The 5G core network faces increasing challenges in handling dynamic traffic growth, stringent latency requirements, and control-plane scalability due to its largely centralized processing model. Software-Defined Networking (SDN) improves network programmability, while edge computing reduces latency by decentralizing control and processing closer to users. However, existing SDN-based approaches for 5G cores still face limitations in dynamic signaling management, mobility handling, and resilience under varying traffic loads. The research proposes an Adaptive SDN-Edge 5G Architecture (ASE-5G) that integrates SDN programmability with edge-assisted control-plane coordination while preserving compatibility with the 3rd Generation Partnership Project (3GPP) service-based architecture. ASE-5G introduces rule-based and threshold-driven mechanisms to dynamically manage control-plane signaling, flow redirection, and controller failover without relying on GPRS Tunneling Protocol-User Plane (GTP-U) tunneling. The proposed architecture is evaluated using Objective Modular Network Testbed in C++ (OMNeT++) simulations and validated through analytical queueing models, focusing on control-plane signaling performance metrics, including latency, throughput, controller utilization, and failover behavior under different traffic conditions. Simulation results show that ASE-5G achieves low and stable signaling latency under moderate load, bounded throughput saturation under realistic protocol overheads, controlled resource utilization, and resilient controller failover behavior. A qualitative comparison with representative prior work indicates that ASE-5G extends existing SDN-based and edge-assisted 5G core designs by providing quantitatively validated performance insights while maintaining alignment with standardized 3GPP control procedures. These findings suggest that ASE-5G is a suitable and practical architecture for edge-centric and private 5G deployments requiring adaptive and resilient control-plane operation.
Afiliasi berasal dari OpenAlex; metadata OAI-PMH jurnal tidak memuatnya. Penulis tanpa ID OpenAlex dicocokkan berdasarkan nama, sehingga dua orang bisa tergabung atau satu orang bisa terpisah.