Standard machine-learning approaches involve the centralization of training data in a data center,where centralized machine-learning algorithms can be applied for data analysis and inference.However,due to privacy res...Standard machine-learning approaches involve the centralization of training data in a data center,where centralized machine-learning algorithms can be applied for data analysis and inference.However,due to privacy restrictions and limited communication resources in wireless networks,it is often undesirable or impractical for the devices to transmit data to parameter sever.One approach to mitigate these problems is federated learning(FL),which enables the devices to train a common machine learning model without data sharing and transmission.This paper provides a comprehensive overview of FL applications for envisioned sixth generation(6G)wireless networks.In particular,the essential requirements for applying FL to wireless communications are first described.Then potential FL applications in wireless communications are detailed.The main problems and challenges associated with such applications are discussed.Finally,a comprehensive FL implementation for wireless communications is described.展开更多
Swarm intelligence in a bat algorithm(BA)provides social learning.Genetic operations for reproducing individuals in a genetic algorithm(GA)offer global search ability in solving complex optimization problems.Their int...Swarm intelligence in a bat algorithm(BA)provides social learning.Genetic operations for reproducing individuals in a genetic algorithm(GA)offer global search ability in solving complex optimization problems.Their integration provides an opportunity for improved search performance.However,existing studies adopt only one genetic operation of GA,or design hybrid algorithms that divide the overall population into multiple subpopulations that evolve in parallel with limited interactions only.Differing from them,this work proposes an improved self-adaptive bat algorithm with genetic operations(SBAGO)where GA and BA are combined in a highly integrated way.Specifically,SBAGO performs their genetic operations of GA on previous search information of BA solutions to produce new exemplars that are of high-diversity and high-quality.Guided by these exemplars,SBAGO improves both BA’s efficiency and global search capability.We evaluate this approach by using 29 widely-adopted problems from four test suites.SBAGO is also evaluated by a real-life optimization problem in mobile edge computing systems.Experimental results show that SBAGO outperforms its widely-used and recently proposed peers in terms of effectiveness,search accuracy,local optima avoidance,and robustness.展开更多
基金This work was supported by research grants from the Engineering and Physical Sciences Research Council(EPSRC),UK(EP/T015985/1)from US National Science Foundation(CCF-1908308).
文摘Standard machine-learning approaches involve the centralization of training data in a data center,where centralized machine-learning algorithms can be applied for data analysis and inference.However,due to privacy restrictions and limited communication resources in wireless networks,it is often undesirable or impractical for the devices to transmit data to parameter sever.One approach to mitigate these problems is federated learning(FL),which enables the devices to train a common machine learning model without data sharing and transmission.This paper provides a comprehensive overview of FL applications for envisioned sixth generation(6G)wireless networks.In particular,the essential requirements for applying FL to wireless communications are first described.Then potential FL applications in wireless communications are detailed.The main problems and challenges associated with such applications are discussed.Finally,a comprehensive FL implementation for wireless communications is described.
基金This work was supported in part by the Fundamental Research Funds for the Central Universities(YWF-22-L-1203)the National Natural Science Foundation of China(62173013,62073005)+1 种基金the National Key Research and Development Program of China(2020YFB1712203)U.S.National Science Foundation(CCF-0939370,CCF-1908308).
文摘Swarm intelligence in a bat algorithm(BA)provides social learning.Genetic operations for reproducing individuals in a genetic algorithm(GA)offer global search ability in solving complex optimization problems.Their integration provides an opportunity for improved search performance.However,existing studies adopt only one genetic operation of GA,or design hybrid algorithms that divide the overall population into multiple subpopulations that evolve in parallel with limited interactions only.Differing from them,this work proposes an improved self-adaptive bat algorithm with genetic operations(SBAGO)where GA and BA are combined in a highly integrated way.Specifically,SBAGO performs their genetic operations of GA on previous search information of BA solutions to produce new exemplars that are of high-diversity and high-quality.Guided by these exemplars,SBAGO improves both BA’s efficiency and global search capability.We evaluate this approach by using 29 widely-adopted problems from four test suites.SBAGO is also evaluated by a real-life optimization problem in mobile edge computing systems.Experimental results show that SBAGO outperforms its widely-used and recently proposed peers in terms of effectiveness,search accuracy,local optima avoidance,and robustness.
基金Project supported by the UK EPSRC(No.EP/N005597/1)the H2020-MSCA-RISE-2015(No.690750)+1 种基金the National Natural Science Foundation of China(No.61728101)the U.S.National Science Foundation(Nos.CNS-1702808 and ECCS-1647198)