Network Convergence and Service Convergence

1 IP Bearer Network Convergence
It has been widely recognized to construct a multi-service integrated bearer network on the basis of the Internet Protocol/Multi-protocol Label Switching (IP/MPLS) technology, and to provide business and enterprise customers with quality services by building a fine network that is relatively independent of the existing Internet. Major international network operators have adopted such a dual-network operation solution. For instance, some first-class transnational operators like BT, AT&T and South Korean Telecommunications all proposed their plans to build IP/MPLS-based multi-service integrated bearer networks. The IP/MPLS-based
multi-service integrated bearer network with large capacity adopts advanced designing concepts and technologies like optimized topology and routing structure, Fast Convergence (FC) of the route, fast re-routing, MPLS-based Layer 2 and Layer 3 Virtual Private Network (VPN), Quality of Service (QoS) guarantee, multicast, Network Time Protocol (NTP), and network management and security. It fulfills scalability, availability, capability of carrying integrated services, manageability and security of the telecom-class IP network. Therefore, it can simultaneously carry out high value-added services like Internet services for VIP clients, VPN services at Layer 2 (including private pseudo-wire and virtual local area network) and Layer 3, QoS services, and multicast services (such as IPTV). At present, network operators follow such a technological path: to build the backbone network based on IP MPLS Dense Wavelength Division Multiplexing (DWDM) technology, and use the IP VPN service platform to carry multi-services; to build the multi-service broadband access network based on Digital Subscriber Lines (xDSL) and Passive Optical Networks (xPON) technologies and attach great importance to the management of network edge; to combine Differentiated Services (DiffServ) and MPLS Traffic Engineering (TE) technologies to improve network QoS performance; to improve network control and intelligence; and to push forward new IP services like IPTV and home network.
Network operators are actively carrying out service-oriented optimization and transformation of the Metropolitan Area Network (MAN). It aims to equip the network with multi-service capabilities such as VPN to fully support the development of audio, video and data services. The optimization and transformation include the following tasks:

    (1) Making Network Layers Clear
    The Layer 3 routing network (MAN backbone network) and the Layer 2 access network (Broadband Access Network) are built with clear physical and logical structure by separating the Layer 2 and Layer 3 functions.

    (2) Fulfilling Flat Network Structure
    The MAN backbone network has large capacity with fewer nodes, and the broadband access network has wide coverage. The optimization uses these characteristics to reduce the number of physical and logical cascades in the IP MAN, and makes the network structure flat.

    (3) Differentiating Network Quality
    The service differentiating mechanism is deployed to offer DiffServ with different QoS for different users and services.

    (4) Centralizing Control and Management Functions
    A clearly defined service access control layer is established by the Broadband Remote Access Server (BRAS) and Service Router (SR), fulfilling centralized service provisioning and control. Moreover, it is enhanced to establish the centralized authentication, accounting and network management system at the provisional level, in order to improve manageability of the network, and fulfill telecom-class service support and network management.

    (5) Perfecting Equipment Specifications
    Equipment specifications are made and perfected, requiring new added equipment to have  the functions, performance, management and interworking necessary for supporting IP MAN services.

    The optimized MAN should have the following characteristics:
    (1) The broadband access network is capable of separating Layer 2 users and giving each user a unique ID, which makes the users traceable.

    (2) The IP MAN is capable of providing differentiated services and has various service levels available.

    (3) The IP MAN has multicasting capability in commercial use, and supports multicast IPTV services.

    (4) The IP MAN has Layer 2 and Layer 3 VPN service capabilities for multiple access networks, and it can fulfill interworking between enterprises on both the physical and logical layers.

    (5) The IP MAN is capable of defending the network layer from attacks, and realizes network security.

    In the research on architecture of the Next Generation Network (NGN), telecommunication operators have been positively seeking for a solution to IP networks. With numerous tests and experiments, the NGN architecture is trying to create a beautiful future network. This network will help operators control and manage the IP-based network and have the ability of service management. It will also enable system convergence, with scalability and capability of integration at the service layer and those of access control, labeling and management at the access layer. These are the reasons that the NGN gets much attention. The Network Attachment Subsystem (NASS) and Resource and Admission Control Function (RACF) in the NGN rightly aim at making access management and control possible. The NASS may implement the connection authentication for access and user configuration management, as well as user location management. The RACF is between the Service Control Function (SCF) and Transport Function (TF), and it fulfills the negotiation and reservation of QoS transport resources based on the factors of the user profiles, Service Level Agreements (SLA), network strategies of the operator, service priority, and availability of resources in the access and core networks.

