Multi-Cloud Flow Admission Automation With Predictive Resource Headroom

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Managing telecommunications networks in multi-vendor cloud environments is challenging due to varying cloud platform characteristics and unpredictable delays in resource allocation, leading to increased complexity and potential over-allocation of resources to meet admission deadlines.

Innovation Solution

An automation system autonomously controls pre-allocated cloud resources across multiple platforms to ensure that new flow admissions meet desired admission deadlines by using data-driven models for control-plane latency and new flow requests, determining the necessary 'headroom' of standby resources based on scaling latency and arrival rates.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of time

If the CSP over-allocates resources to meet admission deadlines, then the admission delay is reduced, but the costs increase

Engineering Contradiction:

Improveadmission delayVSAvoidcosts

Core Design Contradiction:

Loss of timeVSLoss of energy

Solution Approach 1:

The system performs preliminary actions by pre-provisioning resources before they are actually needed. The automation system predicts future resource requirements based on historical data and patterns, and proactively allocates resources in advance. This allows the system to meet admission deadlines without over-provisioning, as resources are prepared beforehand based on accurate predictions rather than conservative over-estimation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements continuous feedback loops where the automation system monitors actual resource usage, admission delays, and performance metrics in real-time. This feedback is used to continuously refine predictions and adjust resource allocation strategies. The system learns from past performance and adapts its provisioning decisions, enabling it to meet deadlines while optimizing costs through data-driven decisions rather than static over-provisioning.

Inventive Principle:

Principle #23Feedback

2Loss of energy

If the CSP allocates resources dynamically without pre-provisioning, then the costs are reduced, but the admission delay increases

Engineering Contradiction:

ImprovecostsVSAvoidadmission delay

Core Design Contradiction:

Loss of energyVSLoss of time

Solution Approach 1:

Inventive Principle:

Principle #10Preliminary action

3Adaptability or versatility

If the system spans multiple cloud platforms with different characteristics, then the flexibility and scalability are improved, but the management complexity increases

Engineering Contradiction:

ImproveflexibilityVSAvoidmanagement complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The automation system serves as a universal layer that manages multiple cloud platforms with different characteristics through a single unified interface. It abstracts the heterogeneity of underlying platforms and provides consistent resource provisioning, monitoring, and optimization across all clouds. This allows the system to leverage the flexibility and scalability of multi-cloud deployment while eliminating the management complexity through centralized automation.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The automation system acts as an intermediary between the CSP and multiple cloud providers. It translates high-level resource requirements into platform-specific provisioning commands, handles platform-specific nuances and characteristics, and provides a unified view of resource allocation. This intermediary layer shields the CSP from the complexity of managing multiple different cloud platforms while maintaining the ability to leverage their diverse capabilities.

Inventive Principle:

Principle #24Intermediary (Mediator)

4Ease of manufacture

If the system uses traditional resource allocation methods, then the implementation is simple, but the ability to meet QoS criteria in multi-cloud environments deteriorates

Engineering Contradiction:

Improveimplementation simplicityVSAvoidQoS guarantee

Core Design Contradiction:

Ease of manufactureVSReliability

Solution Approach 1:

The automation system implements self-service by autonomously making resource provisioning decisions without requiring manual intervention or complex configuration. It automatically monitors system state, predicts resource needs, and executes provisioning actions based on pre-defined policies and learned patterns. This maintains implementation simplicity while dramatically improving QoS reliability, as the system adapts dynamically to changing conditions rather than relying on static, manually-configured allocation methods.

Inventive Principle:

Principle #25Self-service

Data Source

PatentEP4364375B1Over-the-top automation solution to guarantee flow admission delay in multi-vendor cloud systems

Publication Date: 2025.09.03 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

Figure 1
Figure 2
Figure 3

AI summary

The present disclosure provides an automation solution for telecommunications networks deployed in a multi-cloud environment. The automation solution autonomously controls the minimum amount of pre-allocated cloud resources (e.g., CPU, storage, VNFs, etc.) within each underlying cloud platform with the goal of ensuring that the admission delay for accepting new flows into the system is below a desired admission deadline. It fulfills this goal by controlling the amount of pre-allocated resources.