Deadline-Aware Bandwidth Allocation in Cloud Networks

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current bandwidth allocation mechanisms in cloud computing environments fail to optimize resource utilization and predict performance, leading to sub-optimal performance and resource inefficiency, as they do not adequately consider task deadlines and heterogeneous bandwidth demands.

Innovation Solution

A method and system that utilize a virtual network with virtual switches to allocate bandwidth based on task deadlines, employing bandwidth shrinking and time sliding algorithms to minimize cloud resource utilization, while ensuring tasks are completed on time, by reshaping resource requests and conservatively estimating task durations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If bandwidth is allocated based on traditional mechanisms without considering task deadlines, then resource allocation is simpler, but cloud resource utilization is sub-optimal and performance is not predictable

Engineering Contradiction:

Improvecloud resource utilizationVSAvoidbandwidth allocation mechanism complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent changes the allocation parameters from static bandwidth requests to dynamic allocations based on task deadlines. The resource allocation mechanism adjusts bandwidth allocation over time, shrinking bandwidth for tasks with distant deadlines and allocating more bandwidth near deadlines, thereby optimizing resource utilization while maintaining manageable complexity through parameter-based control

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent implements feedback mechanisms where the system continuously monitors task progress and adjusts bandwidth allocation accordingly. The deadline-aware allocation uses task completion status and time remaining as feedback signals to dynamically adjust resource allocation, improving predictability and utilization without requiring overly complex mechanisms

Inventive Principle:

Principle #23Feedback

2Reliability

If bandwidth is allocated to meet all task requirements, then task completion is guaranteed, but resource utilization decreases and costs increase

Engineering Contradiction:

Improvetask completion guaranteeVSAvoidcloud resource consumption

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The patent applies dynamic bandwidth allocation where the bandwidth assigned to each task changes over time based on the task's deadline proximity. Tasks receive minimal bandwidth initially and progressively more bandwidth as their deadline approaches, ensuring completion while minimizing total resource consumption. This dynamic approach replaces static over-provisioning with adaptive allocation

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent uses partial action by allocating only the necessary bandwidth at each time point rather than the full requested bandwidth from the start. The bandwidth allocation is intentionally limited (shrunk) for tasks with distant deadlines and gradually increased, providing just enough resource to ensure completion without excessive consumption

Inventive Principle:

Principle #16Partial or excessive action

3Productivity

If bandwidth is shrunk to optimize resource utilization, then more tenants can be served, but task completion time may be extended

Engineering Contradiction:

Improvenumber of tenants servedVSAvoidtask completion time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent changes the bandwidth parameter dynamically based on the time parameter (deadline proximity). By adjusting bandwidth as a function of time remaining until deadline, the system ensures that tasks receive sufficient bandwidth near their deadlines to complete on time, while allowing bandwidth to be shrunk during early phases when more tenants can be accommodated

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system performs preliminary actions by allocating minimal bandwidth early in task execution when many tenants can be served, then prepares for the final completion phase by progressively increasing bandwidth allocation as the deadline approaches, ensuring timely completion without sacrificing early resource efficiency

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11516146B2Method and system to allocate bandwidth based on task deadline in cloud computing networks

Publication Date: 2022.11.29 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

AI summary

A method implemented to provide a virtual network to tenants requiring bandwidth in a cloud computing environment is disclosed. The method starts with receiving a request for a task at a network device, the request including a first parameter indicating VMs required, a second parameter indicating bandwidths the required VMs need, a third parameter indicating a duration of the task, and a fourth parameter indicating a deadline of the task. The network device then selects a starting time and a bandwidth allocation of the task, where the bandwidth allocation is shrank to be smaller than the second parameter indicating, and where the selection aims at minimizing a measurement of cloud resource utilization considering consumptions of both VMs and bandwidth. Then the network device allocates VMs for the request at the starting time with the bandwidth allocated at a particular location in the cloud computing environment.