DeSlice: An Architecture for QoE-Aware and Isolated RAN Slicing.

Area of Science:

• Telecommunications Engineering
• Computer Networking
• Wireless Communication Systems

Background:

• Network slicing is a core 5G technology enabling customized network resources for diverse applications and traffic types.
• Efficient resource allocation and virtualization are critical for meeting varied Quality of Service (QoS) demands across network slices.
• Radio Access Network (RAN) slicing faces challenges due to spectrum scarcity, necessitating advanced solutions.

Purpose of the Study:

• To introduce DeSlice, a novel architecture for Radio Access Network (RAN) slicing.
• To develop efficient real-time slicing algorithms that meet heterogeneous QoS requirements.
• To enhance the Quality of Experience (QoE) for end-users across different network slices.

Main Methods:

• Development of the DeSlice architecture for RAN slicing.
• Design of real-time slicing algorithms integrated with the DeSlice architecture.
• Utilizing application-to-network communication for enhanced slice management.
• Performance evaluation through simulations, focusing on joint service of cloud VR, video, and web traffic.

Main Results:

• The DeSlice architecture facilitates efficient real-time RAN slicing.
• The developed algorithms effectively satisfy diverse QoS requirements for multiple slices.
• Simulations demonstrate significant improvements in Quality of Experience (QoE) for cloud VR, video, and web traffic.

Conclusions:

• The DeSlice architecture and its associated algorithms provide a robust solution for RAN slicing.
• This approach significantly enhances user Quality of Experience (QoE) by meeting heterogeneous traffic demands.
• DeSlice offers a promising framework for future mobile network operators and service providers.