Micro-UAV swarms usually generate mas-sive data when performing tasks.These data can be harnessed with various machine learning (ML) algo-rithms to improve the swarm's intelligence.To achieve this goal while protecting swarm data privacy,feder-ated learning (FL) has been proposed as a promising enabling technology.During the model training pro-cess of FL,the UAV may face an energy scarcity issue due to the limited battery capacity.Fortunately,this issue is potential to be tackled via simultaneous wire-less information and power transfer (SWIPT).How-ever,the integration of SWIPT and FL brings new challenges to the system design that have yet to be addressed,which motivates our work.Specifically,in this paper,we consider a micro-UAV swarm net-work consisting of one base station (BS) and multiple UAVs,where the BS uses FL to train an ML model over the data collected by the swarm.During train-ing,the BS broadcasts the model and energy simulta-neously to the UAVs via SWIPT,and each UAV relies on its harvested and battery-stored energy to train the received model and then upload it to the BS for model aggregation.To improve the learning performance,we formulate a problem of maximizing the percentage of scheduled UAVs by jointly optimizing UAV schedul-ing and wireless resource allocation.The problem is a challenging mixed integer nonlinear programming problem and is NP-hard in general.By exploiting its special structure property,we develop two algorithms to achieve the optimal and suboptimal solutions,re-spectively.Numerical results show that the suboptimal algorithm achieves a near-optimal performance under various network setups,and significantly outperforms the existing representative baselines,considered.

Wanli Wen;Yunjian Jia;Wenchao Xia

School of Microelectronics and Communication Engineering,Chongqing University,Chongqing 400044,China;National Mobile Communications Research Laboratory,Southeast University,Nanjing 210009,China;Jiangsu Key Laboratory of Wireless Communications,Nanjing University of Posts and Telecommunications,Nanjing 210003,China

中国通信（英文版）

[1]Wanli Wen;Yunjian Jia;Wenchao Xia-.Joint Scheduling and Resource Allocation for Federated Learning in SWIPT-Enabled Micro UAV Swarm Networks)[J].中国通信（英文版）,2022(01):119-135

feder,broadcasts

Joint,Scheduling,Resource,Allocation,Federated,Learning,SWIPT,Enabled,Micro,Swarm,Networks,swarms,usually,generate,mas,sive,data,when,performing,tasks,These,harnessed,various,machine,learning,ML,improve,intelligence,To,this,goal,while,protecting,privacy,FL,has,been,proposed,promising,enabling,technology,During,model,training,cess,may,face,energy,scarcity,issue,due,limited,battery,capacity,Fortunately,potential,tackled,via,simultaneous,information,power,transfer,How,ever,integration,brings,new,challenges,system,design,that,have,yet,addressed,which,motivates,Specifically,paper,micro,consisting,one,station,BS,multiple,UAVs,where,uses,over,collected,by,neously,each,relies,its,harvested,stored,received,then,upload,aggregation,performance,formulate,problem,maximizing,percentage,scheduled,jointly,optimizing,wireless,resource,allocation,challenging,mixed,integer,nonlinear,programming,NP,hard,general,By,exploiting,special,structure,property,develop,algorithms,suboptimal,solutions,spectively,Numerical,results,show,achieves,under,network,setups,significantly,outperforms,existing,representative,baselines,considered

0.563402