    The development trends of the IP bearer network are mainly in the following aspects:

    (1) The network architecture is developing towards bearing integrated multi-services. The operators will optimize and reconstruct their IP networks to support multi-play services. The demand of high-level clients evolves into broad bandwidth, and the network will be improved into high reliability, high scalability, high QoS, low delay and line-speed transfer.

    (2) The network will allocate a unique ID to each user, and it has the capability of service aware.

    (3) The network will support multicast in large-scale commercial use. The router matrix technology will be gradually introduced into super core nodes in order to simplify the overall network architecture. With regard to core router equipment, the 40GPOS interface has begun its trial commercial use. The 10GE interface will take the place of the 10GPOS interface to become a major interface type in the MAN. The SR will support high-performance deep packet inspection capability with its development in integrity, large capacity and high performance.

    (4) The telecommunication operators begin to positively face up to the bandwidth occupied by Point to Point (P2P) traffic and channel off the Internet traffic. The operators are making exploration of architecture of the telecom-class IP bearer network, and gradually introducing the end-to-end resource management and control platform to the entire network.

    (5) Great technological progress will take place in telecom-class Ethernet, which will expand the service range provided by the operators for business clients.

2 Control Network Convergence
Concerning the control architecture in the NGN, there are Softswitch system  and IP Multimedia Subsystem (IMS) defined clearly. The former supports voice services, while the latter supports multimedia session services. Moreover, the control architecture for non-session services, such as stream media services, is under study.

2.1 Standardization of IMS
The 3rd Generation Partnership Project (3GPP), International Telecommunication Union-Telecommunication Standardization Sector (ITU-T), and European Telecommunications Standards Institute (ETSI) all defined the IMS. The three definitions of the IMS frame are basically consistent and mutually quoted. However, their emphases differ. The 3GPP defined the IMS from the angle of a mobile network, while the ETSI and ITU-T took the IMS as an organic part of the NGN and also defined the standard interfaces of the IMS with other subsystems of the NGN.
The 3GPP has three versions of the IMS standards: R5, R6 and R7. R5 mainly defined the basic frame and 3G access capability of the IMS. R6 defined the interfaces and functions in detail, as well as the Wireless Local Area Network (WLAN). R6 was frozen in December 2004. R7 optimizes the mobile parts, introduces the TISPAN R1 standards for supporting xDSL access, defines access-independent strategy control and billing architecture, and takes the Voice Call Continuity (VCC) capability into consideration. The R7 version is still under completion.

     R1, R2 and R3 are three versions of the IMS standards defined by ETSI TISPAN. R1 introduced the 3GPP IMS R5 standards, and improved them to support fixed access modes such as xDSL. R1 focused on the IMS in the NGN architecture, Public Switched Telephony Network (PSTN) emulation subsystem, PSTN simulation subsystem, and NASS/Resource and Admission Control Subsystem (RACS) access subsystem. This version was completed. R2 is studying such service requirements as Fiber to the Premises (FTTP), WiMAX access, and Fixed Mobile Convergence (FMC). R3 will add mobility management function, which has not been started yet.

    To sum up, the IMS standards released in 3GPP R5 have been mature in the control of mobile users. R6 released the standard for large-scale commercial use of mobile networks. TISPAN R1 is the control standards for both mobile and fixed broadband users. TISPAN R1 was released at the end of 2005. The R7 version is still under completion[1-3].

2.2 Application of IMS Network
According to the requirements of quick promotion of new services, network optimization and resources construction, most operators are doing IMS trails while only a few have started IMS deployment for commercial use.

    The IMS is capable of providing diversified services like session-based multimedia conferencing, one number for all calls, Number Portability (NP), Centrex, Push to Talk over Cellular (PoC), presence, group management and instance messaging.
As for the converged and combined services, the IMS can provide Circuit Switching and IMS Integrated Services (CSI) such as video sharing, VCC, mixed Centrex, gaming (including communication services between members, such as instant messaging and multi-party conversation), combination of web searching and
click-to-dial, and combination of audio-video communication and IPTV.

