语义信息驱动的车路协同群智感知服务资源分配方法

打开文本图片集

中图分类号：TP391 文献标志码：A文章编号：1003-3114（2025）06-1279-08开放科学（资源服务）标识码（OSID）：

Abstract：Vehicularnetworks enablecolaborative crowdsourced perception，where mobilevehiclesandroadsideunitscoperate tosensedataandtransitittosrviceprovdersfodpodeliference.Hweerfereeasselyingonotiuoustrasion of high-definitionimagesandvideosimposeasignficantburdenonwirelesscommunication.Meanhile，leaing-basedsemanticcommunicationaistoxtactthmeangfrovisaldataileroutielevantifoationandrasmitteanigfulotenttoiients，resultinginaturaldantagesinduingdatastorageandtansmissionloadsandimproingdatatransmionefiecyi igatethisissue，thispaperpropossasmanticiformaio-drienvehicleoadcoperativecrowdsesingsstem.Thissystemuilies variable-sizelag-basedsmanticomunicatiomodeltoextractandtransmitecoreinforationofsualdatadynaiallyl catingcomputingsoustodirectlimpoveefecyofirelesstrasmisioTocoragetolboatieptipatiof mobilevehiclesandroadsideuits insemanticinformationservices，thispaperthenproposesaservicequalitymeasurethatconsiders delay.Basedonthemeasure，atwodimensionalcontract-torybasedcomputingresoucealocationstrategisdesigned.Tersouce alocationstrategyproposedinthispapercansignificantlyenhancetheservicequalityofsrviceproviders，withanimprovemenipro viderutilitybynearly 20% anda reduction in service delay by approximately 33% ：

Keywords：crowdsourced perception； semantic communication；vehicular collaboration；resource allocation

0引言

车载群智感知是未来6G泛在人工智能的重要应用场景之一[1]。（剩余10441字）