Font Size: a A A

The Physical Layer Security Based Communication Method In Fog Computing

Posted on:2022-02-27Degree:MasterType:Thesis
Country:ChinaCandidate:J L YuFull Text:PDF
GTID:2518306764993109Subject:Telecom Technology
Abstract/Summary:PDF Full Text Request
At present,China’s 5G mobile communication technology is developing steadily.Affected by the epidemic situation,all walks of life have unprecedented demand for digital and intelligent transformation with 5G.As the key technology of future 5G communication,mobile fog computing has become one of the current research hotspots.1)With the help of channel parameters and zero sum game in PLS,this method first deals with the revenue calculation in dynamic environment through Q-learning algorithm,and then obtains the best camouflage attack test threshold,which effectively reduces the false alarm rate(FAR)and miss detection rate(MDR)of the system,which improves the accuracy of camouflage attack detection between users and fog nodes,and ensures the security of users when connecting with fog nodes;2)Secondly,a method of physical layer joint generation based on Q-learning algorithm is proposed,which has been proved by experiment.First of all,in view of the problem that the key power generation rate is not enough to meet the environmental requirements,a key power generation scheme with node support is proposed.Then,in conjunction with the social behavior network,select the best assistant node,and further improve the critical generation rate of the user’s physical layer through the Q-learning algorithm,ensuring the user’s personal privacy;3)Finally,the fingerprint identification method of the device is studied,and the method of classifying the device information carried in the wireless signal is proposed through the LSTM neural network.Based on the LSTM neural network,the processing power of wireless signal I/Q components is improved,and the access control of fog nodes is realized;...
Keywords/Search Tags:mobile fog computing, reinforcement learning, game theory, physical layer security, neural networks
PDF Full Text Request
Related items