End-to-End Network Slicing via Multi-Time Scale Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network slicing technologies in 5G-NR networks primarily rely on network parameters for slice allocation, often neglecting or loosely considering application parameters, which leads to inefficient resource allocation and network management.
Innovation Solution
A method and system for optimal end-to-end slicing in next-generation networks, which involves receiving a demand matrix from User Equipment (UE) that includes Key Performance Indicators (KPIs) for application and network parameters. The system determines UE priority and allocates slices based on the demand matrix and current resource availability, using a combined optimization problem with multi-objective functions and a multi-time scale approach to maximize achievable data rate while maintaining KPIs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If network parameters are used for slice allocation, then resource allocation can be simplified, but application-specific performance requirements cannot be adequately met
Solution Approach 1:
The patent combines network parameters and application parameters into a unified demand matrix that serves as the basis for slice allocation. This merging allows the system to simultaneously consider both network capabilities and application requirements, resolving the contradiction between allocation simplicity and performance satisfaction.
Solution Approach 2:
The patent transforms application requirements into quantifiable KPI parameters that can be directly integrated with network parameters. By changing application parameters into a standardized parameter format (demand matrix with KPIs), the system can process both types of parameters uniformly while maintaining application-specific performance requirements.
2Manufacturing precision
If application parameters are extensively used in slice allocation, then resource allocation accuracy improves, but system complexity increases
Solution Approach 1:
The patent segments the complex allocation problem into distinct components: UEs generate demand matrices with their specific KPI requirements, the gNB collects and processes these matrices, and then performs slice allocation based on the aggregated demand. This segmentation allows extensive use of application parameters without overwhelming system complexity by distributing the processing burden.
Solution Approach 2:
The demand matrix acts as an intermediary structure that bridges application parameters and network parameters. Instead of directly processing raw application requirements, the system uses the standardized demand matrix as an intermediate representation, simplifying the allocation process while maintaining accuracy.
3Ease of operation
If static slice allocation is used, then network management is simplified, but dynamic application requirements cannot be accommodated
Solution Approach 1:
The patent implements dynamic slice allocation where the gNB receives demand matrices from UEs in real-time and adjusts slice allocation accordingly. The system transitions from static pre-configured slices to dynamic allocation based on current application requirements, maintaining simplicity through automated demand-driven adjustment rather than manual reconfiguration.
Solution Approach 2:
The system establishes a feedback loop where UEs communicate their current KPI requirements to the gNB, which then adjusts slice allocation based on this feedback. This continuous feedback mechanism enables the network to adapt to changing application requirements while maintaining simplified management through automated closed-loop control.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
State of art techniques proposing end-to-end slice allocation in next-generation networks are focused on network parameters and even if address application parameters, they do so at higher level. A method and system for optimal end-to-end slicing in next-generation networks is disclosed, The method formulates multi objective functions or a combined optimization with multi-time scale approach to address application and network parameters that use different time scales. Using the multi objective functions, the method aims to minimize a penalty matrix that is indicative of difference between a demand matrix of a User Equipment (UE) and an observable matrix that represents UE experience for received slices. The method is practically deployable, real time and has lower complexity.