    Mobile operators are mainly concerned with providing multimedia value-added services such as PoC and Video Sharing through the IMS based on the mobile network. Fixed network operators, on the other hand, mostly focus on providing IP Centrex to enterprise clients and VoIP second-line service to public customers through the IMS. As for the integrated operators of both fixed and mobile networks, it is attractive to implement access of both fixed and mobile users through a unified IMS core network.

    Lightreading made an investigation on the IMS implementation plans of sixty network operators all over the world in February 2006. According to the investigation results, 8% of the operators were deploying the IMS, over 60% would begin to deploy the IMS before the end of 2007, and 19% would deploy the IMS after 2007. Gartner had a statistics that there were 46 IMS contracts in the world during the 2nd quarter of the year 2006, among which 90% were experimental. And 64% out of the experimental IMS networks were built for checking the capabilities of the IMS. In fact, most existing IMS networks in commercial use worldwide are still in their infancy.

2.3 IMS and Softswitch
The designing idea of Softswitch is that services are separated from control and bearer, and that all entities connect and communicate with each other through standard protocols, in order to provide services more flexibly. Softswitch is a software-based distributed switch/control platform. It separates call control from the gateway, and opens the protocols between service, control, access and switch.
The major advantage of Softswitch is its ability to provide mature voice services. The Softswitch technology is able to take the place of circuit switching, and it has a certain capability of supporting broadband multimedia services. Softswitch and IMS have the same design idea, which is to realize the separation of bearer from services based on IP. They also have overlapping functions. However, IMS has more advantages in roaming management of broadband users and service capabilities, because it, based on the Session Initiation Protocol (SIP), can support mobility management, and has flexibility of service application and certain QoS guarantee. Moreover, IMS exceeds Softswitch in terms of providing session-type multimedia services and standardization.

2.4 IMS and P2P
IMS and P2P are two different architectures. IMS adopts a Client/Server (C/S) architecture and lays more emphasis on management and control centralization. It is based on the SIP. In IMS, the management and control of service capability between clients is centralized, and media capabilities of clients are equivalent. However, P2P adopts a decentralized architecture in both service capability and media interaction.

    These two different architectures learn from each other. The idea of decentralization is introduced to the IMS; IMS takes on a P2P feature gradually. On the other hand, SIP is introduced to the P2P architecture, and standardization of the protocols based on decentralized service capability has begun.

    IMS and P2P seep into each other in service capability. P2P has powerful penetration capability and wide coverage; the IMS mainly offers session-based services, and also supports value-added services such as video-audio communications and stream media services. IMS has a complete service frame, but many of its conceptual service capacities have not come true. Only traditional equipment vendors and traditional operators have participated in the actual IMS application and value-added service development.

    P2P uses distributed computing, together with client-end capabilities, to fulfill large-scale processing of service capabilities. The service capabilities include content distribution, video-audio communications, stream media services and cooperative computing. P2P has been widely applied and there are a remarkable number of professional personnel for P2P development.

    IMS and P2P represent two different development trends of the networks. IMS is in line with the ideas of traditional operators for development directions, management system and service provisioning. However, P2P is a part of Web2.0, representing the development trend of the Internet. International operators in transformation pay attention to both IMS and P2P, resulting to the development of both.

    IMS and P2P technologies may complement and compromise each other. IMS may be the core of the network, and it centralizes the configuration of some functions on the network such as authorization, service authentication management, translation of protocols, and adaptation and conversion of media codes, which leads to simple architecture and good control and management. In the long run, P2P can be used for fulfilling certain functions between IMS terminals when the terminals are able to support P2P technology.

    Since offense is the best defense, operators should explore and develop new transverse integrated networks for P2P services of their own. They should make comprehensive use of the existing advantageous resources in the subscriber, network and operating platform, and explore a new business model of “forward and backward combined running, with participation of clients” in order to develop unconventional telecom services with the P2P technology as soon as possible.

2.5 IMS, NASS and RACS
NASS is a part of the NGN, and is mainly in charge of managing the attachment of users to the access network, including user authentication, network address assignation and location management.

