Anda di halaman 1dari 5

White Paper

Fiber Architecture for Rural FTTx Deployments


Introduction
Broadband service for rural America has been an intense topic of conversation and debate for a number of years. The American Recovery and Reinvestment Act of 2009 helped to jump-start many rural infrastructure builds for advanced communications networks but since the announcement of those broadband stimulus award winners, it has become apparent that there is still a significant gap in broadband services available to most citizens of rural communities. While the debate continues on who should serve these communities (municipal utilities, electric cooperatives, cable television system operators, rural telcos, etc.), the need for an economical, future-proof architecture is clear.
In order to bring service to these rural areas, there are several ber distribution plant architecture options available: two based on splitters, a third based on optical taps, and of course, point-to-point (P2P). Communication providers are probably most familiar with splitter-based networks because of the large number of telco passive optical network (PON) deployments, most of which use centralized splitting architectures. There is an alternative distribution architecture, tapped distribution, which simplies installation relative to a splitter-based approach. By using optical taps instead of splitters, service providers have reduced the ber counts and structure requirements of the distribution plant and also reduced the time associated with plant installation. This paper is intended to present a ber architecture that is optimal for deployment in rural communities. This passive architecture is compatible with all existing FTTx protocols and is independent of the electronics chosen for central ofce/headend and subscriber deployment. The following sections will compare two distinct architectures based on splitters and taps, and examine three key areas of difference between them, including cabling/enclosures, taper/splicing management and access splicing. To avoid the many assumptions needed in a network cost model, the discussion will focus on the cost differences in the three areas, using a model serving area of 256 homes.

FTTx Distribution Architectures


Centralized and Distributed Split
Splitter-based FTTx architectures are a compromise between cost and the exibility of running ber to every subscriber location. Instead of running ber from the central ofce/ headend to every home like in a P2P arrangement, which is extremely expensive, splitter-based architectures run a single ber from the headend/central ofce out to a central distribution point in the network, and via an optical splitter/combiner from that point individual bers are run to each subscriber location (hence the description point to multi-point). This is the physical layer architecture dened in the ITU PON and IEEE PON standards. As shown in Figure 1A, feeder bers are run to a cabinet near the neighborhood to be served. Each feeder ber terminates on a 1x32 optical splitter in the cabinet, which can in turn connect to up to 32 distribution bers. Typically, the cabinet also provides for managing the ber cables, splicing the individual bers and for terminating the distribution bers on the splitter ports through an interconnect panel. The interconnect panel gives an operator exibility in activation of the system and utilization of central ofce/ headend equipment. If an interconnection function is not desired, then the distribution bers can be directly spliced to the splitters output bers. Each distribution ber is then run from the cabinet to a drop pedestal location, and through a drop ber to a subscriber location to serve a single customer. The architecture provides a splitter port and a dedicated ber for every subscriber location in the serving area.

www.commscope.com

Alternatively, instead of a centralized splitting implementation (as shown in Figure 1A), multiple smaller splitters, 1:4 and 1:8, for example, can be distributed within the plant (hence the name distributed split Figure 1B). This reduces the requirements for the splitter cabinet, but it also eliminates the original benets of the splitter-based approach - removes the dedicated bers from each subscriber back to a centralized point, eliminates the advantage of having all the interconnections at a single point, and increases the network complexity without the benet of reducing ber count.

Splitter Based Architecture (Centralized and Distributed Split)


CENTRALIZED SPLIT
Advantages Disadvantages

standards based cost savings over a P2P architecture centralized cabinet for management of cables, splicing and terminating interconnection panel gives operator exibility a splitter port and dedicated ber for every subscriber
Advantages

costly entry point cabinet required taper splicing adds labor intensive cost and complexity taper splicing requires detailed record keeping separate mapping and increases design time and cost higher ber count cables required exiting the cabinet requires mid-span entry technique for larger count bers for cable access, increased time due to management of remaining cables
Disadvantages

DISTRIBUTED SPLIT

exibiltity in split ratios in serving area reduces splitter cabinet requirements

removes the dedicated bers from each subscriber back to a centralized point eliminates the advantage of having all the interconnections at one point increases network complexity without the benet of reducing overall ber count

TABLE 1 FIGURE 1 - FTTx DISTRIBUTION ARCHITECTURES A) Centralized Splitter PON Distribution Architecture
Subscriber Central Ofce/ Headend

