.png){kind=icon}
.png){kind=icon}
Return
SEDViN: Secure Embedding for Dynamic Virtual Network Requests Using a Multi-Attribute Matching Game
.png){kind=icon}
PRE
AI
DOI：10.1016/j.jpdc.2025.105171.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Comprehensive VNE Framework (SEDViN) for Online VNRs: We propose SEDViN, a secure and efficient VNE framework that enhances VNR acceptance rates and revenue-to-cost ratios while ensuring security compliance for VMs and VLs. • Innovative Preference Generation Strategy: We introduce a ranking mechanism combining entropy-based weighting and TOPSIS to generate agent preferences holistically, considering topology, system, and security attributes. • Matching-Theory-based VM and Shortest Path VL Embedding: VM assignment is modeled as a one-to-many matching game, solved using the deferred acceptance algorithm (DAA), while VLs are embedded using the shortest path algorithm, ensuring security adherence at every step. • Concrete Implementation and Performance Evaluation: Implemented using the alib utility toolkit, SEDViN is compared against VNE-NRM, DPGA, and ReMatch. Results show a 56% increase in VNR acceptance and a 44% improvement in revenue-to-cost ratios while meeting security requirements.
Journal
IF:
4
Papers:
3.8K
Citations:
4.8K
View more organizations