Bookmark & share

By Zhang Chengliang

The explosive development of broadband services has created both opportunities and challenges for existing transmission networks. Service pattern changes and the unrelenting emergence of new technologies have pushed them to an unavoidable transition point. In order to exploit possibilities while ameliorating anticipated difficulties, the future direction that transmission networks must follow consists of service-orientation coupled with the selection of suitable network construction technologies.

Different Technologies for Different Scenarios

Forecasts indicate that broadband services will increase by 50% each year over the next 5 years. Backbone broadband traffic will surpass 50T by 2010, and roughly 97% of this will comprise broadband data services. Optical transmission networks clearly reflect the rapid elevation in traffic, not least because the demand for backbone WDM system expansion invariably exceeds supply. While 2.5Gbit/s and 10Gbit/s interfaces currently form WDM backbone networks, the next 2 years will witness the commercialization of 40Gbit/s interfaces. The swift development of new MAN data services in turn necessitates DSLAM acceleration.

Rapid traffic growth and the All IP trend have inexorably driven optical transmission networks to a turning point. It is evident that each layer should utilize different technologies that can holistically be evolved to accommodate different scenarios and processes. For example, IP technology has already been adopted in various network layers – such as in the core layer - and OTN interfaces will soon be deployed. Bearer networks generally adopt FE/GE, or even 10GE, and routers currently employ POS interfaces.

After a lengthy inception process, MSTP has emerged as a competent and cost-effective solution for Ethernet and TDM co-transmission, and is now widely deployed in China. MSTP will exist for some time during All IP evolution, largely because low-speed interfaces such as 2Mbit/s and TDM will continue to exist in both access networks and MANs. Moreover, MSTP evolution will accompany a gradual decrease in TDM processing ability and a steady escalation in packet processing ability.

ASON has been widely adopted in China's telecom networks, especially in MANs. Numerous international carriers have successfully implemented ASON networking and applications, and in doing so have accomplished its correct orientation. ASON should mainly focus on dedicated line networks for VIP customers, traditional TDM and the IP link networks with high QoS demands. Key elements in developing ASON are inter-provincial transmission network and ENNI applications. ASON's successful application in the provincial backbone network will stimulate its development in other network layers. The ENNI interface can realize interoperability among different manufacturers. Therefore, carriers need not worry about equipment selection and the networking cost is reduced.

At present, the point-to-point WDM linear system is the most important application among WDM networks, and the 80×10Gbit/s WDM system represents its mainstream form. At the same time, the requirement for 160/192-wave systems has come up. In regards to 160-wave transmission, considering the lack of L band and its performance, C band with a spacing of 25GHz is more appropriate. In addition, the N×40 Gbit/s system is emerging. The development of the N×40 Gbit/s system is driven by its capability of supporting 80 wave with a spacing of 50GHz, that is, 3.2Tbit/s (80×40Gbit/s) transmission capacity.

Broadly speaking, the All IP trend urgently demands a high-quality, reliable and flexibly configured transmission network to maximize opportunities, while reducing the risks and pitfalls that would otherwise be faced.


Mature OTN Technical Conditions

OTN technology describes an important transmission-layer technology oriented to high-speed next-generation transmission networks. Compatible with SDH and SONET, OTN has inherited their advantages, and -- after 6 years' development -- is now mature for commercial application. On the optical layer, an OTN can enable large-granule processing in a similar manner to WDM. On the electronic layer, it adopts asynchronous mapping and multiplexing, thus facilitating the application of cost-effective time division cross connection technology at key times.

OTN interfaces have already been widely employed in WDM equipment. OTNs current developmental focus will lead to application in MANs and long-distance WDM networks, while its future direction will see it execute control-layer functions to support distributed ASON control.

The inheritance of SDH technologies has enabled OTNs to meet requirements for broadband service development. OTNs can adapt to any customer service, including SDH, ATM, Ethernet, SAN, and Video. It applies asynchronous mapping and multiplexing as opposed to overall network synchronization, which eliminates synchronization limitations. In addition, it is more suitable for handling GE and 10GE services.

Given its ability to support exponential rises in broadband traffic, OTN technology has been accepted in commercial applications and its further opportunities largely rest with the GE interfaces that are widely used in MANs, and also 10GE interfaces which will enjoy wide future application. As an OTN processes large rather than VC-4/12 granules, switching capacity is theoretically enhanced. Such an improvement can better meet the requirement of service development. With the new intelligent ASON/GMPLS feature, an OTN network can offer different levels of QoS guarantees, thus effectively reducing protection costs in MESH networks and considerably increasing its attraction among carriers.

In the future, carriers and equipment vendors must focus on the interworking of multi-vendor OTN systems. As in SDH systems, interworking should occur through at least a single channel where the NMS does not need to be considered. At scale tests should be deployed in OTNs to enable interconnectivity and interoperability between different vendors' equipment. In interworking based on the G.709 Recommendation, FEC interworking is particularly important, and OTNs must enable serial connections with sufficient monitoring applications.

Any technology requires considerable lead time from promotion to large-scale commercial application, and OTNs are no exception, given that the related technology still faces many hurdles. Firstly, SDH is still heavily relied on as GE or even higher speed processing depends on the direct connection of optical fibers or the WDM system. Secondly, OTN cross-connect equipment is relatively immature since cross-connect capacity must at least exceed that of large-capacity, cross-connect SDH equipment. Thirdly, there are only a few manufacturers who currently possess the capability of supporting OTN, and further ratification is required in interoperability terms. However, OTNs technological advantages and its conformity with service development trends evidence that OTN applications will start from cross-connect equipment in MAN and long-distance WDM system. The former will validate OTN's ability in cross connection and circuit reassembly, and the latter will validate G.709 interface's ability in overhead and interworking.


Real Implications of All-IP

Faced with network and service transformation, each carrier must possess a basic network with a stable architecture and scalability potential. Indeed, the huge upsurge in IP traffic together with the increase in folded network bandwidth requirements have been accompanied by a heavy upturn in investment and network construction frequency.

All-IP represents a general reference point since it signals that the future will fully embrace the IP mode. However, this does not mean that coming networks should necessarily form end-to-end IP networks, or that telecom networks' existing three-layer architecture will be replaced by router networking. The adoption of IP signal formatting also does not irrevocably imply end-to-end Ethernet transmission above any other format such as SDH or DTN. In fact, even now, core routers are interconnected via SDH-based POS interfaces. Moreover, transmission layer OAM is vital for ensuring end-to-end transmission and protection recovery. The All IP concept emphasizes the access layer and signal formatting, both of which are more relevant to service user perceptions, such as those associated with VoIP and IPTV.

The IP trend is bringing with it a rapid increase in traffic that necessitates large granular handling capacity, and IP networks good at handling small granules are no longer sufficient. A more promising future solution is represented by transmission networks capable of handling large granular and performing convergence, multiplexing and transmission functions.

The telecommunication industry has witnessed numerous technological innovations and rapid development. Yet, the established hierarchical structure of telecom networks remains intact. Any ambition to utilize one technology with which to face all challenges is simply an illusion. Given that the future direction for transmission network development in terms of network transformation should be service focused, no single'best' technology exists for carriers, just the most suitable one. Only those technologies that can assist them inherit legacy services, as well as driving evolution at a high cost-performance ratio, can guarantee market success.