Yoram Haddad)
For autonomous real-time QoS control in large, dynamic, complex many-domain environments like the IoS, there is a great need for scalable, non-intrusive monitoring and measurement of service demands, service performance and resource usage. Additional constraints regarding e.g. privacy and integrity will further complicate the challenges for monitoring and measurement. In addition, proactive service adaptation capabilities are rapidly becoming increasingly important for service-oriented systems like IoS. In this context, there is a need for online quality prediction methods in combination with self-adaptation capabilities (e.g., dynamic service re-composition). Motivated by this, several approaches have been studied, including data-mining and run-time verification techniques, online testing and statistical analysis approaches to detect anomalies and QoS degradation, and simulation-based approaches to simulate future behavior of service-oriented systems. In this context, the main challenges include handling the heterogeneity (e.g., in service types and in time scales) and assessing the relevance of predictions.
WG3: Smart pricing and competition in many-domain systems (leader: Peter Key)
In the context of multiple composite service providers and multiple single-service providers, where each provider implements some pricing strategy, which can range from simplistic flat-rate pricing to advanced dynamic pricing schemes based on real-time information about sales and stock levels. In this context, WG3 will study the implications of these strategies in terms of the existence and characterization of game-theoretical equilibriums, using for example simulations or techniques form algorithmic game theory.
The research to be pursued in the context of the WG’s is subdivided into a dynamic set of Task Forces.