Research On Offloading Computing And Edge Caching Of Fog Computing Network

As an emerging technology in 5G era,fog computing is an extension of cloud computing which enables the provision of services with the close proximity of user terminal.It provides transmission of ultra-low latency and high-bandwidth of wireless network and decreases the requirements of backhaul bandwidth.Fog computing is characterized by low latency,location awareness,widely geographical distribution,mobility supporting,edge location,etc,so the energy consumption of user terminal can be decreased effectively.At present,fog computing network research mainly includes offloading computing and edge caching.For offloading computing,when user terminals only depend on cloud server to offload computing,the requirements of low delay and small energy consumption cannot be meeted.When user terminals only depend on fog server to offload computing,the requirements of strong computing power with computing intensive applications cannot be meeted.Therefore,a coordination offloading computing way of cloud server and fog server is proposed.For edge caching,when user terminals get all the requested files from the cloud server,frequent requests will cause traffic congestion at the core network and generate a higher energy consumption.When user terminals get the requested files only from the fog server,it is hard for user terminals to acquire all the files requested due to the limited caching capacity.Therefore,a coordination edge caching distribution way of fog server and cloud server is proposed.The main content and contributions are summarized as follows:1.A coordination offloading computing way of fog server and cloud server is proposed in the fog network.At first,the optimization problem is modeled as a minimization of the offloading computing energy consumption problem of the user terminal by optimizing the transmit power of the user terminal while satisfying the delay constraint and the transmit power constraint.Then the energy consumption of user terminals for fog offloading computing and cloud offloading computing are estimated and compared,respectively.Finally,the user terminal chooses the minimal energy consumption of fog server or cloud server to offload computing.The problem is solved by a non-convex partial approximation into convex reduction and an alternating convex optimization(ACO).Compared with the existing methods,numerical results show the proposed algorithm has the advantage of low energy consumption with user terminals.2.An edge caching distribution method in the fog computing network is proposed.At first,problem is modeled as a minimization of energy consumption by optimizing the transmit power of the user terminal and the coordination ratio(the file ratio that is provided by the fog)while meeting the delay and the transmit power constraint.Then,cloud server and fog server deal with the requested file of the user terminal depending on the ratio which is calculated by optimization.Finally,fog server cooperates with cloud server to offer corresponding proportion requested file to the user terminal.The successive convex optimization(SCA)and alternating convex optimization are adopted to solve the proposed problem.Numerical results show the proposed algorithm makes the energy consumption of the fog network much smaller than any other way.