    The existing access network has acquired some of the NASS functions such as user authentication and address assignation, but does not have service system address assignation and location management functions. It is possible to combine the IMS and NASS technologies to fulfill access authentication and access location identification.

    RACS is also a part of the NGN. It correlates the resource demand of the service layer (IMS) with the resource distribution of the bearer layer, and fulfills strategy control, resources reservation, admission control, Network Address Translation (NAT), and firewall traversal. RACS and the network-layer QoS technology may coordinate and cooperate with each other. The network-layer QoS technology is used to execute the resource distribution strategy of RACS, and DiffServ and MPLS are necessary conditions for the deployment of RACS. The deployment of RACS is determined by the requirements of services.

    NASS and RACS can support both IMS and other services. The deployment strategy of NASE and RACE depends on the demand of services and on economic cost.

2.6 IMS and Service Platform
The service platform consists of an Integrated Service Management Platform (ISMP) and an integrated service provisioning platform. The latter has functional entities such as the Service Access Gateway (SAG) and Service Engine (SE), and can offer integrated services with reusable service capabilities and oriented for the
third-party Content Provider (CP)/Service Provider (SP).

    The IMS may provide the SE capacities such as IMS session control and presence. It can also act as the SE of the mobile network.

2.7 IMS and FMC
FMC enables users in both fixed and mobile networks to enjoy the same services. Technologically, FMC may include a variety of service binding, terminal convergence, service convergence and network convergence. It involves convergence at layers of service, control and access terminals. IMS is suitable for realizing the FMC at the control layer and acts as the technological basis for convergence of fixed and mobile core networks. However, the realization of FMC is not necessarily based on adopting IMS. Without IMS, operators can still provide users with FMC service experience by other means.

    Considering there are more than one existing core control network, FMC may be fulfilled at the service and access terminal layers first, which requires the cooperation of IMS. In the long run, IMS will gradually become the unified IP multimedia services control core after the fulfillment of FMC.

    As an important part of the NGN, IMS represents a good prospect by providing diversified and flexible IP multimedia services and fulfilling FMC. However, the development of IMS is still confronted with challenges. Taking a general view, there are many problems preventing the practical operation of IMS. For examples, the service positioning of IMS in the fixed network is not clear, the varieties of services are not profuse, and the application of the IMS in the fixed network is limited by the access control capability of the network. Moreover, IMS technology itself is challenged by other Internet technologies and business models such as P2P. In other words, IMS is currently in the initiation stage, and needs further research and experiment.

3 Service Layer Convergence

3.1 Definition of Service Layer
With continuous convergence of the telecom network and Internet, functions of the service layer are expanding from the development, execution and management of traditional telecom value-added services to the development, execution and operation management of application services including telecom value-added services, content provisioning, content distribution and Internet applications.

    However, the service layer can still be divided into the service provisioning system and the service management system. The service provisioning system includes both the service execution and development environments. Application service logic is compiled in the service development environment and loaded into the service execution environment. Application services operate in the service execution environment, and visits and invokes the network capabilities of the control layer. These capabilities include traditional telecom network capabilities, as well as content distribution network capabilities. The service management system provides managerial functions for the process of service development and execution.

3.2 Driving Forces of Service Layer Convergence
    (1) There are demands of experiencing customer-oriented binding services and unified service, such as bound and combined services and charging packages, unified service accepting, unified service portals, unified bills and unified customer care service.

    (2) Operators will make full use of their advantageous resources to meet customers’ demands of integrated service experience. The integrated services include traditional telecom value-added services that integrate multiple network capabilities such as voice, video and data and that are available in both fixed and mobile networks, as well as converged services that integrate telecom value-added services with Internet and enterprises applications, such as FMC-based Centrex service, unified messaging service, inter-role communications embedded into online games (such as video group communications), and IPTV combined with incoming call notification.

    (3) Operators expect to rapidly and flexibly meet various demands of customers with a low cost. Now, it is difficult to predict the future demands of customers. On the other hand, the highly competitive market urges operators to make rapid and flexible response once they find out market opportunities. However, the network resources and service system owned by operators are limited. It is neither practical nor economical to build up a new service system for a new service when a market demand is discovered. Therefore, it will be an important means of gaining edge over competitors to build up a reusable and scalable service system with clear architecture and converged functions, and thus quickly meet the flexible and ever-changing demands of customers with a low cost.

