Research Of Key Technologies Of End-User-Oriented Service Creation In Mobile Networks

With the advancement of modern technology,wireless network technology,and the mobile Internet industry have developed rapidly.The functions of mobile smart devices,i.e.,smartphones,tablets,mobile wearables,etc.are becoming more powerful.Based on the rich functions of mobile devices and Web services in mobile networks,developers have developed a large number of mobile applications,making mobile services more pervasive in people's daily lives,while stimulating users'demand for more personalized and diversified mobile services.Traditional mobile applications and services are developed by professional developers in programming environments who research and analyze the requirements of users,then design and implement corresponding mobile services and applications for specific platforms.Such a development process is complex for end-users and difficult to adapt to the personalized needs and cross the different platforms.To address the above issues,this paper researches the key service creation technologies for end-users based on Service-Oriented Architecture.On the one hand,we propose a cross-platform graphical mobile application development environment for end-users,allowing users to independently develop applications according to their own needs.On the other hand,we research the QoS-aware automatic service composition technology driven by users' requirements to realize dynamic service provision and propose corresponding optimization algorithms.The main work and contributions include the following three aspects:1)Research on the cross-platform mobile application creation model based on Web components for end users.Aiming at the problem of mobile application development in end-user service generation,a componentized cross-platform mobile application creation model is proposed.In the application creation model,a service component model and component plugin model are defined,which generate and build cross-platform mobile applications based on an event-driven component aggregation method.EasyApp sbased on the application generation model provides end-users with an editor for graphically developing mobile applications and a series of web component libraries that can be used for application development,enabling rapid construction of cross-platform mobile applications.Finally,the application development process was demonstrated,and the usability of the development environment among end users was evaluated.Compared with the baseline system,the time for end users to develop applications using EasyApp was reduced by about 18%.2)Research on QoS-aware automated service composition.Aiming at the problem of service provision in the mobile network,we research the QoS-aware automated service composition problem and propose two optimization algorithms.The automated service composition can automatically orchestrate existing Web services according to users' requirements,and obtain a service chain that can output expected data.On the one hand,we research the Graphplan algorithm and the heuristic search algorithm in automatic planning technology to combine Graphplan with heuristic search and propose the Q-Graphplan algorithm to solve the QoS-aware automated service portfolio problem.Compared with the baseline service composition algorithm,the experimental results show that Q-Graphplan reduces redundant services by about 77%,and the average composition time is reduced by about 24%.On the other hand,to solve the problem of increasing numbers of QoS attributes that need to be considered in the service composition optimization problem,we propose a service composition algorithm with over three QoS optimization objectives,the MaSC algorithm,which employs a novel temporal model to divide the original complex service composition problem to be solved into several simple sub-problems.An evolutionary process is adopted to finally obtain an approximately optimal composite service solution set.The performance of the algorithm was tested using the standard test set,and compared with the existing algorithms.The deviation of the service combination solution obtained by the MaSC algorithm from the optimal was reduced by about 28%,and the average combination time was reduced by about 73%..3)Research on service composition in mobile dynamic networks.Aiming at the service provision in the dynamic mobile network environment,we focus on the problem of service discovery and real-time composition in the mobile ad hoc network.The mobile self-organizing network environment has the characteristics of a lack of central management nodes and the continuous movement of service hosts.We formulate the mobile service composition problem as a real-time planning problem and propose the RTASC algorithm to dynamically construct the composite service in the mobile network and execute it in real-time.RTASC can be divided into two steps:service discovery and service execution.In the service discovery phase,RTASC uses a decentralized heuristic service discovery mechanism to discover currently available mobile services in decentralized service host nodes and based on the service dependency relationship constructs a heuristic overlay network.In the service execution stage,RTASC adopts a real-time planning and real-time execution strategy with lookahead,real-time planning,and execution of composite services.Finally,we test the performance of the algorithm in a simulated environment and compare and analyze the results with the existing algorithms.The performance of the algorithm was tested in a simulated environment and compared with the existing algorithms.The results showed that the composition time of RTASC was reduced by about 20%,and the combination failure rate was reduced by about 35%.