Next-Generation Networks and Convergence

A next-generation network is a high-speed packet- or cell-based network capable of transporting and routing a multitude of services, including voice, data, video, and multimedia, and it is a common platform for applications and services that is accessible to the customer across the entire network as well as outside the network. The main physical components of the next-generation network are fiber and wireless media, routers, switches, gateways, servers, and edge devices that reside at the customer premises. A next-generation network is designed for multimedia communications, which implies that it has broadband capacities, multichannel transport with high data rates, low latencies (80 milliseconds or less is the target), low packet loss (less than 5%, with the target being less than 1%), and QoS guarantees.

A next-generation network has a worldwide infrastructure that consists of fast packet-switching techniques, which make maximum use of transport and provide great transmission efficiencies. A next-generation network involves optical networking. Today's electronic systems are going to be the bottlenecks to delivering tomorrow's applications, so we will see a replacement of the electronic infrastructure with optical elements that will provide end-to-end optical networking.

A next-generation network has a multiservice core, coupled with a multiservice intelligent edge. The application of next-generation telephony in the edge environment may potentially replace the existing architectures associated with the PSTN. Next-generation networks will be characterized by intelligent networking for rapid service delivery and provisioning. They will also have video and multimedia elements, to deliver the content for which the broadband infrastructure exists. Their access media are broadband in nature and encompass both wired and wireless facilities.

Next-generation networks stand to change how carriers provision applications and services and how customers access them. End-user service delivery from a single platform provides many benefits: It decreases time to market; it simplifies the process of moves, adds, and changes; and it provides a unique connection point for service provisioning and billing. Full-service internetworking between the legacy circuit-switched network and the next-generation packet networks is mandatory going forward. Next-generation networks also must be interoperable with new emerging structures, which implies that they have to be able to support the most up-to-date transport and switching standards. They also must support advanced traffic management, including full configuration, provisioning, network monitoring, and fault management capabilities. In a next-generation network, it is important to be able to prioritize traffic and to provide dynamic bandwidth allocation for voice, data, and video services, and this enables management of delay-tolerant traffic and prioritization of delay-sensitive traffic.

Convergence in Different Industry Segments

One of the central themes in next-generation networks is the notion of convergence. What convergence is depends on who you are and what segment of the industry you represent because convergence is actually occurring in a number of different areas. As discussed in the following sections, the concept behind convergence varies a bit depending on whether you're a service provider, an equipment manufacturer, or an applications developer. In the end, though, they all focus on one thing: bringing together voice, data, and video to be happily married at the network level, at the systems level, at the applications level, and at the device level.

Convergence in Transport

Convergence in transport refers to voice, data, and video traffic all sharing a common packet-based network, generally based on IP at present. This can apply to LANs, MANs, WANs, and wireless alternatives to each of those domains. From the standpoint of a service provider, convergence has to do with the idea of bringing together all the different technologies we've known in the past to come up with one common infrastructure, rather than working in today's world that consists of a separate voice network, the PSTN, separate data solutions, the Internet, IP backbones, Frame Relay, and ATMa variety of different packet-switched alternativesand then generally also separate networks engineered specifically for video and broadcast. The problem is that each of these represents a separate control plane, which means separate network management systems and procedures. It means technicians who are trained and knowledgeable in those specific systems, services, and products; and, of course, it means greater cost because there's duplication among platforms, and you have to make redundant components available and power backups to a large number of what, in essence, are overlays. So, to the service provider community, convergence means the convergence of the transport network.

Convergence in Systems

There has to be some convergence from the standpoint of the systems. To equipment manufacturers, this means creating systems that allow voice, data, and video traffic to all be commonly served through one device. In the context of next-generation network infrastructures, this most commonly refers to the use of softswitches, also known as call servers, to replace the use of traditional circuit switches, allowing the support of voice communications over IP backbones rather than the circuit-switched PSTN. From the standpoint of an enterprise network, this can also involve the use of IP PBXs at the customer premises or a service provider making IP centrex available to the enterprise, as well as integrated access devices (IADs) that combine voice and data traffic before presenting it to the external network.

Convergence in Applications

In the realm of applications, convergence refers to the integration of voice, data, and video at the desktop or in servers. Examples of this include integrated messaging, instant messaging, presence management, real-time rich-media e-learning and training products, multimedia sales presentations, and a variety of interactive programs, such as videogames. A host of such applications are unique and specific to various vertical industries, including health care, education, entertainment, government, and warfare.

