Research On Wireless Location Sensing And Data Caching Technolog

With the rapid advancement of Internet of Things(Io T)digital technology,a vast array of terminal devices are being connected to the internet via wireless access points,resulting in an unprecedented increase in data traffic demand.To meet the demands of these novel applications for ultra-high transmission rates,millisecond-level ultra-low latency,and enhanced data security,servers will face unparalleled backhaul traffic loads,posing a challenge that traditional network architectures are obliged to confront.Edge caching technology is viewed as a key solution enabling synergistic operations between cloud computing and edge computing,and it is anticipated to be a necessary support in addressing this challenge.The primary research in this thesis includes:Firstly,in the context of large-scale millimeter-wave wireless communication scenarios,we investigate rapid direction perception technology based on beamforming and propose a rapid beam direction search scheme based on Huffman coding.This scheme provides real-time,accurate user location information at a faster rate,offering data support for channel measurements in caching decisions.Secondly,for centralized edge caching scenarios,a strategy for edge caching that minimizes transmission error rates is introduced.Initially,mathematical relationships between the bit error rate and average fading duration in burst fading are derived,leading to the expression for the burst bit error rate.Subsequently,an optimization equation is constructed based on the burst bit error rate,with the aim of minimizing the system’s overall average delivery bit error rate through adjusting the caching strategies of multiple edge small base stations.Finally,a meta-learning algorithm is proposed,enabling rapid adaptation and prediction of caching strategies in timevarying communication scenarios.Simulation results demonstrate that the proposed algorithm exhibits faster convergence speed compared to existing algorithms.After a few sample trainings,the cache hit rate is higher,and the average transmission error rate is lower,proving the effectiveness and superiority of the algorithm.Thirdly,considering cache decision scenarios in heterogeneous networks,a distributed multibase station cooperative caching strategy is proposed.This strategy takes into consideration more realistic wireless communication scenarios.Initially,each small base station observes the channel quality of users in its local service area,then realizes cooperative caching through minimal information exchange with surrounding base stations.Users are grouped according to different file set requests.A decentralized cooperative cache allocation algorithm based on meta-learning is proposed,which quickly adapts to time-varying communication scenarios through few-shot learning.Lastly,considering the diverse computational capabilities of edge devices,a deep model multi-exit mechanism is introduced.Edge nodes can choose different depth submodels for prediction based on their computational capabilities.Simulation results indicate that the proposed algorithm,compared to other existing decentralized algorithms,exhibits faster convergence speed and better performance;additionally,its performance is very close to that of centralized algorithms.