The browser version you are accessing is too low. To provide better experience, it is recommended that you upgrade the browser toEdgeBrowserOr, it is recommended.GoogleBrowser

Architecture and Key Technology of Distributed Intelligent Open Systems

Release Date:2011-06-21  Author:Xiaoyu Tong , Yunyong Zhang , and Bingyi Fang  Click:


1 Introduction
    Since the reshuffle of the Chinese telecom industry in 2008, Chinese telecom operators have been deploying 3G networks, and upgrade of the Internet has sped up. Competition between operators is fiercer than ever. Considerable bandwidth is now being consumed by a diversity of applications, but telecom operators are not obtaining reasonable ROI on these services. Nor are they adding to the value of their enterprises by capitalizing on the high value-added services.  The Internet has become a huge industry that integrates services and applications, creates greater enterprise value, and has significant social influence. It is entering an era of marketing channels, content, and applications, and the dominant players in the industry chain are changing. Radio and TV operators are converting their radio and TV networks into digital networks in order to promote tri-network convergence. They have become new competitors in the information communication service sector.

    With the arrival of 3G and the development of mobile Internet, IT, the Internet, media, telecom and retail industries are quickening convergence and penetration. Their boundaries are becoming indistinctive, and a new industry ecosystem is coming into being. All participants in this ecosystem need to be open and cooperative so that they can coexist and develop together. A new post-telecom era is coming. Telecom operators are platform providers and should be more open than ever to creating and sharing value, constructing platforms, engaging in standardization, redesigning and optimizing networks, and creating cooperation mechanisms between platforms and terminals.

    In light of the current information service environment and the transformation of telecom business models, this paper proposes a distributed intelligent operation system (DIOS). DIOS is regarded as the core system for future public computing communication networks (PCCNs) [1].  DIOS is capable of intelligent network management, intelligent business development, and intelligent service provision. By using and integrating public platform resources, DIOS shortens service innovation cycles and maintains an ecosystem where several industries can co-exist.

2 Objectives of DIOS
    DIOS is designed to transform the business models of telecom operators. The transformed models of leading Chinese telecom operators may be one of the following [2]:

  • a communication and marketing channel model. Controlled channels, application combinations, and terminal experience are enhanced along with user demand analysis to ensure the operator’s products and services satisfy demands. The operator establishes prompt, economic, and on-demand marketing capabilities.
  • a communication and media advertising model. The operator exploits rich customer demand information, improves its data analysis capability, and uses its marketing channel and terminal delivery capability to develop a “pan-terminal,” “rich media,” segmented all-round advertising model.
  • a communication and information service model. The operator uses the infrastructure and trusted, secure public communication network (PCN)—as well as emerging broadband communication and cloud technologies—to offer services of information facilities, information production, and information application.
  • a prepaid and postpaid model. With the channel and media feature-based postpaid model, communication and information expenses of users are reduced. The operator can quicken the growth its user base (quick aggregation), increase service usage time (increased browsing traffic), enhance the channel and media utility, and improve its profit-making capability from agent fees and advertising fees.

    In transforming its business model, a telecom operator should make use of its network and customer resources to operate commercial platform and interface resources. To adapt to new communication, marketing channel, media advertisement and information service models, an operator should exploit its convergence capability (for network transport and platform access) and its operational capability (for service delivery and support). It should also add a BDCS core operation system (Fig.1). After the transformation from network-centered operation to customer-centered operation, a communication network is simply a channel for users to access services and a means for user aggregation. All new business models operate as follows: First, the BDCS mines user information. Then, the feature database and commercial intelligence model analyzes the collected information. Finally, customized, segmented services are delivered to users via the PCCN, allowing them to have a better experience on their terminals.


