| With the continued integration of telecommunications networks and the Internet, the service abilities for Mobile Internet open up and evolve further. The network architecture, service capabilities, data resources of Mobile Internet is toward open sharing, allocated on demand, fast response trends. The development of distributed computing technology as the core of service computing, takes into account the convergence of the mobile Internet structure and the diversity of the terminal environment. It makes the Design and construction of mobile systems/applications decouple from the topology relationship of underlying network and the heterogeneity of terminal platforms, greatly enhancing the development efficiency of service systems, flexibly adapt to business needs.Service Computing bases on services as the basic elements, using the openness of service abilities, through the reuse of component-based, large-grained orchestration and flexible assembly, builds loosely coupled distributed applications on demand. Mobile Internet Open Service bases on Service Computing Theory, find the attributes, relationships, structure between services, by studying the integration of telecommunications network and Internet, the openness of component-based services, and achieve the Architecture incremental expansion and the facilitation delivery. The focus of the current study are the openness of service abilities, the Standardization of service description, multi-attribute selection, flexible service orchestration, personalized service recommend, resource-based service storage, and so on. This paper aims to study and resolve how to improve the efficiency of the development of the Mobile Internet Open Service at the architecture level, and reduce the threshold of service operation and delivery; How to achieve multi-QoS open service selection, reduce the complexity of service selection in the multi-dimensional, high spatial; How to study the network characteristics of open service orchestration relations groups, and achieve the dynamic prediction and security of the service network; How to implement open services active Push mode, and improve the recommended accuracy in service sparsity; How to improve the organization and management of the resource-based services open source, to solve the storage bottleneck of the massive resources.The main innovations of this paper include:(1) In the converged network environment and heterogeneous terminal platform for Mobile Internet, the problem is the low efficiency of systems/applications development, the high threshold of service operation, the service marketing cost, and difficult to efficient response to the massive diversified and personalized needs of mobile users. This paper has proposed the specific design of Open Service Architecture for Mobile Internet (OSAMI) on the network side and the terminal side. The Open Service Architecture for Mobile Internet has been the study basis of the Mobile Internet open services, achieved process-oriented, component-based, open technology architecture, reduced the degree of coupling between modules, and completed the longitudinal reuse of functions and the transversely extending of abilities. It has designed a Open Service Distributed Cache Policy (OSDCP), by setting the service cache table and the service request forwarding table in the edge node, and made Token mark the status and effectiveness of service copy, to achieve the service caching retrieve and the forwarding of service requests, ensuring the consistency between the service copy, and shortening the average response time approximately26.85%.(In chapter2, Papers[3][8][15], Standards[1][3])(2) In the Mobile Internet, network infrastructure, communication protocols, interface form of domain and inter-domain are different. The problem is How to select the open service fast and accurately, which matches the user demand in the same functional service sets. This paper has built an Overlay QoS Model (O-QoS) and a Open Service Selection Algorithm based on QoS Overlay Model (OSSA), into consideration of nearly30QoS parameters. The Algorithm has reduced the dimension of QoS parameters, narrowed the space of service sets, gradually refined the optimal range, and obtained the optimal solution or suboptimal solution of open service, improving service quality dynamic adaptability, increasing the accuracy of the open service selection based on QoS, reducing QoS parameters service selection algorithm time spent approximately28.43%.(In Chapter3, Papers[1][6][13], Standard [2])(3) With the increasing number of Mobile Internet Open Services, the service orchestration relationship is complex, and service groups characteristics are significant. This paper has built a Service Orchestration Network, based on Complex Network Theory, and analyzed Network Topology, Degree Distribution, Clustering Coefficient, Average Shortest Path, Degree Assortativity, Betweenness Centrality, and so on. It has discovered that the Service Orchestration Network is a centralized Hub network; Degree Assortativity is approximately0.23, and the network will migrate into the higher Hub with increasing density; the average shortest path is1.75, therefore the higher the efficiency of service discovery and service search. The findings have a high value of theoretical guidance for the entire network connectivity, validity, strong protection, with.(In Chapter4, Papers[2][7][14], Patent[1],Standard [4])(4) In the early time of mobile service system, the initial user total is small, service evaluation is a small number, and mobile end-users interest is in change. This paper has built a Feedback-Loop Service Recommend Model (FLSRM) by feature-based prediction method, formed a Open service recommendation algorithm based on feedback model (OSRA), effectively solved the dynamic offset of users concept in the process of open service recommended, avoided the sparsity of service recommended, and improved the recommendation accuracy rate about4.17%-18.75%.(In chapter5, Papers[4][12], Patent [2])In the heterogeneous, fragmented, bulky massive open service resource, the problem is How to regulate the resource management strategies, improve the efficiency of resource lookups, reduced resource maintenance costs. This paper has built a Structured Resource Object Model (SROM), and designed an Open Resource Transparent Storage Algorithm (OSTSA) based on the model. It has formed a resource storage path, by hash, conversion, and splice, achieved storage path no maintenance, and resource objects evenly distributed. Compared with relational database storage, SROM has improved the efficiency by about87.67%, in the resource attribute retrieval of millions resource management.(In chapter6, Paper [9], Patent [3])... |
PDF Full Text Request