HYPER-AI

In HYPER-AI, we work with smart virtual computing entities (nodes) that come from a variety of infrastructures that span all three of the so-called computing continuum's layers: the Cloud, the Edge, and IoT. It focuses on intensive data-processing applications that present the potential to improve their footprint when hyper-distributed in an optimized manner. In order to give targeted applications access to computational, storage, or network services, HYPER-AI implements the idea of computing swarms as autonomous, self-organized, and opportunistic networks of smart nodes. These networks may offer a diverse and heterogeneous set of resources processing, storage, data, communication) at all levels and have the ability to dynamically connect, interact, and cooperate. HYPER-AI proposes semantic representation concepts to enable heterogeneous resources’ abstraction in a homogeneous way, under a common annotation (computing node), across the whole range of network infrastructures. The main orchestration and control concept of HYPER-AI is inspired by autonomic systems (self-CHOP principles) which employ swarmed computing schemes. Its objective is to make smart multi-node (swarm) deployment scenario design, execution, and monitoring easier, through appropriate AIs for self-configuration (nodes assigned resources), self-healing (swarmed nodes lifecycle), self-optimizing (exploiting built-in situation awareness mechanisms) and self-protecting (intrusion detection, privacy, security, encryption and identity management) at application runtime. In order to support dynamic and data-driven application workflows, HYPER-AI suggests the flexible integration of resources at the edge, the core cloud, and along the big data processing and communication channel, enabling their energy, time and cost-efficient execution. Finally, distributed ledger concepts for security, privacy, and encryption as well as AI-based intrusion detection are also considered.