WDM-PON for the MSO market
With WDM-PON, MSOs can leapfrog competitors and move beyond GPON.
Residential and business customers continue to demand more bandwidth and services, and MSOs are deciding how to respond. Most MSOs still rely on legacy coax cable plants and DOCSIS to deliver voice, data and video services, but they recognize the need to upgrade in response to competitive threats, as well as customer demand. With incumbent telephone carriers and other competitors rolling out fiber-based networks that are capable of delivering far more bandwidth than coax cable, MSOs need a new strategy.
In this article, we’ll look at how MSOs can use wave division multiplexed passive optical network (WDM-PON) to leapfrog their competitors and move beyond technologies such as gigabit passive optical network (GPON) to deliver scalable, dedicated and secure high-bandwidth services to residential and business customers.
Figure 1: Typical GPON deployment.
THE COMPETITIVE TECHNOLOGY LANDSCAPE
Fundamentally, the migration to HDTV and IP HDTV changes the game for residential deployment. Consumers require 20 Mbps per channel for uncompressed HDTV alone, and that’s on top of a competitive 10+ Mbps of bandwidth for Internet access and voice services.
Most incumbent carriers are saddled with legacy TDM-based networks. Companies like AT&T are using fiber-to-the-node (FTTN) technology as a bridge to future network upgrades and can provide up to 25 Mbps per household via ADSL2+ over copper. Verizon, on the other hand, has taken this a step further with its FiOS offering. Initially, FiOS started as a broadband PON (BPON) network, but it is being migrated to GPON. BPON and GPON are well suited for incumbent carrier networks because they are all TDM-based technologies supporting legacy POTS, RF video and data. In a few years, incumbent carriers may well migrate to 10GPON for even more bandwidth. At that point, MSOs will need an answer that goes beyond the current bandwidth limitations of DOCSIS.
In most cases, GPON uses a single feeder fiber to connect the Optical Line Terminal (OLT) to a remote splitter, which takes the signal coming from the OLT and replicates it onto 32 distribution fibers for service into a residential area. Typically, a single residence is associated with a distribution fiber, which is terminated by an optical networking unit (ONU). As a result, GPON’s 2.5 Gbps of bandwidth is shared among 32 households. The bandwidth need not be split evenly among households, but if it is, this provides about 78 Mbps per residence.
On the other hand, GPON doesn’t scale easily to support a future in which residential consumers will demand multiple HDTV channels and other services requiring roughly 200 Mbps per household. With HDTV and digital video recorder services becoming ubiquitous over the next few years, carriers will need to support as many as four simultaneous HDTV channels per residence, plus voice services and broadband capabilities in the 50-100 Mbps range.
Moreover, GPON is not very attractive to small- and medium-size businesses (SMBs). SMBs want cheaper alternatives to SONET-based services, which can cost thousands of dollars per month. While major telephone carriers tout SONET’s protected architecture, most SMBs care more about getting dedicated high bandwidth at a lower price than they do about SONET’s protection capabilities. Also, GPON doesn’t provide the secure, symmetrical bandwidth that financial and government customers desire. With GPON, customers are connected over a shared infrastructure, which poses a security risk and complicates network changes.
Figure 2: Typical WDM-PON deployment.
WDM-PON: THE MSO UPGRADE PATH
In this landscape, MSOs have a key opportunity to develop upgrade strategies that go well beyond what the incumbent carriers will offer. WDM-PON is an IP-based technology that today offers MSOs the ability to deliver 1.25 Gbps per customer, and higher-bandwidth versions will soon become available.
Architecturally, WDM-PON uses a single feeder fiber to take advantage of the same economics associated with traditional PONs; but logically, WDM-PON uses a point-to-point architecture. Therefore, it is far more scalable and secure than other PONs. Today, WDM-PON delivers 20 Gbps per fiber (1.25 Gbps dedicated per user on a 1:16 split). In addition, WDM-PON enables a dedicated wavelength for each user, ensuring the security that SMBs demand and providing greater provisioning flexibility – essentially, WDM-PON is a fat pipe that can support Ethernet, Metro Ethernet or TDM, depending on what the provider wants to offer.
Within the next two years, WDM-PON will offer 80 Gbps per fiber, which will allow 2.5 Gbps per subscriber on a 32:1 split.
WDM-PON is also less expensive to deploy, maintain and upgrade. For example, it uses colorless optics, which eliminates the sparing issue associated with typical DWDM network elements. In addition, if a bandwidth upgrade becomes available with better line terminals and network terminals, the provider can simply upgrade these without affecting service for existing customers. Also, WDM-PON channel plan changes can be accommodated simply by swapping the arrayed wavelength grating (AWG) at the OLT and remote splitter, rather than having to pull new fiber or replace the terminals themselves.
The architectural and operational benefits mentioned above make WDM-PON a good option for MSOs to serve multi-tenant units (MTUs) or multi-dwelling units (MDUs), SMBs and IP-based cell sites today.
With these significant advantages over GPON, WDM-PON seems like the obvious upgrade path for MSOs. However, it will take time to spread throughout the provider community. There are a few hurdles to overcome before WDM-PON sees widespread deployment.
First of all, WDM-PON’s density needs to improve. Current systems deliver 1.25 Gbps dedicated per subscriber for a total of 20 Gbps on a 16:1 split; 2.5 Gbps systems should be available within the next two years on a 32:1 split. While this density mirrors GPON, the OLT must increase the feeder fiber count to increase the subscriber count from 16 today to 640 subscribers or more in the future.
Secondly, although WDM-PON is significantly (four to five times) less expensive per Mbps, it is currently about three to four times the cost of GPON on a per-subscriber basis. This is a function of density, but it is also due to low production volumes as the technology makes its market entry. The cost will move toward parity with GPON over the next year or two.
The last and most significant hurdle is market readiness. Despite the need to offer more bandwidth, most MSOs have enormous investments in DOCSIS-based networks, and they will doubtless want to get maximum leverage from those investments until competitive pressures and market demand force the upgrade. In greenfield areas where it makes sense to deploy fiber from the start, WDM-PON can future-proof the network, but upgrades will be rolled out slowly.
These challenges will be overcome over the next few years, and WDM-PON will offer low-cost deployment, reduced maintenance costs and a transparent upgrade path to new levels of bandwidth. For MSOs considering ways to move into highly scalable fiber deployments that put them ahead of incumbent carriers in the era of IP-based television and higher-bandwidth Internet services, migrating to WDM-PON is a winning strategy.