Target Edge Application Server

Description

The Target Edge Application Server (T-EAS) is a functional entity defined within the 3GPP Edge Computing architecture (EDGE). It represents the specific Edge Application Server (EAS) instance that is designated to serve a User Equipment (UE) after a decision has been made to relocate, migrate, or initially assign an application session. The selection of the T-EAS is a key outcome of the EAS discovery and selection procedures orchestrated by the Edge Enabler Client (EEC) in the UE and the Edge Enabler Server (EES) in the network. When an application requires edge computing resources (like low latency or local data processing), the network assists the UE in finding a suitable EAS. The T-EAS is the final chosen destination.

The process involves several steps. The EEC, potentially triggered by an application request, contacts its associated EES. The EES, which has knowledge of available EAS instances and their capabilities (through registration with an Edge Configuration Server), performs EAS discovery. Based on factors like UE location, application requirements, EAS load, and network policies, the EES selects a candidate EAS and provides its information (including IP address or FQDN) to the EEC as the T-EAS. The UE then establishes a direct application-layer connection (e.g., TCP/IP) to this T-EAS. In scenarios of mobility or changing conditions, the application session may need to be transferred from a previous EAS (Source EAS) to a new T-EAS. This relocation procedure is managed by the network with the goal of maintaining session continuity.

The T-EAS hosts the actual application logic or data processing function. It is typically deployed at the network edge, close to the radio access network, to minimize latency. The role of the T-EAS is central to fulfilling the promises of edge computing: enabling ultra-reliable low-latency communications (URLLC), supporting computational offloading for IoT devices, and facilitating advanced services like augmented reality. The 3GPP specifications define protocols and APIs (e.g., in 23.558 and 29.558) for the discovery, selection, and relocation procedures that ultimately identify the T-EAS.

Purpose & Motivation

The T-EAS concept was created to formalize the endpoint selection in 3GPP's standardized Edge Computing framework. Prior to this standardization, deploying applications at the edge relied on proprietary or cloud-centric methods, lacking integration with mobile network control. This made dynamic, network-assisted selection of the optimal edge server based on real-time conditions (like UE mobility or network load) difficult. The EDGE work item aimed to integrate edge computing seamlessly into the 5G system.

T-EAS addresses the problem of how to dynamically and efficiently direct a UE's application traffic to the most suitable edge server instance. It provides a clear target for the network's selection algorithms, enabling optimized service delivery. This is critical for applications sensitive to latency or location, such as vehicle-to-everything (V2X) communication or industrial IoT control. By having a standardized 'target' entity, the procedures for session establishment, relocation, and termination can be uniformly managed, ensuring interoperability between UE, network functions, and edge application providers.

Key Features

• The selected Edge Application Server instance for a UE application session
• Identified through EAS discovery and selection procedures
• UE establishes direct application-layer connection to T-EAS
• Can be a new EAS or a destination for session relocation
• Hosts the edge application logic or data processing
• Deployed at network edge to provide low-latency service

Evolution Across Releases

Defining Specifications