Network Compute Load Balancing for Cloud Service Embedding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing software-defined networks struggle to efficiently embed cloud services across data centers with heterogeneous resources, leading to suboptimal cloud service embedding and increased cloud demand blocking, particularly in optical transport SDN networks where wavelength, spectrum, and modulation format conversion capabilities are limited.

Innovation Solution

A method that first maps virtual links over physical links in a software-defined flexible grid optical transport network, selecting modulation formats and routing solutions based on probability to minimize spectral fragmentation, and then maps virtual nodes over physical nodes, ensuring resource isolation and capacity constraints while optimizing network and compute resources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If cloud services are embedded in software-defined optical transport networks with limited wavelength, spectrum, and modulation format conversion capabilities, then network resource utilization improves, but cloud demand blocking increases

Engineering Contradiction:

Improvenetwork resource utilizationVSAvoidcloud demand blocking

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent performs preliminary network optimization before compute load balancing by pre-calculating optimal modulation formats, wavelength assignments, and spectrum allocations for virtual link mappings. This advance preparation ensures that when cloud demands arrive, the network is already configured to maximize resource utilization while minimizing blocking, resolving the contradiction between improved utilization and reduced blocking.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent implements a dynamic two-stage procedure that adapts to changing network conditions. The first stage optimizes network resources (wavelengths, spectrum, modulation formats) while the second stage performs compute load balancing. This dynamic adaptation allows the system to maintain high resource utilization while responding flexibly to cloud demand patterns, reducing blocking without sacrificing productivity.

Inventive Principle:

Principle #15Dynamics

2Adaptability or versatility

If virtual links are mapped over physical links with spectral fragmentation, then embedding flexibility improves, but mapping likelihood decreases

Engineering Contradiction:

Improveembedding flexibilityVSAvoidmapping likelihood

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent performs preliminary network optimization that includes pre-calculating spectrum allocations and identifying fragmentation patterns before compute load balancing. By addressing spectral fragmentation in advance through optimized wavelength and spectrum assignments, the system maintains embedding flexibility while ensuring high mapping likelihood when cloud services need to be deployed.

Inventive Principle:

Principle #10Preliminary action

3Productivity

If network optimization is performed before compute load balancing, then spectral efficiency improves, but procedure complexity increases

Engineering Contradiction:

Improvespectral efficiencyVSAvoidprocedure complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent segments the embedding procedure into two distinct stages: network optimization (first stage) and compute load balancing (second stage). This segmentation allows each stage to focus on specific optimization goals - spectral efficiency in the first stage and load distribution in the second - making the overall complex procedure more manageable and implementable while achieving high spectral efficiency.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Inventive Principle:

Principle #15Dynamics

Data Source

PatentEP3014811B1Network optimisation followed by compute load balancing procedure for embedding cloud services in software-defined flexible-grid optical transport networks

Publication Date: 2023.08.09 NEC CORP

Figure 1
Figure 2
Figure 3

AI summary

A method for solving a cloud embedding problem includes first mapping virtual links over physical links followed by virtual nodes over physical nodes. The inventive method entails an efficient procedure, namely network followed by compute load balancing (NCLB), that first maps virtual links over physical links while balancing network resources, and finally, maps virtual nodes over physical nodes while balancing different types of computational resources.