What Is Edge Computing? Key Concepts, Architectures, and Future Trends

Overview

Edge computing is presented as a new computing model that follows distributed computing, grid computing, and cloud computing. It places cloud computing at its core, uses modern communication networks as the conduit, and brings massive intelligent terminals to the forefront, optimizing resource allocation for computing, storage, transmission, and applications.

Key Characteristics

Global Model: Edge computing covers both central and edge resources, treating the edge as a resource pool that collaborates with centralized clouds, data centers, and supercomputing facilities.

Resource Edge‑ization: Computing, storage, cache, bandwidth, and services are distributed closer to the demand side, delivering high reliability, efficiency, and low latency.

Intelligent, Dynamic Service: The model possesses situational awareness, dynamically configuring resources based on user needs, workload size, and data volume.

Physical and Logical Boundaries: Physical boundaries refer to devices at the edge gateway; logical boundaries distinguish edge functions from centralized cloud functions.

Collaborative Computing: Edge and central resources cooperate; tasks that can be processed locally are handled at the edge, while more complex tasks are offloaded to the central side.

Heterogeneity: Edge environments involve diverse devices with varying specifications, vendors, and functions, making inter‑device connectivity a major challenge.

Service Model Differences: Unlike the traditional three‑layer cloud model (IaaS, PaaS, SaaS), edge computing adapts its service layers to specific application requirements and integrates with cloud services.

Dynamic Architecture: Edge computing combines cloud, network, and terminal layers into a four‑dimensional model (cloud, network, terminal, intelligence) that supports automatic, intelligent resource orchestration.

Reference Architectures

1. ETSI MEC Reference Architecture

Based on the ETSI GS MEC 003 specification, the mobile edge system consists of two layers: the network layer and the mobile edge host layer, forming a hierarchical structure for edge services.

2. Intel MEC Architecture

Intel’s MEC architecture places the mobile edge between the wireless access point and the wired network. It includes routing, capability exposure, platform management, and edge cloud infrastructure subsystems. The first three subsystems run on MEC servers, while the edge cloud infrastructure resides in small or micro data centers at the network edge.

3. ECC (Edge Computing Consortium) Reference Architecture 1.0

The ECC reference architecture is layered into four functional domains: application, data, network, and device. It provides open interfaces for edge industry applications, data optimization services, inter‑device connectivity, and real‑time intelligent integration of sensors, robots, and machine tools.

Application domain: enables edge industry applications via open interfaces.

Data domain: offers data extraction, aggregation, interoperability, semantics, analysis, and ensures security and privacy.

Network domain: supplies connectivity for system interconnection and data transport.

Device domain: embeds sensors and actuators to support real‑time intelligent interconnection.

4. OpenFog Reference Architecture

Published by the OpenFog Consortium, this architecture targets IoT, 5G, and AI workloads. It is built on eight core principles—security, scalability, openness, autonomy, reliability/availability/maintainability (RAS), agility, hierarchical design, and programmability. The architecture spans horizontal layers (hardware platform, virtualization, node management, application support, application services) and vertical perspectives (performance, security, management, data analytics, IT business, cross‑fog applications).

Overall, edge computing extends intelligent capabilities from centralized clouds to the network edge, enabling low‑latency, high‑reliability services and fostering collaborative, heterogeneous, and dynamically orchestrated resource environments.