CPU Frequency Scaling for Wireless Network Power Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In telecom data centers, there is a significant excess power utilization during off-peak hours due to high CPU frequency usage, even though network traffic is low, leading to inefficient energy consumption.
Innovation Solution
A method and system for dynamic frequency control of CPU cores based on predictive analysis of network resource utilization patterns, adjusting frequencies during lean workload times to reduce power consumption, using AI-assisted user plane allocation and ARIMA time series models to forecast CPU utilization periods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If CPU frequency is maintained at high levels to ensure data traffic processing capability, then network service reliability is improved, but power consumption increases significantly during off-peak hours
Solution Approach 1:
The patent implements dynamic CPU frequency adjustment by transitioning from static high-frequency operation to adaptive frequency scaling based on real-time traffic conditions. The system dynamically switches between high frequency during peak hours and low frequency during off-peak hours, resolving the contradiction between maintaining service reliability and reducing power consumption.
Solution Approach 2:
The patent changes the operational parameter of CPU frequency from a fixed high value to a variable parameter that adapts to traffic demand. By using predictive analytics to forecast traffic patterns and adjust frequency accordingly, the system maintains reliability when needed while minimizing power consumption during low-demand periods.
2Use of energy by moving object
If CPU frequency is reduced during off-peak hours to save power, then power consumption is reduced, but data traffic processing capability may be affected
Solution Approach 1:
The patent uses predictive analytics and machine learning models to forecast future traffic patterns before they occur. By anticipating peak and off-peak periods in advance, the system proactively adjusts CPU frequency to match upcoming demand, ensuring processing capability is maintained when traffic increases while saving power during predicted low-demand periods.
Solution Approach 2:
The patent implements a feedback mechanism where actual traffic measurements are continuously compared with predicted patterns. The system uses this feedback to refine its predictions and adjust CPU frequency in real-time, ensuring that power reduction during off-peak hours does not compromise the ability to handle traffic when it actually occurs.
3Speed
If static high frequency is used to guarantee immediate response to traffic spikes, then response time is improved, but energy efficiency deteriorates during lean periods
Solution Approach 1:
The patent transforms the static CPU frequency into a dynamic parameter that responds to changing traffic conditions. By using predictive analytics to anticipate traffic spikes and pre-adjusting frequency, the system maintains fast response times when needed while avoiding energy waste during lean periods, thus resolving the contradiction between speed and energy efficiency.
Data Source
AI summary
The disclosure is related to a method for power management in a wireless communication system. The method includes measuring network resource utilization levels for a plurality of virtual network functions (VNFs) over a time period based on at least one network parameter, determining a behavioral pattern of the network resource utilization levels based on a predictive analysis of the measured network resource utilization levels, forecasting a lean workload time interval of the network resource utilization levels based on the determined behavioral pattern and current network resource utilization levels of the plurality of VNFs, and adjusting central processing unit (CPU) core frequencies of a network server based on the forecasted lean workload time interval.