Low-Latency Orientated Resource Allocation Approach In Wireless Cellular Networks

As intelligent user equipment springs up and low-latency application comes into fashion,current mobile cellular neeworks is supposed to support high-reliable and low-latency wireless connection.Mobile equipment application,e.g.smart phone,tablet computer and wearable devic,calls for a large amount of computation,and has a strong requirement on the life of the batter.To addree the challenge from energe consumption and computing capacity,Mobile Edge Computing(MEC)is deemed as a potential solution,which supplies users in network edge with computation offloarding through powerful edge computing servers.Unmanned Aerial Vehicle(UAV),widely applied to temporarily support aeras beyond wireless cellular networks,has natural advantages in high mobility and low cost.Well air-to-ground wireless channels makes UAV sutable for delay-sensitive data collection.The research on low-latency applications over wireless cellular networks involves complicated comutation and communications resource management problems.Specially,wireless communications leads excess transmission delay to delay-sensitive applications.Therefore,to overcome inconvenience caused by excess delay and meet low-latency requirement,The paper studies the low-latency orientated resource allocation approach in wireless cellular networks.Firstly,the paper considers migrating virtual reality(VR)applications to small base stations with computation resource in small cell networks,with the goal of guaranteeing quality of experience(QoE).In order to estimate the QoE determined on delay,the paper introduce mean opinion score(MOS)to capture VR users' immersive experience.In addition,VR users can randomly switch their field of view(FOV).Considering the actual scene,local VR users' probabilities of FOV are unknown to the small base station.Thus,under the constraints by dynamic and incomplete information,the paper utilize stochastic theory to solve the resource allocation problem on the system-wide MOS maximization.Secondly,the paper focuses on the transmission delay in wireless sensor networks(WSN)and studies the problem of delay-sensitive data collection.In system,it incurs serious time-varying wireless channels that the object fixed-wing UAV operates in straight or circular flight mode.With respect to wireless sensor nodes,the flight trajectory of the UAV and the strategies of other nodes are unknown to them.Therefore,limited by dynamic and incomplete information,the paper applies stochastic game theory to solve the channel selection probem with the goal of minimizing transmission delay.Finnally,the paper studies the multi-user MEC-based VR transmission model,which offords 3D FOV preprocessing to MEC servers.For purpose of jointly optimizing the cache and computation resource in VR devices,communication resource in access point(AP)and CPU resource in MEC server,the paper foemulates the resource management problem with th goal of minimizing systemwide delay,which is a nonconvex constrained noconvex optimization problem.Furthermore,the paper applies successive convec approximation(SCA)approach to solve the proposed problem.