Arguments for Convergence

One of the primary arguments for convergence has to do with cost reductions. As discussed earlier, packet switching is a more efficient approach than circuit switching for carrying conversations, particularly those that may be bursty in nature. With the continuing reduction in the cost of electronics, with the growth in traffic levels, and with the emergence of an increasing number of competitors, the price of delivering a packet on the backbone has been dropping by about 45% to 50% per year. Of course, we're now approaching a stabilizing point; nevertheless, packet switching has continually become more cost-efficient than the traditional circuit-switched network, or PSTN.

Another cost-reduction argument for convergence is in support of VoIP, and the major savings here occur through bypassing normal toll operations. Of course, this is most dramatic on international calls.

There are some other arguments for convergence as well. One is that you get improved productivity, from both user operations and the ICT staff. Another benefit is easier administration of the network. Again, you have a single network infrastructure and a single system that needs to be administered versus a number of separate platforms and networks consolidating the network management systems.

However, the real value in and the real argument for convergence lies in the applications. There are many synergies between converged transport, IP telephony, and converged applications. As functions such as instant messaging, presence, video communications, and streaming media merge with IP telephony, it becomes something greater than a cheaper voice solution: It becomes an integrated application. When you have an application that integrates voice, data, video, and streaming media, you most certainly require converged transport.

Regulatory Effects on Convergence

While we can achieve advances in technology very rapidly, what tends to delay their deployment is both a human factor and a political factor: We have to resolve the regulatory issues that surround the argument for convergence.

The integration of all communication modes under the control of IP has a powerful impact on regulatory models. Regulation has historically been quite different for various parts of the industry (e.g., voice versus broadcast versus cable). The way our regulatory system and regulations have been structured and still currently operate is that they are largely based on a service definition. If you offer telephone service, telephone regulations apply. Cable TV regulations apply if you offer cable TV service. But IP and converged networks part with the vertical model traditionally used in regulation. How do we regulate within a converged network? One idea considers a horizontal model, whereby regulations would be applied to the layers of activity versus the service definition. However, this is an area very much under hot debate, and ideas vary around the world. The main point is to remember that before you proceed to make decisions based on technology platforms and promises, you need to make sure you also consider what your regulatory agency is thinking and doing.

Converging Public Infrastructures

Today we have converging public infrastructures. The PSTN and the Internet are well on the path to convergence. There has been a steady, albeit slow, migration to packet-based networks. Today there are many networks running converged voice, data, and video over a common WAN infrastructure. Meanwhile, new developments, especially in the optical era, stand to alter the path of migration for all concerned. Truly magnificent new network designs are facilitated through the introduction of optical elements end to end. As a result, a new generation of networks is emerging.

If we look at where the PSTN and the Internet stand, side by side, we see that they both have the same goal (see Figure 10.1). Where originally the PSTN was a voice solution and the Internet was a data solution, they have now converged their objectivesthat is, both are striving to become high-speed networks capable of accommodating interactive multimedia applications. Both are also striving to apply QoS guarantees, for two reasons. First, the performance of applications is critical to customers, and without QoS guarantees, we are not able to guarantee the performance of interactive rich-media applications. Second, and perhaps more importantly in the mind of the service provider, having differentiated levels of service allows differentiated levels of pricing, and in an era where the cost of transporting a bit continues to drop dramatically, revenues have to be gained from new sources. One of those sources is potentially to offer customers a wide range of service levels and prices, from best-effort/lowest-cost to platinum service, with the highest QoS resulting in the highest cost.

Figure 10.1. Converging public infrastructures

Where there is a difference between the PSTN and the Internet is that in the 1980s, the PSTN took the approach of deploying ATM to support multimedia and QoS. Interestingly, the telco community is now looking toward one of the newer standards developed by the IETF, MPLS, as a means of reducing some of the costs traditionally associated with ATM. (MPLS is covered at the end of this chapter.) The ultimate goal, at least based on what we know today, is to marry IP and the optical realm via a control plane based on a more robust version of MPLS called Generalized MPLS (GMPLS).

The Internet has the same objective as the PSTN: to support high speeds, accommodate multimedia traffic, and support QoS guarantees. But the Internet community took a slightly different approach than the PSTN. Rather than rely on ATM and its QoS architecture, it began developing various architectures that address class of service (CoS) and QoS. Included in these is an overall umbrella architecture known as Integrated Services Architecture (ISA). ISA encompasses various solutions, including IntServ and DiffServ, along with MPLS; these strategies have the ultimate goal of blending IP and optical and migrating to GMPLS.

