Debunking Ten Myths

There are a number of industry myths surrounding both TDM-PON and switched Ethernet that have created confusion in the industry. This section attempts to debunk 10 industry myths and issues related to the two architectures.

Myth #1: Ethernet solutions do not have QoS.

QoS was not a part of early generations of Ethernet. However, with the emergence of Carrier Ethernet, Ethernet can now provide the QoS robustness of ATM-class services. With Carrier Ethernet solutions, it is possible to provide guaranteed bandwidth with CIRs and Excess Information Rates (EIRs) that allow the ultimate flexibility in creating, managing, and billing new services. Today, Carrier Ethernet is being deployed widely for business services and is enabling revenue-generating, differentiated services. Carrier class Ethernet is also necessary for triple play and is a key enabler for future residential services.

See the World Wide Packets white paper, Carrier Ethernet: Its Attributes and Opportunities, to learn more about this subject.

Myth #2: TDM-PONs are “Passive”.

In many scenarios it is most convenient to deploy the OLT and splitter at the same physical location as the OLT. This scenario for example occurs when the natural flex point for the fiber distribution cable is close to the CO or an RT. Another example is MDUs or MTUs with fiber to the apartment or business.

In these cases the TDM-PON is in effect an active deployment and optical splitters and the TDM-PON protocol only add an unnecessary layer of complexity and cost.

Myth #3: Switched Ethernet is “Active”.

As fiber and optical component costs have decreased, it is now possible to deploy switched Ethernet without any active components used in the outside plant for much longer distances at cost points not previously possible. At the current cost of fiber and optical components, point-to-point fiber connections between the CO and customer premises are possible for distances that cover the majority of loops deployed.

Where there is a need or preference for the physical layer topology of a TDM-PON, WDM is positioned to provide solutions at acceptable costs per subscriber. Such conditions might include very long feeder distances, circumstances where there is an existing cable with a limited number of fibers available for use as a feeder, inadequate conduit space, or fiber management constraints in the CO. It is important to note that WDM-PONs while enabling the physical layer attributes of a TDM-PON will easily exceed the 20-km distance limits of TDM-PONs.

Figure 3: Switched Ethernet Deployment Examples

Carriers can choose passive, active, or a combination of passive and active switched Ethernet options, depending on factors such as subscriber density, and feeder cable constraints.

Also to be considered is the fact that RTs can be deployed in rings that can provide redundant paths to protect against fiber path or electronic equipment failures. This configuration delivers levels of reliability and resiliency that are difficult to achieve without active RTs.

The choice between passive and active RTs weather using a TDM-PON or a switched Ethernet solution is most importantly a matter of the deployment scenario. For example in MDUs and MTUs powering is usually easily available and an Ethernet switch in the basement is an obvious choice over a TDM-OLT with splitters in the basement. For fiber-to-the-home (FTTH) applications the decision to install an active RT would depend on the ideal distribution fiber flex points and the level of customer concentration at those flex points.

Thus active RTs are not a requirement but it can be an attractive an option for Switched Ethernet solutions depending on the deployment scenario and carrier's preference.

Myth 4: TDM-PON networks can scale and are more flexible than point-to-point architectures.

TDM-PON networks have very limited scaling ability to meet increased customer demand. Because TDM-PON deployments have a fixed, shared bandwidth, service providers wanting to increase bandwidth per subscriber would have to upgrade the TDM-PON to a higher bit rate. For example, if every customer required a guaranteed bandwidth of 100 Mb/s, then a 1 x 32 PON would have to be upgraded to 3.2 Gb/s, and a 1 x 64 PON would have to be upgraded to 6.4 Gb/s. These upgrades if possible would be forklift upgrades involving a change-out of the OLT and every attached CPE.

It should be considered that there will be practical limits to upgrades by this method as the speeds required of a TDM-PON would inevitably become unrealistic. As an alternative the split ratio on a PON can be reduced, allocating more bandwidth per customer, but this would require a re-organization of the splitters and customer fiber connections, as well as potentially requiring the installation of new splitters and additional feeder fiber cable. In the limit, the system would turn into a point-to-point optical system, defeating the purpose of a TDM-PON and leaving the TDM-PON protocol as a liability.

With switched Ethernet, on the other hand, moves, adds and changes can be managed from a central location without the need to disrupt existing services or subscribers. The same switched Ethernet solution that provides 100 Mb/s to each subscriber can also support fully symmetrical 1 Gb/s, or even 10 Gb/s subscriber connections which. The connections for those customers needing additional bandwidth would be moved to a higher-capacity switch port if needed, and the CPE which could be supplied for 100 Mbits or 1 Gbit would be changed only if necessary. This is a level of scalability and flexibility that only switched Ethernet can permit and it is a level of scalability and flexibility that essentially renders the switched Ethernet architecture future proof.

Myth #5: TDM-PONs have superior revenue generation potential.

Switched Ethernet's fully symmetrical connections while easily enabling triple-play services today, will enable a full range of additional simultaneous services that TDM-PON's bandwidth limitations will render impossible in the future.

