Capsule-Based RAN Slicing for Dynamic Resource Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing software-defined radio access network (RAN) architectures struggle with inflexibility, inefficiency, and inability to adapt to dynamic computing resources and real-time requirements, particularly in future networks with increasing traffic demands and diverse application needs.
Innovation Solution
A flexible software-defined RAN architecture utilizing modular capsules and a Liquidity Controller for dynamic composition and cooperation, enabling real-time adjustments, seamless communication, and on-demand scaling, with support for edge-cloud deployments and cooperative RRM.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional fixed RAN architecture is used, then implementation is simpler, but adaptability to dynamic computing resources and real-time requirements deteriorates
Solution Approach 1:
The RAN architecture is segmented into independent functional units called capsules (e.g., PDCP capsule, RLC capsule, MAC capsule, PHY capsule) that can be dynamically allocated and executed on different computing resources. Each capsule represents a separable functional block that can be independently managed, migrated, and scaled based on real-time requirements.
Solution Approach 2:
The architecture implements dynamic composition where capsules can be added, removed, migrated, or modified at runtime based on changing network conditions and requirements. The system transitions from a static fixed architecture to a dynamic flexible architecture where functional units can be reassigned to different computing resources as needed.
2Adaptability or versatility
If modular capsule architecture is implemented, then adaptability and scalability improve, but system complexity increases
Solution Approach 1:
The capsule architecture creates universal functional units that can serve multiple purposes and be deployed in different contexts. Each capsule (e.g., PDCP, RLC, MAC, PHY) is a self-contained functional block that can be instantiated multiple times and executed on various computing platforms, providing multi-functionality and reducing the need for specialized dedicated components.
Solution Approach 2:
The architecture introduces intermediary components such as the Near-RT RIC (Radio Intelligent Controller) and E2 interface that mediate between the capsule-based RAN functions and the core network or external systems. These intermediaries manage capsule lifecycle, coordination, and communication, simplifying the complexity by providing standardized interfaces and control mechanisms.
3Productivity
If dynamic capsule composition is enabled, then productivity and resource efficiency improve, but control and management difficulty increases
Solution Approach 1:
The architecture implements feedback mechanisms where the Near-RT RIC continuously monitors RAN performance, resource utilization, and capsule execution status. Based on this feedback, the system dynamically adjusts capsule allocation, migration, and configuration to optimize resource efficiency and meet service requirements while maintaining manageable control through closed-loop automation.
Solution Approach 2:
The capsule-based architecture enables self-service capabilities where capsules can autonomously manage their own execution context, resource requirements, and coordination with other capsules. The system automatically handles capsule instantiation, deployment, and migration without requiring manual intervention, improving productivity while reducing operational complexity through automation.
4Adaptability or versatility
If capsule migration is supported, then flexibility and fault tolerance improve, but communication overhead increases
Solution Approach 1:
The architecture performs preliminary actions by pre-configuring and caching capsule images, dependencies, and execution contexts before migration is needed. When migration is required, the system can quickly instantiate pre-prepared capsule versions at the target location, reducing the actual migration overhead and communication requirements compared to on-demand capsule creation.
Data Source
AI summary
A radio access network (RAN) system comprises a plurality of capsules that operate in parallel to cooperatively perform RAN stack operations, each capsule utilizing a processor and a memory configured to perform a corresponding component of the RAN stack operations, wherein the plurality of capsules are configured to support slicing functionality to separately process different types of signals; wherein the different types of signals correspond to different data flows, wherein each data flow of the different data flows serves as a pipe to transfer user data from one or more applications in accordance with a corresponding quality of service requirement and wherein the RAN stack operations result in: communicating backhaul communications with a core communication network; communicating fronthaul communications with a plurality of radio units configured to engage in wireless communications with a plurality of user equipment (UEs) via a radio channel of a radio network; converting, in accordance with a communication standard, received fronthaul communications from the plurality of radio units into backhaul communications transmitted to the core communications network; and converting, in accordance with the communication standard, received backhaul communications from the core communications network into fronthaul communications transmitted to the plurality of radio units.