Hierarchical Network Resource Allocation via Traffic Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hierarchical communication networks face challenges in minimizing energy consumption due to inefficient resource allocation, as existing methods rely on actual traffic volume rather than predictive forecasting.
Innovation Solution
A method that determines data traffic forecasts at the lowest network level and signals these forecasts to higher levels, allowing for the adaptive activation or deactivation of network resources to match expected traffic demands, thereby minimizing energy consumption across the network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If network resources are activated based on actual traffic volume, then energy consumption is reduced, but resource allocation efficiency deteriorates due to reactive rather than proactive control
Solution Approach 1:
The patent creates traffic forecasts in advance for future time periods and uses these forecasts to proactively control network resources before the actual traffic occurs. This preliminary action allows the system to prepare and allocate resources ahead of time based on predicted demand, transforming reactive resource management into proactive management while maintaining energy efficiency.
Solution Approach 2:
The patent implements a feedback mechanism where actual traffic measurements are continuously compared with forecasted traffic values. The deviations between actual and forecasted traffic are used to adjust and refine future traffic forecasts, creating a closed-loop control system that improves prediction accuracy over time and enables more precise resource allocation decisions.
2Productivity
If network resources are allocated based on traffic forecasts, then resource allocation efficiency improves through proactive control, but system complexity increases due to forecast creation and signaling mechanisms
Solution Approach 1:
The patent divides the hierarchical network into multiple levels (access network, aggregation network, core network) and creates traffic forecasts separately for each level. Each network level independently generates its own traffic forecasts based on local traffic measurements, rather than requiring a centralized forecast for the entire network. This segmentation reduces the overall system complexity by distributing the forecast creation burden across multiple independent components.
Solution Approach 2:
Each network element in the hierarchical network performs its own traffic measurements and creates its own traffic forecasts independently. The system is self-sufficient at each level, with network elements autonomously generating forecast data without requiring complex external coordination or centralized control, thereby reducing system complexity.
3Loss of energy
If traffic forecasts are created at the lowest network level, then energy consumption is minimized through targeted resource control, but information transmission requirements increase due to forecast signaling to higher levels
Solution Approach 1:
The patent creates traffic forecasts specifically tailored to each network level's local requirements and characteristics. Each level generates forecasts relevant to its own traffic patterns and resource needs, rather than transmitting unnecessary forecast information across the entire hierarchy. This localization of forecast creation reduces the volume of information that needs to be signaled while maintaining the energy optimization benefits.
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method for determining the expected data traffic in a hierarchical communication network (5) and, depending on this, controlling the network resources in the various network levels (10, 20, 30, 40) of the hierarchical communication network (5) so that energy consumption can be minimized. For this purpose, the traffic volume in the lowest network level (10, 11) can be determined, and a traffic forecast for the lowest network level can be generated from this. The energy demand of the lowest network level can be directly derived from the future expected traffic volume. The traffic forecast is then signaled to the higher network levels (20). In each network level, the respective network resources are then activated or deactivated depending on the traffic forecast.In this way, each hierarchical level of the communication network (5) is allocated the network resources necessary to handle the expected traffic demand.