Evolvable design of network-oriented services based on a core/periphery structure

Area of Science:

• Computer Science
• Network Engineering
• Telecommunications

Background:

• Emerging network-oriented services require virtualization in multi-access edge computing (MEC) environments.
• Fifth-generation (5G) network technology standardization is concurrent with MEC development.
• Service function networks face dynamic changes, including API evolution, impacting service delivery and cost.

Purpose of the Study:

• To propose a method for evolving service function networks adaptable to user requirements and environmental changes.
• To reduce the cost of maintaining and modifying services within dynamic network environments.
• To ensure future adaptability of network structures to evolving conditions.

Main Methods:

• A core/periphery structure is proposed to divide service functions into core and periphery components.
• The method focuses on evolving the entire service function network based on this structure.
• Maintaining core and peripheral functions at appropriate scales is key to cost-effective evolution.

Main Results:

• The proposed method enables low-cost evolution of service function networks.
• It accommodates nearly 100% of randomly generated service chains.
• The length of accommodated service chains is kept below twice the minimum required length.

Conclusions:

• The core/periphery structure facilitates continuous, low-cost evolution of service function networks.
• This approach maintains a high service accommodation ratio.
• The method enhances network adaptability to dynamic environmental changes in MEC.