A CPN-based Model for Resource Allocation in Multi-Access Edge Computing Supporting URLLC
Caio Souza, Renata K. Gomes Dos Reis, Maria G. Lima Damasceno, Marcos Falcão, Andson Balieiro · 2024
Although the integration of Multi-Access Edge Computing (MEC) and Network Function Virtualization (NFV) paves the way for supporting URLCC services in 5G and beyond 5G networks, it raises some practical issues such as limited resources of edge nodes, overhead caused by virtualization, and failures during service processing, which may lead to URLLC requirement violations. By incorporating these features, this work proposes a Coloured Petri Net (CPN)-based model to address the dynamic resource allocation for URLLC services in NFV-MEC-based 5G Networks. Additionally, to mitigate the virtualization overhead, a resource pre-initialization strategy is employed in consonance with service buffering and resource scaling on demand. Results showed that resource pre-initialization has positive impacts on overall system performance without significantly affecting power consumption, achieving similar gains to those achieved by setup rate improvements.