Track Description

With the dynamic and transient nature of the Edge/Fog network topology, any node can initiate applications deployment, host operators of an application, or serve as a data source or sink. Therefore, the complete deployment landscape must be progressively discoverable and resilient to a network topology change. As a result, applications should be scheduled in a decentralized fashion, operators should be (proactively) migrated, and data sources should be discoverable via data marketplaces. The operators’ implementation should take into consideration the hosting node’s capabilities. Therefore, different data structures should be used to cope with the nature of the target host. For example, approximation and succinct data structures can be employed on resource-constrained nodes

Topics of interest include

* Data allocation, management, and processing on the Fog/Edge
* Data storage systems on the Fog/Edge
* Data pipelines for Fog/Edge architectures and applications
* Decentralized scheduling of Fog/Edge applications
* (Stateful) osmotic computing frameworks for streaming applications on the Fog/Edge
* Data marketplaces on the Fog/Edge
* Containerization for data operator deployment on the Fog/Edge
* AI for decentralized scheduling of Fog/Edge streaming applications
* (Stateful) operator migration on the Fog/Edge
* Succinct data structures and sketches for aggregate operators on the edge
* DSL for network topology-aware edge applications specification
* Real-time communication interfaces and protocols
* Resource scheduling that enhances the reliability and scalability.
* FMEC Resources monitoring mechanism and utilization measuring mechanism.
* Real-time load prediction model to optimize the user satisfaction.

Track Chairs and TPC Members

* Ahmed Awad, University of Tartu, Estonia
* Mohamed Ragab, University of Southampton, UK