Figure 4 illustrates this point with potential services that we know off or can conceive off today. It is impossible to predict what the future holds in terms of applications but it is safe to assume that those applications will require increasing amounts of bandwidth. As switched Ethernet solutions are capable of scaling far in excess of the fastest TDM-PON implementations they are arguably the only network solutions capable of scaling to meet future requirements and generating the associated revenues.

Figure 4. Scaling to meet future demands

Myth #6: TDM-PON varieties are all standards-based and offer easy interoperability.

Every TDM-PON flavor-A-PON, B-PON, E-PON and G-PON-is governed by a different standard. Interoperability between these different technologies is impossible. E-PON does not even interoperate with other IEEE 802.3-based Ethernet devices.

In contrast switched Ethernet uses fully compliant IEEE 802.3 standard components to extend Ethernet to the subscriber in a point-to-point topology offering a fully symmetrical 100 Mb/s or 1 Gb/s of bandwidth to every subscriber. If additional bandwidth is required, 10 Gb/s standard-based Ethernet solutions (IEEE 802.3ae) are available, and efforts are underway to standardize faster speeds. Using IEEE 802.3 standard components results in very low-cost solutions due to the millions of ports of Ethernet connections already deployed worldwide. It also allows full interoperability with other standards-based Ethernet solutions from a number of vendors.

Myth #7: TDM-PONs have the lowest Capex costs.

Ethernet equipment costs are able to leverage the huge component volumes generated by Ethernet networks being deployed in LANs, MANs and WANs.

TDM-PON equipment on the other hand are able to leverage only FTTX deployments which are relatively modest in volume. In addition TDM-PONs require highly specialized equipment such as burst mode transceivers which are inherently more expensive than the optics required for Ethernet. The high complexity of TDM-PON components and modest volumes translates to an inherent cost disadvantage relative to Ethernet .

While TDM-PON technologies can be deployed with lower split ratios resulting in more available bandwidth per subscriber, this effectively increases the cost basis of the solution. For example, decreasing the split ratio of E-PON from 1:32 to 1:16 increases the available bandwidth per subscriber to just over 60 Mb/s, but significantly increases the cost basis of the solution, because the OLT port is now shared among fewer ONUs.

Myth #8: TDM-PONs have the lowest Opex costs

PON vendors assert that using passive splitters instead of active elements between the OLT and optical network unit (ONU) will yield cost savings by eliminating the need to power and service active components. The argument is that this lowers overall maintenance costs, because fewer powered elements will require service. But it is important to note that in any FTTP deployment (TDM-PON or switched Ethernet) it is powered ONTs/ONUs that make up the vast majority of the total deployed powered elements in the field, and they will require service technicians and truck rolls if they fail.

Even in a worst-case scenario, the total additional number of active sites is less than 1% more than a passive deployment for the same number of subscribers. This small percentage of additional powered RTs must be evaluated in the context of the unmatched bandwidth subscriber scalability, deployment, and topology flexibility of switched Ethernet, as well as its increased revenue potential.

Ethernet has been widely deployed in LANs for more than two decades with the result that there are large numbers of IT personnel who are familiar with the technology. It is also a technology that was developed from its inception to be a plug-and-play technology. This attribute has been maintained as the technology has evolved to support more complex operations.

By contrast, TDM-PON technology has seen limited deployment even though its basic concepts were developed more than 15 years ago. Carriers must train people to deploy and maintain these systems, which are inherently very complex with limited upgradeability, requiring frequent retraining.

Finally only switched Ethernet allows a service provider the flexibility to seize full advantage of both their optical or copper infrastructure and maximize their investment by offering any mix of services to any customer, business or residential, with one converged services network.

Myth #9: TDM-PONs are easy to test.

Troubleshooting fiber problems on the distribution side of a TDM-PON with an optical time domain reflectometer (OTDR) is not straightforward. The OTDR receives the reflected signature of all 32 distribution fibers overlaid on each other making it impossible to distinguish the signatures of each individual fiber. Testing would require a truck roll to the splitter point and testing the distribution fibers individually.

Myth #10: TDM-PONs have the lowest OSP construction costs

Total FTTP costs are dominated by outside plant and construction costs, and improvements in this area will remain critical to rapid and cost-effective deployment of FTTP.

TDM-PONs have very tight power budgets because of the loss through the optical splitter (17dB through a 1 x 32). Connector and splice losses become critical, constraining the ability to leverage multi-fiber connectors and mechanical splicing. With switched Ethernet, power budgets are not tight and are much more forgiving. This will enable a migration to construction techniques based on pre-terminated multi-fiber connectors and other techniques that will reduce both cost and deployment times.

Conclusion

Although the TDM-PON concept was once a powerful solution it is a concept that may have passed its prime. At this point in the history of technology development and price points, it is a less optimal solution that will require additional significant investment to correct later.

Switched Ethernet mitigates all complexity and scalability problems, as well as cost issues, with a mature technology that continues to excel. In the end, relying on a proven technology with hundreds of millions of deployments worldwide that is still growing rapidly is a far more reliable and future-proof prospect than one with a decaying lifecycle.

• Part 1:
  Why Rethink TDM-PON?
• Part 2:
  Debunking 10 Myths