3.3 Definition and Development Status of Service Layer Convergence
The service layer convergence includes service provisioning system convergence and service management system convergence. It aims at fulfilling unified service management and provisioning functions, and obtaining reusable and scalable system architecture.

3.3.1 Service Provisioning System Convergence
Different from the existing service provisioning system (such as the fixed intelligent network), the converged service provisioning system can offer a unified service provisioning platform, and use the capability of simultaneously accessing and controlling multiple networks to provide users from different networks with trans-network converged application services that integrate voice, video and data. There are four types of technologies for implementing the service provisioning system convergence: the Integrated Intelligent Network (IIN) technology, SIP service provisioning technology, Parlay/OSA/OMA open service interface technology, and Service-Oriented Architecture (SOA) technology (including Web services middleware). Each of the above four technologies have its own features in system architecture.

    (1) IIN
    The core of IIN is the integrated Service Control Point (SCP), which may simultaneously support the Intelligent Network Application Protocol (INAP) and Customized Applications for Mobile Network Enhanced Logic (CAMEL) Application Protocol (CAP). The IIN may provide FMC-based intelligent network services such as combined Centrex through visiting and controlling the Service Switch Points (SSP) in the fixed and mobile networks respectively. The IIN technology is almost mature, but its capability is limited to providing voice value-added services, and it does not support opening to the third party.

    (2) SIP-based Service Provisioning Technology
    A typical representative of the SIP-based service provisioning technology is the service provisioning architecture in the IMS system. The architecture is implemented by cooperation of the Call Session Control Function (CSCF) and the SIP Application Server (SIPAS). Since SIP (including the IMS core control architecture) enables itself independent from any access modes, and has good scalability, this technology can support both the voice and video real-time session control functions and the presence and IM control functions. Moreover, the SIP session control mechanism provides the SIP AS and CSCF with opposite repetitious interaction. Therefore, the SIP-based service provisioning technology can offer service-and-control-separated,
FMC-based converged services that can combine voice and video service functions with data service functions, such as presence and instant messaging, and that is flexible to trigger and may be easily embedded into other services. Products based on this technology are still under development and being perfected. The SIP-based service provision technology does not support opening to the third party either.

    (3) Service Provisioning System Based on Parlay/OSA/OMA.
    The core features of this technology are that it defines abstract service interfaces that are not related to specific technical details of the bottom network, and that it supports the authentication function that is necessary for opening to the third party. With these interfaces, telecom operators can offer their network capabilities to the third-party service providers with security and manageability, and thus support their service development and operation. Therefore, this service provisioning technology can implement converged services with multiple network capabilities such as independence from specific networks and integration of voice, video and data, and can thus realize the combination of telecom services and the third-party Internet applications.

    Moreover, the architecture of this service provisioning system is universal and may adapt to the quick changes of the bottom network through the addition of interface adapting modules (such as Parlay SCS) to meet the flexible requirements on service provisioning.

    (4) Service Provisioning System Based on SOA Architecture and Web Services Middleware

    In the SOA system architecture, all related service functions are standardized and defined as service modules. Through defining the combination sequence of these service modules (including control stream and data stream) and a dynamic module discovering and acquiring mechanism, and accomplishing information interaction through Web services middleware, which is more suitable for the Internet environment, new application services can be not only produced automatically at real time, but also deployed and operated in a short time. This architecture can better meet the ever-changing service demands, as well as realize the flexible reuse of existing functions and rapid service development and deployment. The SOA architecture and Web services middleware have been widely used in enterprises and IT areas, and have started its application in the telecom field. However, in order to meet the requirem

[Abstract] Convergence is the main trend for future network development. The multi-service bearer network, built upon the Internet Protocol/Multi-protocol Label Switching (IP/MPLS) technology, is a converged network with Quality of Service (QoS) guarantee for business and enterprise customers. The integrated and unified service provisioning platform (including both service execution and development enviroments) can provide integrated audio, video and data services over multiple access networks. However, achieving network convergence and service convergence are long-term goals, which can only be realized step by step after comprehensive consideration of many factors like market demands, technologies, services and costs.