Multi-Access Edge Computing (MEC) is an ETSI-standardised architecture that places cloud-like compute and storage resources at the radio access network edge, enabling low-latency application hosting within milliseconds of end-user devices. It decouples latency-sensitive processing from centralised cloud data centres by co-locating compute with base stations and other network nodes.
Content
- The concept of placing compute at the mobile base station was first formalised by ETSI’s Mobile Edge Computing ISG in 2014, which later expanded the scope to “Multi-Access” in 2016 to encompass Wi-Fi and fixed-access networks. The standardisation effort produced a suite of APIs allowing applications to access real-time network context — radio conditions, user location, and traffic patterns — unavailable in centralised cloud deployments.
- MEC platforms typically expose a RESTful API surface conforming to ETSI GS MEC 011 and GS MEC 012 specifications, running on COTS x86 hardware installed at base stations or in regional edge data centres. Virtualisation relies on VMs or containers orchestrated by platforms such as OpenNESS (now Smart Edge Open) or proprietary vendor stacks. Application packages (MEC apps) are onboarded via a management-and-orchestration (MANO) layer that handles placement, instantiation, and fault management.
- The MEC ecosystem includes network equipment vendors (Ericsson, Nokia, Huawei), cloud providers offering edge extensions (AWS Wavelength, Azure Private MEC, Google Distributed Cloud Edge), and a growing segment of vertical-specific software vendors targeting manufacturing, logistics, and public safety use cases. Operator deployment has been closely tied to 5G rollout, as the ultra-reliable low-latency communication (URLLC) slice of 5G networks is the natural complement to MEC infrastructure.
- In 2024–2025, MEC adoption has moved from trials to selective production deployment, particularly in industrial and smart-city contexts. Challenges around standardisation fragmentation, business model clarity between operators and application vendors, and the difficulty of achieving sub-10 ms latencies outside dense urban deployments persist. The convergence of MEC with Open RAN architectures and AI-driven Edge Orchestration represents the leading edge of current development efforts.