Multi-access Edge Platform

Description

The Multi-access Edge Platform (MEP) is a standardized framework, originally defined by ETSI ISG MEC and integrated into 3GPP architecture, that provides a cloud-computing platform and service environment at the edge of the mobile network. It is deployed within the operator's network, typically at locations like central offices, aggregation points, or even at the base station (gNB) site, to host applications and services in close proximity to end users. The MEP's primary function is to expose edge services and capabilities to authorized applications, which are packaged as Virtual Network Functions (VNFs) or Container Network Functions (CNFs). It manages the lifecycle of these applications, including their instantiation, termination, and relocation.

Architecturally, the MEP consists of several key components. The MEP Platform Manager (MEPM) is responsible for the overall management and orchestration of the MEP platform and the applications running on it. The MEP Orchestrator (MEO) handles the onboarding of application packages, manages application rules and requirements, and triggers application instantiation and termination. Within the MEP itself, the MEP Host runs a virtualization infrastructure (e.g., a hypervisor or container runtime) and hosts the MEP Platform. This platform includes the MEP Service Registry, which catalogs available edge services, and the Traffic Rules Control function, which can steer user plane traffic to the appropriate hosted application based on policies. Crucially, the MEP integrates with 3GPP's 5G Core Network, particularly the Network Exposure Function (NEF) and the Policy Control Function (PCF), to access network and user context (like location and QoS) and to influence traffic routing via the User Plane Function (UPF).

How it works involves a coordinated flow. When a service request requiring edge processing is identified (e.g., an AR session), the 5G core network, via the NEF or PCF, can communicate with the MEP Orchestrator. The orchestrator selects an appropriate MEP Host based on the user's location and the application's needs, instantiates the application, and configures the necessary traffic rules. These rules are then enforced by the UPF, which diverts the relevant user plane traffic to the MEP-hosted application instead of sending it to a distant data center. The application processes the data with minimal latency, leveraging local context exposed by the MEP platform, and returns the result directly to the user or to the core network. This closed-loop interaction between the 5G core and the MEP is essential for delivering ultra-reliable low-latency communication (URLLC) and enhanced mobile broadband (eMBB) services.

Purpose & Motivation

The Multi-access Edge Platform was created to address the fundamental limitation of centralized cloud computing for latency-sensitive and context-aware applications. As mobile services evolved towards augmented reality, autonomous driving, industrial IoT, and real-time video analytics, the round-trip delay to distant hyperscale data centers became a critical bottleneck. The purpose of the MEP is to bring cloud capabilities—compute, storage, and networking—to the very edge of the access network, thereby drastically reducing latency, conserving backhaul bandwidth, and enabling access to real-time network and user context.

It solves the problem of service silos and vendor lock-in at the edge. Before standardization, edge computing solutions were often proprietary, making it difficult for application developers to write once and deploy across different operator networks. The MEP, as a standardized platform, provides a common set of APIs and a consistent environment. This allows third-party developers to create applications that can run on any compliant MEP deployment, fostering an ecosystem of edge applications. It also solves the challenge of efficient resource utilization at the edge by providing a managed, orchestrated platform that can dynamically scale applications based on demand.

Historically, the concept originated from Mobile Edge Computing (MEC) in ETSI, which was later renamed Multi-access Edge Computing to include fixed and other access technologies. 3GPP's integration of MEP, particularly from Release 15 onwards, was motivated by the 5G design principle of native support for edge computing. It addresses the limitations of previous approaches where edge processing was an afterthought, requiring complex and non-standard integrations. By defining clear reference points between the 5G core (especially NEF, PCF, UPF) and the MEP, 3GPP enables automated, policy-driven traffic steering and service exposure, which is essential for realizing the full potential of 5G vertical services.

Key Features

• Standardized execution environment for hosting VNFs/CNFs at the network edge
• Integrates with 5GC via NEF and PCF for context-aware service orchestration
• Includes MEP Platform Manager (MEPM) and MEP Orchestrator (MEO) for lifecycle management
• Provides traffic steering capabilities to divert user plane flows to local applications via the UPF
• Exposes edge services (like location, bandwidth management) to applications via standardized APIs
• Supports application mobility and state relocation for seamless user experience

Evolution Across Releases

Defining Specifications