Source Edge Application Server

Description

The Source Edge Application Server (S-EAS) is a core functional entity within the 3GPP Edge Application Server (EAS) discovery and mobility framework defined for 5G System integration with edge computing. It resides in the data network (DN), typically at a specific edge location (e.g., a Multi-access Edge Computing (MEC) host). The S-EAS is defined by its state relative to a UE's application session: it is the EAS that currently provides the application service to the UE before a potential migration event is triggered. Its architecture is part of the larger EAS ecosystem, which includes the Target EAS (T-EAS), the Edge Enabler Server (EES), and the Edge Enabler Client (EEC) in the UE.

The S-EAS works in concert with other edge entities to support application context mobility. When conditions warrant a migration (e.g., UE moving, load balancing, or resource optimization), the system initiates a procedure to transfer the application session from the S-EAS to a T-EAS. The S-EAS's key role is to provide the application context to the T-EAS. This context includes the internal state of the application session necessary for seamless continuity, such as user session data, transaction states, or media buffers. The exact content of this context is application-specific and defined outside 3GPP, but the 3GPP architecture provides the signaling framework to enable its transfer.

How the migration works involves several steps. The EEC or network triggers the discovery of a more suitable T-EAS via the EES. Once a T-EAS is selected, a context transfer procedure is executed. The S-EAS receives a request (relayed via the EES) to provide the application context. It then packages this context and sends it securely to the T-EAS. After successful transfer and confirmation, the UE's traffic is redirected to the T-EAS, and the S-EAS may eventually terminate the old session. The S-EAS interacts with the EES using the EDGE-9 reference point (between EES and EAS), and the context transfer may use the EDGE-10 reference point (EAS-to-EAS) as defined in 23.558.

Its role is critical for realizing the promise of edge computing: low-latency, high-bandwidth services that follow the user. By formally defining the S-EAS and the procedures for handing off from it, 3GPP enables stateful application mobility across edge nodes. This is a significant advancement over simple connection rerouting, as it maintains the application's interactive state, which is essential for immersive services like augmented reality, cloud gaming, and industrial control, where session interruption is unacceptable.

Purpose & Motivation

The S-EAS concept was created to solve a fundamental challenge in mobile edge computing: how to move a running, stateful application instance along with a mobile user without service disruption. Prior to its standardization, edge computing deployments often treated application servers as static endpoints. If a user moved, they would disconnect from one server and reconnect to another, losing all session state—a break-before-make approach unsuitable for real-time interactive services. The motivation for the S-EAS/T-EAS model was to enable make-before-break application mobility.

This was driven by the requirements of new 5G verticals outlined in Rel-16 and Rel-17, such as vehicle-to-everything (V2X), industrial IoT, and immersive media. These applications demand ultra-low latency and high reliability, which edge computing provides, but they also assume user mobility. The existing network-based mobility (handover) only manages the IP layer connection, not the application layer state. The S-EAS entity formalizes the 'source' of the application state, providing a clear anchor point from which context can be migrated.

It addresses the limitation of stateless edge services by introducing a standardized procedure for state transfer. This allows application providers to implement mobility-aware edge applications that can expose their context via a defined interface. The creation of the S-EAS role within the 3GPP architecture ensures that edge computing is not just about proximity but also about continuity, making the edge a seamless and dynamic extension of the cloud that follows the user, which is a key differentiator for 5G-enabled services.