AlTrOPiC – AI-based Anomaly Diagnosis and Traffic Prediction for Service Orchestration and Placement in Clouds
The AITrOPiC project focus on the research and development of AI-based methods to enable automated provisioning of services and processing resources in cloud-based computing infrastructures.
Cloud-based computing enables the provisioning of computing, storage, and networking resources in the dynamic fashion. This concept has become the cornerstone of many modern software applications and services. Cloud computing has, however, evolved into a vast and complex ecosystem of technologies, products and services. As a consequence, efficient and effective cloud management has become a major challenge.
Artificial Intelligence (AI) technologies are regarded as useful tools and potential solutions to make networks self-aware and self-managed by incorporating intelligence in the Cloud networks. Particularly, AI may enable networks to learn from experience to make them more robust against vulnerabilities and failures. AI-based applications have the potential to further automate network management and thereby improve performance and efficiency.
The AITrOPiC project focuses on the research and development of AI-based methods to enable automated provisioning of services and processing resources in cloud-computing infrastructures. Thereby, the following 4 major aspects of the service management life-cycle are considered:
- Automatic composition of service and network
- Intelligent placement of Services and virtual machines
- Cloud components anomaly diagnosis
- Traffic prediction and anomaly diagnosis
SRFG will primarily contribute to traffic prediction and anomaly diagnosis. The approach will be based on pattern clustering and neural network modelling.