The MESH Project

The Multi-access Edge Computing (MEC) and Fog Computing paradigms are enabling the opportunity to have middle-boxes acting as local proxies, with virtualized resources, for supporting and enhancing service provisioning in edge localities. One of the most challenging issues in MEC/Fog solutions is how to ensure that mobile users always receive the requested performance independently from their runtime mobility across different edge localities.

The purpose of the MEC Service Handoff (MESH) framework is to provide effective support for Virtual Machine (VM) and container (Docker Container) service handoff. MESH significantly differs from state-of-the-art work by combining the following original technical aspects:

  • MESH is proactive. It predicts users movement in order to start handoff process and to proactively move as many as possible pieces of service to target edge node.

  • MESH adopts an application-aware approach for service migration. It can fully exploit application-specific knowledge to determine finer grained migrations in order to minimize latency and to optimize the usage of possibly limited resources such as inter-edge bandwidth and edge storage.

  • MESH supports both VM and Container. It supports MEC/Fog-enabled handoff management with migration of both VM-based and container-based virtualized resources.

  • MESH runs on Raspberry Pi. It runs on edge nodes that are implemented by resource-poor devices, such as Raspberry Pi nodes.

  • MESH is tailored on ETSI MEC specification. ETSI MEC started defining APIs and technical requirements for implementing solution compliant with the standard. In the GR MEC 018 V1.1.1 document ETSI defines the end to end mobility aspects.

In the context of the MESH project, we are also tackling new application-aware approach for service migration by studying the data variability. Our original contribution includes various extensions aimed to speed the migration process up by define a probability of migration based on data access frequency. For more information please refer to the