A Novel Approach to Multi-Provider Network Slice Selector for 5G and Future Communication Systems.

Area of Science:

• Telecommunications Engineering
• Computer Science
• Network Management

Background:

• The Network Slice Selection Function (NSSF) is crucial for managing network resources in heterogeneous environments, especially with the rise of 5G and future networks.
• Existing solutions for NSSF in complex technological settings are insufficient, necessitating novel strategies to meet growing application demands.

Purpose of the Study:

• To present an integrated, distributed framework, the Network Slice Selection Function Decision-Aid Framework (NSSF DAF), for addressing the NSSF problem.
• To evaluate the efficacy of multicriteria decision-making methods and machine learning in slice classification and selection within heterogeneous mobile networks.

Main Methods:

• A hybrid approach combining multicriteria decision-making methods (VIKOR, COPRAS, TOPSIS, Promethee II) and K-means clustering for NSSF.
• A distributed architecture with components on user equipment and network edge, minimizing user-side computational load.
• Implementation and testing within multi-domain slicing environments of heterogeneous mobile networks.

Main Results:

• The NSSF DAF demonstrates a viable solution for network slice classification and selection, achieving adequate Quality of Service (QoS) mapping.
• Validation through testbeds confirms the framework's effectiveness in heterogeneous mobile network environments.
• The proposed solution can be implemented at the Edge, Core, or 5G Radio Base Station without increasing end-user computational costs.

Conclusions:

• The NSSF DAF provides a comprehensive and efficient solution to the NSSF problem in complex, heterogeneous network environments.
• The framework ensures an adequate Quality of Experience (QoE) for users by optimizing network slice selection.
• The distributed and resource-efficient design makes the NSSF DAF suitable for deployment across various network segments.