Research On Service-oriented Network Architecture And Routing Mechanism

ABSTRACT:The Internet is increasingly integrated into all aspects of human society, the emergence business models such as cloud computing and Internet of Things are widely recognized. Computing resources, storage resources, and network infrastructure are increasingly combined with one another closely. The Internet provides not only transmission of information, but also a heterogeneous computing environment for com-plex mass differentiated services. In order to adapt to this trend, the industry designed a variety of flexible and efficient network architecture to meet content sharing, mobility, large-scale computing, and energy consumption. Therefore, in the future network evolu-tion process, how to provide efficient support for the large-scale services and massive information transmission efficiency is a difficult problem in the future network research.This paper presents a generic service-oriented network model i.e. SONM, focusing on the key resources planning routing mechanism. This new model increases network and services’mutual awareness, and provides reliable, selective information transfer ability. In order to validate the effectiveness of the proposed model, data center network and wireless sensor networks are analyzed by the SONM. In this paper, the main work and contributions include the following four aspects:(1)A service-oriented network model is proposed. This article focuses on service routing and network routing through the design of two layer planning. This model composes of three layers, including application layer, resource negotiation layer, and network layer. Application layer provides the semantic information of services, in-clouding universal naming, query, location, composition and execution of the service. Resource negotiation layer provides the integrated rules of resource negotiation with the resources including communication, computing, storage and etc. The resource negotia-tion layer selects the appropriate service programs and network routing according to the application service layer service programs and the underlying physical network status. The network infrastructure layer integrates heterogeneous physical network resources, which establishes a more effective, energy-efficient network routing. The model has universal significance for the future research of network architecture.(2)A resource integration routing mechanism based on two layers planning is pro-posed. The service layer is as a decision-making layer, network layer as a rock-bottom layer. Colored Petri net is used for service process modeling, and the choice of service plans work as the map search on Petri nets. Demand for the resource of services and network is complex. The multi-objective utility theory is introduced, which builds re-source semantics and utility function. On this basis, the use of two layers planning es-tablishes an optimization model of service routing resources and network routing re-sources which is used to provide better user experience with less service resources ac-cording to the characteristics of the distribution network resources and network status.(3)In order to verify the service-oriented network architecture model and the re-sources integration routing mechanism, data center network and sensor network service is analyzed by the model. Routing mechanism for the sharing of resources between dif-ferent wireless networks is proposed. According to wireless sensor network coverage, a resources-shared routing mechanism can enhance the WSN working time. In this paper, the service process in virtual machines of the data center can use the fewer number of hops between the virtual machines. According to the simulation results and service characteristics of the network infrastructure, this routing mechanism can increase the life cycle of the service process and improve the utilization of network resources.(4)This paper attempts to propose a service-oriented network architecture DHSR-CCN, introducing the CCN network elements. This network architecture is able to provide a more efficient service sharing and composition mechanism to further en-hance the network throughput and network utilization. For service name routing scala-bility problem, this paper proposes a hierarchical name routing mechanism. The method selects the computing power and storage capability of the node as a super node, and the CCN network is divided into core networks and edge networks. The user queries ser-vices and content at the edge of the network they belong to with higher priorities. The simulation results show that the network architecture is able to provide more efficient content sharing and service composition mechanism, further reducing network over-head.