Research On Interference Prediction And Base Station Switching Technology Based On Intelligent Algorithm

Due to global climate change and operators’ focus on cost,the deployment and operation of future networks will follow the concept of green development.However,to meet the exploding growth in user traffic demand,future networks are trending towards dense deployment,leading to severe interference among small cells and increased overall energy consumption of mobile communication systems.Base station switching technology plays a crucial role in promoting green wireless networks.By enabling dynamic control of base station hardware components,the system can effectively align its energy consumption with the actual traffic level.However,existing studies either rely on theoretical environmental model assumptions that are difficult to apply to complex real-world network environments,or they depend on information that is hard to accurately obtain in actual networks,resulting in wireless or physical resource costs.Alternatively,decisions may be based on coarse-grained telecommunications measurement data,which can lead to significant performance losses during periods of low network traffic.The main contributions of this paper are as follows.Firstly,a LinkLevel Simplified Interference Identification(SII)scheme was designed,and an algorithm called SII-NLR based on Non-Linear Regression(NLR)was proposed,which provides complete wireless channel perception without requiring additional hardware or wireless resources.SII-NLR,benefiting from the serving base station centric design concept and special model design,has extremely low computational complexity.The computational complexity of the state of art solutions is approximately multiplied by the user density of SII-NLR.In simulation scenarios,SIINLR can achieve the target performance with less training data,significantly reducing the data collection time that does not decrease with the increase in computational power.At the same time,the proposed algorithm has a shorter model training time and is more robust to changes in user density,making it well-suited for real-world network environments.Secondly,a Base Station(BS)switching algorithm based on Wireless Position Awareness(WPA)was proposed,and a load estimation formula based on user traffic records and interference identification data was derived.The performance of the BS switching algorithm based on WPA and the traditional Geographic Awareness(GA)was verified in a scenario of sparse user distribution and low traffic demand.Simulation results show that the WPA-based algorithm,benefiting from its data-driven approach and precise wireless channel perception,can achieve excellent performance in energy-saving and service quality assurance.