1:32

B) Distributed Splitter PON Distribution Architecture


Central Ofce/ Headend

Subscriber

1:4 1:8

C) Tapped Distribution Architecture

Central Ofce/ Headend

Tap

Tap

Tap

Subscriber

Subscriber

Subscriber

www.commscope.com

Distributed Tap Architecture

Distributed Tapped Architecture

Advantages Disadvantages In a tapped distribution architecture, optical simplies installation eliminates the advantage of having all the taps are used to divide the optical signal interconnections at one point power instead of splitters, which eliminates reduces ber counts can be more expensive to deploy in dense the need for a centralized splitting location. urban and suburban areas As shown in Figure 1C, like the splitter-based compatible with existing FTTx protocols network, feeder bers are run to a serving independent of electronic chosen for central area. Instead of being terminated on splitters, ofce/headend the bers are carried into the serving area. eliminates the need for a centralized splitting location At a drop pedestal location, an optical tap is spliced in-line with a single ber in the cable. provides a tap port for every subscriber Like an RF tap, the optical tap couples (taps does not require a serving area entry cabinet off) some of the power from the distribution cable handling/management benet, and no cost to taper the cable ber, and divides it among the drops con nected to the tap. The tap value sets the loss standard design map used for splicing, information placed directly on the tap between the input of the tap and the tap out smaller central tube ber cable used, smaller diameter, put ports. The remaining optical signal power more exibility and lower cost is passed back into the same distribution ber pigtail splicing is optional, but not required via the tap through port, and is carried on to the next tap in the cascade. This is continued TABLE 2 until the link loss budget of a particular ber is consumed, at which point another ber in the distribution cable is used to create the next tap run. In general, 32 tap ports can be served Feeder Cable 72F over a single ber run. The architecture provides a tap port for every subscriber location in the serving area, and a single Splitter 72F Cabinet ber for every 32 subscriber locations. Splice

A) Splitter-Based

72F 72F
256 Home Subdivision

Architecture Impacts
Cabling/Structure
The biggest difference between the two architectures is certainly the cabling requirements within the distribution plant and the structure associated with the entrance to the serving area. With the 256 home serving area model, the minimum number of bers required in the splitter-based architecture is 256. This would not be implemented as a single cable, but instead as shown in Figure 2A, as a number of smaller cables running from the serving area entry point down different paths into the serving area. Four cables are shown in the Figure, but the actual number of cables used will depend on the amount of taper splicing (discussed below) desired by the system operator. In addition to the ber cables, eight 1x32 splitters are required to serve the area. Because of the number of ber cables and bers required, and the number of splitters, generally some form of structure such as a cabinet will be required at the serving area entry point to house all of these components as well as facilitate easy access. These types of cabinets can cost thousands of dollars (w/o splitters), not including the cost of the mounting pad or the labor associated with its installation. In comparison, the tapped ber architecture does not need a serving area entrance structure (cabinet) since the bers

Feeder Cable

4F

Splice

B) Tapped

4F
256 Home Subdivision

FIGURE 2 - FIBER CABLING AND STRUCTURES

feeding the area can be brought directly into the serving area and there are no splitters to house. Based on a 256 home serving area, the tapped architecture would need a minimum of 8 bers to serve the area instead of 256 as is required in the splitter-based architecture. As shown in Figure 2B, two 4 ber cables can be run directly into the serving area, avoiding splicing at the serving area entry point, and fed down different paths within the area. The reduced ber count offers a signicant savings as a 4 ber count cable costs about $350/KFt less than a 72 ber count cable. Total savings depend on the length of ber runs to the drop points, but could easily exceed $10,000 per serving area.

www.commscope.com

Taper Splicing
Because the cable used in the tapped architecture is so small there is a huge cable handling/management benet and no cost associated to transition (taper) the cable to a smaller cable as bers are used up serving customers. Because of this, an operator can use the same 4 count ber cable throughout the serving area and need only stock a single distribution ber cable type. This, however, is not the case with the splitter-based approach since the number of bers within each cable is much larger. For example, if a 72 count ber cable is used on a particular path in the serving area, it can be more cost effective to taper splice that cable, say once 24 bers are used, to a 48 count ber cable than to use the 72 count ber down the entire path until all of the bers are used up. This is actually a way for the splitter-based architecture to offset some of the additional costs of its high ber counts. The trade-off is the cost of the larger ber cable versus the cost of the smaller count cable, plus the cost to install a splice closure, and splice the larger ber cable to the smaller ber cable. To offset even more cost, taper splicing can be done at natural branching points (side street off of a main street for example) in the distribution plant where a splice is already required. An example of taper splicing and taper splicing at a branch location are shown in Figure 3. In this example the cable is tapered every 24 homes passed (HP).

and cumbersome to place on the standard plant map, so a separate splice map is needed. A portion of a splice map is shown in Figure 4A. This map shows a single splice location (branch) and how a 12 ber cable is spliced to a second 12 ber cable, and a 6 ber cable. Remember this is for a single splice location. There will be many in a typical serving area, and the splitters used in the design along with customer access points will also be included in a complete map. The map requires additional design effort and costs, and must be maintained and updated if any eld or customer changes are made. This can add signicant cost to the design.
FIGURE 4 - PARTIAL SPLICE MAP

