After many years of false starts fiber to the home (FTTH) is finally taking off worldwide. The combination of voice, data and video, known as the triple play, is often mentioned as the driver behind FTTH adoption but, in reality, FTTH is being driven by much more. While triple play can certainly be offered on these networks, the primary motivator for FTTH is to create a super broadband infrastructure on which new services and applications can be created and delivered for the rest of this century. It is obvious that although service and application innovations on the internet, such as, e-commerce, distance learning and video-on-demand (VOD), have already changed the way we live today, these innovations are only the beginning of much more to come. We can only guess as to what new innovations will come, but we can be certain they will come and the key to supporting these new services and applications will be the ability of broadband access networks to scale.

In this context TDM-PONs, whether APON, BPON, EPON or GPON, are short-term solutions capable of meeting our needs only for a few more years as they have a very limited ability to scale. On the other hand, Switched Ethernet, based on point to optical connections, can easily meet our present needs and, most importantly, is the only access infrastructure capable of scaling with no known limit to meet the needs of the future. Furthermore, TDM-PON and Ethernet access networks may be comparable in cost today, but the cost trends of Ethernet access, taking into account system electronics, optoelectronics and outside plant facilities, will inevitably make it less expensive than TDM-PON networks. Ethernet has demonstrated over and over again its ability to deliver record-breaking performance at record breaking cost points. So, today as FTTH deployment finally takes off, we make the case that TDM-PONs are past their prime and Ethernet access based on point-to-point optical connections is the optimum optical access infrastructure with which to move forward.

The TDM-PON concept, conceived in the late 1980s, was an optimum architecture for that time. It addressed critical cost and technology issues while also promising to deliver the service expectations of the time. Since that time the landscape has changed dramatically with the emergence of the Internet. The Internet has created limitless possibilities in terms of new services and applications that were impossible to imagine in the late ‘80s. However, realizing this potential depends on our ability to scale the network with an appropriate networking technology. Fortunately, Ethernet has emerged to meet the challenge.

The Emergence of Ethernet

In the late 1980s, Sonet and ATM were emerging technologies that were expected to dominate transport and switching in the MAN and WAN. Instead, Ethernet, initially considered suitable only for data traffic in the LAN, has expanded beyond. With new carrier class capabilities and new levels of performance, it is now displacing ATM and Soent in the MAN and WAN as the dominant new aggregation and transport technology.

Furthermore, the first deployments of Ethernet in the metro are rapidly being upgraded to “carrier-class” Ethernet, a.k.a. Carrier Ethernet, from Enterprise-class Ethernet. Carrier Ethernet embodies five key attributes not found in Enterprise class Ethernet: 1) “Hard” quality of service (QoS) capable of guaranteed bandwidth and latency, 2) Scalability in several dimensions, including bandwidth, number of users supported, services and services per user, 3) Resiliency capable of 50 msec restoration, 4)TDM support that enables Ethernet to carry legacy TDM services, 5) Operations, administration and maintenance (OAM) with network and service management features that exceed the capabilities of legacy TDM networks.

Using Carrier Ethernet in conjunction with VLAN and MPLS techniques enables a great deal of flexibility in segregating traffic into tiered layer 2 virtual private networks (VPNs) and managing multiple traffic streams within each VPN with its own guaranteed QoS. This can be done on an individual link or through a complex interconnected network. For example, with guaranteed QoS, voice services can be supported with the performance and predictability of voice on a circuit switched network. These features have been integrated into Ethernet, while still preserving the ease of deployment and plug-and-play characteristics that are inherent to Ethernet.

Since the late ‘80s, Ethernet has moved from typical speeds of 10 Mb/s to typical speeds of 100 Mb/s, 1 Gigabit and 10 Gigabits drastically increasing the networks ability to deliver services at real-time speeds. Today, 100 Gigabit speeds are being explored.

Metro Ethernet services to business locations are already based on switched Ethernet. DSL Networks are also moving to Ethernet aggregation with GigE transport to the Node (FTTN) and curb (FTTC). FTTH based on switched Ethernet therefore becomes a natural extension of these networks.

Ethernet component volumes are driven by extensive deployment of LANs, business FTTP Metro Ethernet, and DSL aggregation, which dwarf TDM-PON component volumes and give it an inherent price advantage.

Finally, Ethernet also has operating cost advantages because of its inherent plug and play characteristics and the fact that it is a well-developed standard, with established interoperability and trained personnel readily available.

Bandwidth Demand and Scaling

Nominal broadband access speeds over DSL and cable have been growing at roughly 2x every two years. Continued bandwidth growth at this rate, with current nominal bandwidth levels of 4 Mb/s will create a need for 1 Gig of bandwidth per subscriber in about 15 years. Until now, users have been increasing their access speeds mainly for browsing the internet, but one or two killer applications can easily accelerate the bandwidth need requiring the access network to scale much sooner.

Scaling TDM-PONs vs. Ethernet

The primary purpose of a TDM-PON protocol is to function as a traffic manager that regulates bits on the point- to multi-point, shared-bandwidth optical connections created by the splitter-based architecture. It must abstract point-to-point connections on the point to multi-point physical layer while maintaining security in a broadcast environment. This adds significant processing complexity and creates many compromises--most critically, an inability to scale.

