Enricher: An Edge-Based QoS Optimization Framework for Vertical Apps Over 5G via Network-Exposed APIs
Emerging fifth-generation (5G) capabilities such as network slicing and network-exposed application programming interfaces (APIs) aim to support vertical industries with stringent Quality of Service (QoS) requirements. However, current solutions lack per-user observability and real-time adaptation, limiting their effectiveness in dynamic, safety-critical environments. This paper introduces Enricher, a modular edge-based framework that autonomously optimizes application-level QoS by leveraging standardized APIs such as CAMARA. Enricher operates as middleware between vertical applications and 5G infrastructure, combining localized telemetry, artificial intelligence (AI)-based forecasting, and intent-driven control, without requiring telecom-specific logic from vertical developers. We validated Enricher’s performance in a real-world 5G standalone (SA) deployment for teleoperated vessel control at the Port of Antwerp-Bruges. Compared to baseline approaches, Enricher reduces Service-Level Agreement (SLA) violations by 84% and resource overprovisioning by 36%, achieving 97.3% SLA compliance. These results position Enricher as a deployable and domain-agnostic solution for per-application QoS adaptation in 5G and beyond.