    Where network services tend to be information services and information management services, providing services requires a much computing. As network bandwidth increases, computing capability gradually spreads over the entire network. Communication technologies and services are developing towards computing technologies and applications, and computing technologies and applications are developing towards delivery of networks and services. Today’s PCN will evolve into PCCN. The DIOS is the core system of future PCCN. It is distributed, intelligent, open, and integrated, and it is capable of intelligent network management, intelligent business development, and intelligent service provision.


    Using distributed storage, distributed computing resources, distributed databases, and file systems, DIOS can be used to achieve several objectives. First, it allows for on-demand scalability by integrating all resources into the cloud network. Second, it enables the network to be intelligently self-organized and reconfigurable by means of data mining, analysis, and intelligent scheduling. Third, it finds the best match between services and resources, between applications and services, and between terminals and users. Fourth, it quickly and comprehensively responds to service demands by establishing an open-capability engine platform that allows a third party to access cloud resources. Fifth, it maximizes the sharing of integrated resources by unifying virtualization standards and reducing costs. Sixth, it is capable of carrier-grade operability and manageability because it sets up a unified management platform to improve centralized management efficiency.

    PCCN is an information processing network based on virtualization and cloud computing. It integrates the communication network and computer network. By using cloud virtualization, PCCN establishes a support network, service network, and uniform infrastructure resource pool.  By introducing cloud computing, it organizes and uses the infrastructure resource pool efficiently.  Using access, switching, routing and transmission elements of existing PCN, the PCCN implements computing processing, virtual allocation, scheduling management, and a service development environment (Fig.2).


    The DIOS is the core system and the implementation element of PCCN.
Fig. 3 shows the six-layer DIOS topology.


    In the following section, we will discuss the six layers in detail.


    Layer 1: BDCS Layer

    The Business data center system (BDCS) layer consists of three subsystems: network data center, user data center, and operation data center.

    Layer 2: Cloud Resource System

    The cloud resource system comprises three parts: cloud storage, cloud network, and cloud computing. Cloud storage devices include a massive database and a distributed file system. The cloud network device is a compact frame comprising cloud routers and a programmable, virtual cloud switcher with enhanced resource sharing function. The physical cloud devices include mini computers and X86 servers.

    Layer 3: Capability Engine Layer

    The capability engine layer also comprises three parts: event processing engine, business control engine, and data analysis engine.

    Event processing involves implementing workflow-based business procedures and management procedures such as business support systems (BSS), partner relationship management (PRM), operations support system (OSS), open data services (ODS), and office automation (OA).
Data analysis involves computing structured and non-structured data using specific computing rules to find out general business intelligence (BI), customer relationship management (CRM), and search engine services from mass of the data.

    Business control involves strategic control over business, network organization, application adaption, and service delivery according to predefined rules such as session border control (SBC), call session control function (CSCF), subscriber data center (SDC)/home location register (HLR) /home subscriber server (HSS), authentication, authorization, and accounting (AAA), application programming interface (API), and service delivery platform (SDP). Typically, business control involves the business control module of the value-added platform as well as the user credit control module of the billing system.

    Layer 4: Cloud Resource Control Node Layer

    The cloud resource control node layer is based on distributed architecture and shields the complicated physical and logic structures in the cloud. It uses scalable, adaptive load balancing and dynamic, intelligent resource adaption to match services with service engines. In this way, automatic and intelligent scheduling is achieved.

    Layer 5: Cloud Access Gateway Layer

    The cloud access gateway layer, comprising physical access gateways and service platform access gateways, allows terminals to access the cloud. It shields the difference between physical devices and service platforms and allows different terminals to intelligently access the cloud. Thus, unified access is achieved.

    Layer 6: Cloud Terminal Layer

    The cloud terminal layer comprises physical devices and client software. The physical cloud devices can be further divided into thin terminals, dumb terminals (such as sensors of the Internet of things), intelligent soft terminals, and browsers.

    The DIOS architecture is shown in Fig. 4.