For example, when the nominal CIR to each customer increases to 100 Mb/s in a 1x32 TDM-PON, the system would have to operate at an aggregate speed of 3.2 Gigabits and, in the future, when the nominal CIR becomes 1 Gigabit, a TDM-PON would have to operate at 32 Gigabits. And, so on. These extremely high speeds are unrealistic operating speeds for a TDM-PON from both technical and cost perspectives. Besides requiring extremely fast processing, fundamental constraints are in the way. Every time the bit rate is doubled, the power budget of an optical system goes down by 3 dB. With the loss through the splitter (17dB for a 1x32) the power budget for TDM-PONs will inevitably become a fundamental constraint. In fact, it was an issue for GPON necessitating the incorporation of forward error correction (FEC) that has created its own set of complex issues.

Because PON solutions are built for a particular speed, high bandwidth demand by just a few users would force an upgrade of the whole PON system, including the optical line terminal (OLT) and every attached CPE. The alternative of reducing the split ratio is at best, a temporary measure because, in the limit, TDM-PONs turn into point-to-point fiber systems making the TDM-PON protocols redundant and a liability because of its complexity.

Scaling Switched Ethernet

By contrast, Switched Ethernet systems are inherently simple. They operate on a point-to-point optical circuit over dedicated fibers or over shared fiber with a wave-division multiplexed (WDM) overlay. Upgrading bandwidth from 100 Mb/s to 1 Gigabit today is a switch port change at the central office only for customers requiring the higher speed connection. However, as Ethernet is already moving from 10/100 Mb/s ports to 10/100/1000 Mb/s ports, even port changes will become a thing of the past.

Comparisons between Ethernet access and TDM-PON access have too frequently focused on “Passive” vs. “Active” with TDM-PON being positioned as passive and Ethernet as active. However, this discussion is misleading as Ethernet can now be deployed economically in both passive and active configurations due to the advancements in optical technology and reductions in optical component costs.

There were three key drivers that led to the TDM-PON architecture in the late 1980s. First, fiber and optical transceivers were very expensive, requiring an architecture that shared them where possible. Second, there was a desire to eliminate powered RTs because of reliability concerns and power availability. Third, there were concerns about terminating and managing a large number of fibers in a CO. Today, the cost of fiber and optoelectronics is a small percentage of what it was in the late 1980s. The incremental cost of fiber,* which was more than $300 per mile then, can be as low as $30/mile today. Laser transceivers that were thousands of dollars are now less than $50. These new cost points make it possible to build point-to-point fiber networks without intervening active RTs over much longer distances than previously possible. [For example at carrier serving area (CSA) distances point-to-point fiber solutions are comparable in cost to PON today.] Finally, significant improvements in fiber management since the late 80s now make it much easier to terminate large numbers of fibers in a CO.

As the costs of optical fiber and optical components have dropped dramatically, WDM has also matured through extensive deployment in long distance and metro applications. It is emerging as an alternative that enables the same “passive” physical topology of a TDM-PON, with fewer fibers in the feeder plant and terminating in the CO. Of course, “active” RTs are still an option and, with the MTBFs of today’s OSP electronics, they may be the best option, depending on the availability of powering at the feeder/distribution cable plant interface.

Therefore Switched Ethernet solutions can be deployed in both “passive” and “active” configurations while at the same time bringing a level of performance and scalability to the access network that is unimaginable with TDM-PONs.

Other advantages of Ethernet over TDM-PONs

It is beyond the scope of this paper to go through every advantage of Ethernet over TDM-PONs but it is important to highlight one key advantage that has not been widely recognized as yet.

Construction Techniques

FTTH costs are dominated by OSP construction. If cables with factory terminated multi-fiber connectors can be used in place of fusion splicing, it can lower deployment costs and deployment times significantly. However, because of the higher losses and reflections of mechanical connectors, the scope of using this technology will be limited in the case of TDM-PONs. On the other hand, because there are no splitters in a switched Ethernet system this technology can be used extensively to lower construction costs and deployment times.

Summary

TDM-PONs were conceived in the late 1980s at a time when the expectations for bandwidth scaling were much lower. TDM technologies dominated the network, and optical technologies were primitive and expensive. Since then, the Internet has emerged and has become the platform for limitless new services and applications while Ethernet has emerged as a proven networking technology, with a demonstrated ability to deliver new levels of performance at record-breaking cost points. It has been rapidly displacing TDM services to businesses and is becoming the aggregation layer for DSL networks. In the same time period, optical fiber and optical components have matured and dropped significantly in price, and WDM is emerging as a viable option for the access network. These factors now position switched Ethernet in both passive and active configurations to deliver performance levels at cost points not possible with TDM-PONs for FTTH.

TDM-PONs were a concept that addressed the issues and expectations of the late 80s, but they cannot scale to meet the needs of the future. FTTH based on switched Ethernet provides unmatched performance and scalability and has become the optimal solution for maximizing revenue generation potential for the future.

*It should be kept in mind that trenching and installations costs are about the same regardless of the actual cable size. Therefore, for new deployments of feeder cable, it is the incremental cost of fiber that is the key factor that needs to be considered.

Ram A. Rao is Director of Market Management at World Wide Packets.

Visit World Wide Packets online.