Research On Cooperative Edge Caching In Fog Radio Access Networks

With the continuous popularization of intelligent mobile terminals and the rapid growth of social networks,the multimedia services in wireless communication networks also show an explosive proliferation,which makes the fronthaul links face unprecedented data traffic pressure and security threats.To cope with this challenge,fog radio access networks(F-RANs)can effectively address this problem by allocating storage and computing resources to fog access points(F-APs)deployed at the network edge.On the one hand,due to the limited storage and computing resources of F-APs,the social relationship among F-APs becomes a promising scheme to reduce the transmission delay and energy consumption.On the other hand,due to the open access nature of F-APs,security guarantees are strictly needed and ensuring the security of F-APs becomes an effective solution in terms of avoiding file retransmission and improving the cache hit rate.Motivated by the aforementioned problems,a social-aware and secure-aware cooperative edge caching in fog radio access network is studied in this thesis.First,the social-aware cooperative edge caching scheme is investigated.According to the system model and the content transmission modes,a joint optimization problem is constructed to minimize the content transmission delay and system energy consumption.Considering the communication probability between user equipments,the similarity of local popularity,transmission cost and the distance between F-APs,the social relationship among access points is modeled.On the basis of the proposed social relationship model,this thesis designs a clustering method based on the hedonic coalition formation game,which clusters the F-APs to enhance the cooperation gain and improve the content diversity.Finally,the clustering method is combined with the traditional solution of the caching problem to cache contents.Simulation results show that the proposed secure-aware cooperative edge caching scheme can reduce content transmission delay and system energy consumption.Then,on the basis of the above scheme,a cooperative edge caching scheme based on swarm intelligence algorithm is investigated.In order to illustrate the inherent flaws of the original firefly algorithm,the original firefly algorithm is introduced.In order to satisfy the integer constraint in the optimization problem,it is solved by using the heaviside step function.In order to meet the cache capacity limitation in the optimization problem,a mutation strategy based on the local content popularity is designed to improve the original firefly algorithm.The improved algorithm can meet the capacity limitation and jump out of the extreme point with a greater probability,which has a lower computational complexity.Finally,the thesis analyzes the convergence of the algorithm and derives the the parameter conditions of convergence.The simulation results show that the proposed cooperative edge caching scheme based on the swarm intelligence optimization algorithm can accelerate the convergence of the algorithm and improve the network performance effectively.Finally,the cooperative edge caching problem with security guarantees in F-RANs is investigated.Based on the selfish behavior of content providers and the open access nature of F-APs,security influence factors quantify the confidentiality of content and the vulnerability of F-APs.Then,a security prediction model by training a long short-term memory network is designed to predict which F-APs are trusted by the user equipments.The prediction model leverages the file integrity and malicious user requests to determine in advance which F-APs can be used to assist in delivering files.Using historical security data,the long shortterm memory network is trained to automatically learn a security prediction model that reflects the trend of vulnerability and confidentiality.In order to optimize the content transmission delay,the thesis formulates a cooperative edge caching optimization problem by consider of security risk.In view of the complexity of the formulated problem,the thesis resorts to a soft actor-critic based deep reinforcement learning algorithm for solving.The simulation results show that the proposed secure-aware cooperative edge caching scheme has a considerable promotion in the cache hit rate and the transmission delay.