A) Splitter-Based

Splitter Cabinet

72F
24F Branch Splice Point

B) Tapped

10 BL

72F
24F

72F
24F Branch Splice Point 24F 24F

24F

72F
24F 24F

Because only a single ber is used on a particular tap run in a tapped architecture, the indication of which ber is spliced is readily done on the standard design map. As shown in Figure 4B, the information can be placed directly on the tap. A 4 port, 10 dB tap is shown in the Figure with the letters BL placed in the tap. This indicates that the tap is spliced to the blue ber in the cable. No splice map is needed, and the design time and cost associated with the map is avoided.

FIGURE 3 - CABLE TAPER SPLICING IN SPLITTER-BASED APPROACH

Access Splicing and Fiber Management


The large ber counts in a splitter-based architecture will generally drive the system operator to a stranded loose tube cable design. These cables have a central strength member that is surrounded with 5 or more buffer tubes, with each tube containing up to 12 individual bers, as shown in Figure 5A. The overall bundle may then be covered in armor, when required, and nally with an outer sheath, typically of polyethylene. To access the bers needed at a particular pedestal drop location, the cable must be entered with a mid-span entry technique. This technique removes a section of jacket and armor to expose the underlying buffer tubes. The reason this is done, versus simply cutting the cable, is that most of the bers will need to just pass

While taper splicing does save cable costs, it does add cost and complexity to the design process by creating the need for additional record keeping beyond the standard design maps. Since there is not a one to one mapping of bers from the larger cable to the smaller cable(s), a Splice map must be created which shows a splicer how to splice the cables together. In other words a diagram is needed that shows which cables and which bers are connected to which, and in what splice location it occurs. This must be done for each taper and branch point in the distribution network, as well as for the splitters in the serving area cabinet. This information is too detailed

www.commscope.com

through the access point to the next location and only a small number will actually need to be spliced to drop cables. Cutting the cable would mean that all bers would need to be spliced instead of just the bers intended for drops. Once the buffer tubes are exposed, a single buffer tube is also accessed via the mid-span technique to expose the individual bers in the tube, which will usually be 12 in count. Assuming that a pedestal location serves 4 homes, 4 of the 12 bers would be cut and directly spliced to drops or to pigtails, which in turn would be connected to terminated drops. Figure 5A also shows access splicing in a closure with pigtails, and a simplied version (2 tubes) of a stranded loose-tube cable.

must also be managed and stored. This along with the physical size and limited exibility of the larger cable size increases the time associated with and installation, and the components/ features needed in the associated ber closure. Like the splitter-base architecture, the cable used in a tapped distribution architecture would also need to be accessed via the midspan entry technique. However, because of the lower ber counts needed with this architecture, a smaller central tube ber cable design can be used. A smaller count cable means a smaller diameter and more exible cable which make access, prepping and handling of the cable easier for the installer. As shown in Figure 5B, a central tube cable consists of a single buffer tube, holding up to 12 bers, that is wrapped in armor when required, and then an outer jacket. Once the buffer tube is exposed, it is also accessed via the mid-span entry technique. As shown in Figure 5B, a single ber is then cut and 2 splices are made; one to the input of the tap, and the other to the output of the tap. Pigtail splicing is not required since the tap can be congured with connectorized drop ports. The remaining 3 bers are manFiber Closure aged and stored in the tap enclosure. Since there is only a single tube in the design there is no other buffer tube TAP management required. With its smaller cable size and ber count, the tapped architecture reduces the access installation time, compared to the splitterbased approach, by eliminating 2 splices, and the need to handle and store additional buffer tubes.

FIGURE 5 - CABLE ACCESS AND SPLICING

Again because of the ber count and cable design needed for splitter-based architecture, the installation at an access point consists of not only splicing 4 drops (or pigtails), but also managing and storing the remaining 8 bers from the accessed buffer tube. Likewise, the additional buffer tubes in the cable

Summary
Several ber architectures have been developed to support FTTx deployments. While centralized and distributed splitterbased architectures are suitable for dense urban or even suburban environments, they are not optimal for rural broadband networks. A major benet of the tapped distribution architecture is the signicant reduction in ber that is required to serve a rural area. With the long distances typically involved in rural FTTx deployments, this reduction in ber count can dramatically reduce up-front network costs.

1100 CommScope Place, SE P .O. Box 1729 Hickory, North Carolina 28603 Tel 1.800.982.1708 or 828.324.2200 Fax 828.328.3400 www.commscope.com
2012 CommScope, Inc . All Rights Reserved B-WP-BOS-1 04.12

Anda mungkin juga menyukai