5 Key Technologies of DIOS



5.1 Unified Virtualization
    Virtualization technologies are the basis of DIOS, and they fall into several types. Virtualization in DIOS involves server virtualization, storage virtualization, and network virtualization. The virtualization technologies include heterogeneous resource virtualization, live migration of heterogeneous virtual machine, virtual fault-tolerance and disaster recovery, bearer layer virtualization, control plane virtualization, and reconfigurable intelligent network. Of these, heterogeneous resource virtualization is essential for the coexistence of various systems, and it has become a powerful tool for providing customized resources in the context of network and data center. It provides a logic view rather than a physical view of data, computing capability, storage resources, and other resources. Thus, it shields many physical, structured details. Only with unified virtualization can heterogeneous resources be fully exploited and easily managed.


5.2 Unified Cloud Management
    The DIOS matches and schedules resources concurrently for complicated applications in order to maximize the use of resources and enhance execution capability and running performance across the entire network. Thus, users are provided with higher quality services. Unified cloud management includes virtual machine scheduling management (for automatic configuration and scheduling of virtual server resources), deployment management (for automatic installation, configuration, and batch deployment of OS and applications), and storage management (for integrating heterogeneous storage resources, remote synchronous and asynchronous copying functions, remote mirror copy, and snapshot).


5.3 Open Service Capability Engine
    Third-party service providers are the main driving force of development in the telecom industry. Creating a diversity of services requires the participation of a large number of third-party developers. Applications that combine different service capabilities and have different characteristics require many third-party users to be part of the development process. Two main functions of a telecom operator are: to create an open service capability engine, allowing third-party developers to develop low-cost services; and to enhance service innovation to attract more users.

    DIOS can provide new application scenarios for telecom services. With the open business control, event processing, and data analysis platforms of the capability engine, DIOS opens cloud resources to third-party developers for quick response. The open service capability engine of the DIOS has basic service capability and integrated service capability. The basic service capability includes business control, event processing, and data analysis, and the integrated service capability supports access of various networks to provide ICT services. The DIOS controls the open service capability engine for uniform user and capability authentication, load balancing, routing allocation, and session control for all service requests from the open service interfaces. The management component of the service capability engine has control functions so that BSS-related and OSS-related functions can be realized. These control functions include access control, strategy management, configuration management, billing management, fault management, monitor management, and statistical analysis.


5.4 Intelligent Resource Adaption
    When the service capability engine is open to third parties, resources at the lower layer may be used. The intelligent resource adaption function intelligently matches services with different network and computing resources centrally schedules and manages resources of the service capability engine to maximize resource use in the upper layers.

6 Conclusion
    Chinese telecom operators are faced with unprecedented challenges and are in a critical period of transformation and innovation. As platform providers, telecom operators need more intelligent and open business systems. One solution is to form new operation architecture by adding a core BDCS operation system to existing operation architecture. This allows resources to be provided in a distributed and virtual way. Operators should also integrate research and development, product manufacturing, network construction, application development, and service providing in order to establish more open ecosystem and business model.


[1] Xiaoyu Tong, Yunyong Zhang and Yuanshun Dai, “Architecture and key technology of public computing communication network,” J. Communications, vol. 31, no. 8. pp.134-140, 2010.
[2] Xiaoyu,Tong, G.Wu and Yunyong Zhang, Post Telecommunications Age, Beijing: Posts and Telecom Press, 2004. ch.2, p.27.

[Abstract] High-speed large-bandwidth networks and growth in rich internet applications has brought unprecedented pressure to bear on telecom operators. Consequently, operators need to play to the advantages of their networks, make good use of their large customer bases, and expand their business resources in service, platform, and interface. Network and customer resources should be integrated in order to create new business ecosystems. This paper describes new threats and challenges facing telecom operators and analyzes how leading operators are handling transformation in terms of operations and business model. A new concept called distributed intelligent open system (DIOS)—a public computing communication network—is proposed. The architecture and key technologies of DIOS is discussed in detail.

[Keywords] DIOS; public computing communication network (PCCN); cloud computing