Convergence in the Service Environment

An important perspective of convergence lies in the operating assumptions of both service providers and enterprises. Let's look at both the traditional assumptions and current thinking.

Traditional Operating Assumptions

In the service environment, the operating assumptions are those of the traditional carrier mindset. Much of the carrier's business operations have been driven by industry regulations. Voice is a regulated environment, whereas data is not, and because more than 70% of carrier revenues come from voice, most of the traditional carrier's attention has been focused on that world. For example, network engineering has been focused on the requirements of voice traffic, and the network has been optimized to support voice. Before the current era of bandwidth abundance, controlling latency was an easier task than providing bandwidth. Bandwidth was considered to be at a premium, requiring high utilization and oversubscription by customers. Because QoS was a prerequisite to supporting voice, controlling latency was the dominant traffic-engineering issue. In the traditional world, all carriers and some enterprises desired measured use for network chargeback, so accountability was a critical concern. In terms of network ownership, traditional thinking was that carriers should own the facilities to maintain control. Finally, the basic business model was one where transport was considered to be the business.

In the traditional enterprise, the main goal of an enterprise network was seen to be the linking of enterprise sites. The network staff was generally divorced from the Web site developers. The main objective of network management was to manage the network as a cost center, with the major focus being on controlling the size of the "phone bill." In terms of network architecture, the traditional enterprise saw the use of separate voice and data networks, with most traffic being local to a nearby server. The key consideration was to provide high availability, first and foremost for voice, with data being a secondary consideration. Network optimization activities were supported by predictable traffic, predictable service providers, and predictable rates and tariffs. The preference in terms of network infrastructure was to build private networks; they were the norm and favored over public networks, whose use was the exception.

Current Operating Thinking

The assumptions at play with the new enterprise are vastly different from the traditional thinking. Today, using the public Internet for communications is considered as important as using enterprise networks. More often than not, the networking staff works with the Web site and e-commerce operations. The new network management objective is to manage the network as an application-enabling infrastructure, with flexibility being more important than cost. The new network architecture consists of converged voice/data applications and transport, with most traffic destined for remote servers. Given the increasing emphasis on data and multimedia, data network availability is today the key consideration. Network optimization, however, is more challenging and can be said to be nondeterministic due to unpredictable traffic, changing service providers, and changing pricing. When it comes to network infrastructure, the preference is for utilization of public networks and outsourcing, relying on private networks only where necessary.

In the mind of the new-era service provider, voice, data, and video in converged networks are all just bitsthey are all the same thing to service providers and regulators alike. However, although regulatory changes are indicated, they have yet to be fully articulated and instituted. The revenue stream is seen to come primarily from data and advanced applications involving multimedia. Voice is likely to move to wireless or may become "free," as part of IP service. With the focus shifting to data, the main network concern is to optimize the infrastructure for IP traffic. Traffic engineering in the current environment is governed by the philosophy that providing bandwidth is easy but controlling QoS is much more difficult, with the need to provide various levels of service designed to control latencies, losses, and bandwidth allocation. In terms of bandwidth, optical bandwidth has been driving down the cost and price of long-haul WAN transport. Lots of bandwidth, made available through optical technologies such as DWDM, beats the complexity of QoS-based service levels. Many of today's strategies therefore involve throwing bandwidth at the problem. However, QoS is increasingly emphasized as one of the main objectives behind building next-generation networks. When it comes to accountability, usage-based chargeback is being replaced by multiple flat-rate service levels. From the standpoint of network ownership, today extensive wholesaling and reselling of other carriers' facilities replace the view that complete network ownership is required for success. Finally, the business model is also changing: New-era service providers don't want to be just transport businesses any longer.

In the new service environment, there is a growing commonality between the network infrastructures of both service providers and enterprises. Fewer and fewer networks will be 100% facilities based. The service provider hosting sites are looking a great deal like enterprise data centers. As discussed in Chapter 11, service providers and large enterprises alike are taking advantage of many developments in the optical realm, including dark fiber, wavelength services, and WDM. There is a common emphasis on user service-level management, accounting, and rapid deployment. Finally, IP and Ethernet are becoming more pervasive in both worlds.