Research On Computation Offloading Approaches Supporting Complex Applications In Dummy Areas Of Edge Computing

With the rapid development of mobile internet and mobile communication technology,complex applications(e.g.,interactive games)continue to emerge,which puts forward higher requirements for the task processing capability of a mobile device.Simply processing complex applications on the mobile terminal is gradually unable to meet the service requirements of mobile users.Although traditional cloud computing can provide mobile users with sufficient resources(e.g.,computing resources),it inevitably causes a large round-trip delay due to its long distance from the user.As an emerging service model,edge computing provides the possibility to solve the above problem.Using computation offloading to offload complex tasks that are difficult to handle by mobile devices to edge servers gradually becomes an important technology to alleviate the problem of insufficient resources of mobile devices.Although the current research on computation offloading provides certain technical support for the above application requirements,the types of mobile applications and user requirements are diversified.These features bring new challenges to providing efficient and reliable computation offloading service.Moreover,the environment of the mobile terminal is unstable.The blocking of buildings,signal interference,and imperfect facilities may cause mobile devices to be in the dummy area where they cannot normally obtain edge service.The existence of the dummy area causes edge service to be no longer available.The multiple types of applications and diverse user requirements put forward higher requirements for computation offloading.To resolve the above problems,this thesis focuses on computation offloading of complex applications in dummy areas.Specifically,the main contributions of this thesis are summarized as follows:(1)In order to support different types of complex applications for computation offloading in the dummy area,this thesis designs a computation offloading framework that supports complex applications in dummy areas.This framework enables mobile devices in the dummy area to obtain edge service and can switch among different computation offloading approaches according to the type of the application.Specifically,the framework can be divided into three layers from bottom to top,namely the target device layer,the communication link layer,and the edge computing layer.The target device layer is composed of different mobile devices.Some devices may be located in the dummy area,and each mobile device may run complex applications(e.g.,compute-intensive application).The communication link layer contains a variety of communication technologies that support the interaction between mobile devices and edge servers.To address the problem that mobile devices cannot obtain edge service in the dummy area,the framework introduces the communication technology between different devices.The edge computing layer is composed of multiple edge servers,which mainly provide edge service for mobile devices.(2)In order to improve the computation offloading performance of computeintensive applications in the dummy area,this thesis proposes a way to indirectly obtain edge services by D2 D communication technology and designs a Compute-intensive Application-Supported Offloading Approach in dummy areas(CASOA).Specifically,the approach first selects a mobile device that can communicate with the edge server and can establish a D2 D connection with the mobile device in the dummy area as a relay.After that,the mobile device in the dummy area will establish a D2 D connection with the device and offload the compute-intensive applications through the device.At the same time,in order to make the approach more feasible,this thesis also designs an incentive strategy to enable more mobile devices to actively join the computation offloading.Finally,through a large number of experiments,this thesis verifies the feasibility and effectiveness of the approach from the perspective of time delay and energy consumption.(3)Considering that the single-relay assistance strategy adopted by CASOA cannot support interact-intensive applications with higher bandwidth requirements,this thesis proposes an Interact-intensive Application-Supported Offloading Approach in dummy areas(IASOA).In order to solve the problem that interact-intensive applications in the dummy area cannot obtain edge service,this approach proposes a strategy of multiple relays in parallel transmission.Specifically,the approach selects multiple mobile devices that can communicate with the edge server as relays,and the mobile devices in the dummy area will establish D2 D connections with these devices.Through the parallel transmission of multiple relay devices,the interact-intensive applications in the dummy area can be offloaded to the edge servers.At the same time,this thesis also verifies the feasibility and effectiveness of the proposed approach from the perspective of time delay and energy consumption through a large number of experiments.