Parallel Protocol Stack Partitioning for Higher 5G NR Throughput
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 5G NR wireless communication systems face bottlenecks in throughput due to inefficient functional and data decomposition, leading to suboptimal processor utilization and dependency between data segments, which limits data rate improvements.
Innovation Solution
The method and system divide data link layers and/or physical layers based on parallel processing capabilities of user equipment (UE), enabling parallel and independent data processing, with dynamic adjustments to resource utilization based on UE capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If functional decomposition is applied to increase throughput, then parallel processing capability is improved, but dependency between functions prevents true parallelization
Solution Approach 1:
The patent segments the data processing function by dividing data into multiple independent flows, each handled by separate protocol stack instances. This segmentation eliminates functional dependencies by ensuring each flow is processed independently through separate RLC, MAC, and physical layer entities, enabling true parallelization without the critical sections that plague traditional functional decomposition approaches.
2Productivity
If data decomposition into multiple segments is applied, then parallel processing is enabled, but overhead in managing common RLC window and handling RLC procedures increases
Solution Approach 1:
The patent applies segmentation by creating separate RLC entities for each data flow, eliminating the need for a single common RLC window that must be managed across all segments. Each RLC entity maintains its own window and processing state, dramatically reducing the management overhead associated with coordinating multiple segments through a shared window mechanism.
Solution Approach 2:
Instead of dividing data into segments while maintaining a common processing framework (which increases overhead), the patent inverts the approach by creating separate processing frameworks for each data flow. This inversion of the decomposition strategy eliminates the need to manage common windows and procedures across segments, as each flow has its own dedicated processing entities.
3Productivity
If existing quad-core system architecture is used, then processor utilization is limited, but increasing cores introduces complexity in parallel processing coordination
Solution Approach 1:
The patent segments the protocol stack into multiple independent instances that can be distributed across processor cores. Each data flow has its own complete protocol stack instance (PDCP, RLC, MAC, physical layer), which can be assigned to different cores without complex coordination. This segmentation approach simplifies parallel processing coordination compared to traditional multi-core architectures that require sophisticated thread management and synchronization mechanisms.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure discloses system and method for managing operations of one or more applications on an electronic device. The method includes monitoring at predefined instants, device parameters associated with the electronic device and user parameters associated with usage of a plurality of applications in the electronic device. Application usage pattern is identified based on the device parameters and the user parameters by using predefined techniques. Further, the one or more applications are clustered into one or more groups using a real-time learning model stored in the electronic device. The learning model is trained dynamically based on the application usage pattern for clustering. Thereafter, the operations of the one or more applications are managed on the electronic device based on the one or more clustered groups.