Cloud applications, network convergence, and new ways of working such as desk-sharing and home-working have significantly boosted bandwidth demand worldwide. Nexans says that the fast-growing number of Internet of Things (IoT) and Building Internet of Things (BIoT)-based devices in smart buildings, for example, have resulted in a vast increase in the number of required ports.

In the current post-pandemic world, cabling needs to be more flexible and easily configurable than ever. Passive Optical LAN (POL) and Fibre to the Office (FTTO) solutions can both offer this – but there are marked differences.

To register for the webinar click above

Traditional enterprise network infrastructure, standardized in ISO/IEC 11801, is mainly separated in horizontal and vertical cabling. Physical infrastructure is realized in a tree topology. Horizontal cabling is based on copper wire (primarily CAT6, CAT6A or CAT7A) and installed from a floor distributor to the office environment, where cabling is terminated in RJ45 outlets in a cable duct or floor box. Terminal devices such as PCs, IP Phones, or Wireless LAN Access Points connect to this. In the floor distributor, copper cabling is usually terminated in 19” racks and connected to standard 19” access switches to provide Ethernet services to connected devices. Horizontal cabling has a maximum channel length of 100m (90m fixed cabling + two 5m patch cords). This can result in multiple distributors per floor. Although available bandwidth is up to 10Gbps per port, most connections by far are based on 1Gbps.

Figure 1: Traditional network infrastructure

The floor distributor contains a great deal of active and passive equipment and therefore requires a power concept, cooling and a minimum of security to protect the network infrastructure. Vertical cabling is based on fibre. It connects all access switches to central core switches and forwards the aggregated bandwidth to a central distribution room. These connections require more bandwidth and technologies such as 10, 25 or 40Gbps.

Unlike traditional network infrastructure, a ‘full fibre’ approach exclusively uses fibre optic cabling. Fibre is installed between a central building or campus distributor and the office environment. Depending on the fibre type, distances range from a few hundred metres to several kilometres. Floor distribution rooms are not required. Using fibre cabling offers the option to install a physical tree topology, as well as a ring topology, by using extractable cabling (Fig 3).

Figure 3: Ring topology based on fibre optic cabling

Main drivers for installing a fibre-optic cabling-based infrastructure

  1. There is no distance limitation. This avoids floor distributors and all related electrical installation work, increases security, and saves space, CAPEX and OPEX.
  2. Single-mode fibre offers practically unlimited bandwidth reserves. Today, it is already possible to transmit hundreds of Gb/s over single-mode, resulting in future-ready network infrastructure.
  3. The lower volume of fibre optic cabling avoids big (copper) cable bundles in the floors. This can be essential in historical building with limited options for rolling out new network infrastructure.
  4. Reduced energy consumption over the whole lifetime and helps save our planet’s resources.

Technical basics of two main technologies for realizing a full fibre enterprise infrastructure

A Passive Optical Local Area Network (POL or POLAN) migrates provider infrastructure and Passive Optical Network (PON) technology to enterprise, campus or hospitality network infrastructures. It is based on a point-to-multipoint topology and connects one central port with multiple clients. The used technology is mainly GPON or GEPON derived from Internet Service Provider networks.

Table 1: Wording used in PON networks.

In a typical POL installation (Fig 4) the central equipment (OLT) is placed in the building or Campus Distributor and provides the signalling that is forwarded via one single-mode fibre to a passive optical splitter that splits/divides the optical signal for the connected ONTs/ONUs. The passive optical splitter is used instead of floor distributors to connect multiple end user. The optical network terminates at the end user side in an ONU/ONT in the office next to the client. All terminal devices connect to this using RJ45 connectors.

Figure 4: Typical POL installation
Table 2: Technical details of technology used in a POL network.

Fibre To The Office (FTTO) is a second concept based on full fibre infrastructure. This point-to-point topology connects one central port to a single FTTO switch installed in the office environment (Fig. 5). The used technology is standard Ethernet Point-to-Point.

Figure 5: FTTO topology.

Small managed Ethernet switches convert the optical signal to electrical and vice-versa. FTTO switches are placed in cable ducts or floor boxes and usually offer four RJ45 ports to connect terminal devices in the office environment. On the central side, standard Ethernet switches are used to connect the FTTO switches to the core network. FTTO uses standard enterprise network equipment and technology, most commonly Gigabit Ethernet, providing one or two optical uplinks supporting 1Gbps and four user ports that also support 1Gbps.

Table 3: Comparing POL and FTTO: key topology differences
Figure 6: Topology overview

A shared medium is used for communication in a point-to-multipoint POL network, raising several issues:

  • From a security perspective, sharing multiple connections on a single fibre is less than ideal.
  • Available bandwidth has to be split into an existing ratio - either 1:16, 1:32 or 1:64.
  • Limited flexibility during rollouts may result in overspecifying. You might require just one port in a specific location but have to install 16, for example. It is impossible to upgrade one single client. Upgrades affect every connected client.
  • POL is based on single-mode fibre optics, which requires bidirectional optical transmission technology to filter the used wavelength. Using multi-mode optical fibre is not possible.
  • The shared fibre and splitter act as a single point of failure for all connected ONTs and clients. Risk of failure is normally mitigated by introducing redundancy topologies, such as ring topologies. In POL networks a redundancy can only be archived on the central side when using 2:n splitters with two central OLTs which results in a vast cost increase.
Table 4: Comparing POL and FTTO: key equipment differences in design, installation and feature setup.
Figure 7: 7-Port FTTO switch

A fibre-based infrastructure that offers the best of both worlds

What type and level of performance do users and devices require – now and in the future? Which conditions exist in the building or buildings? Which distances need to be bridged? Are there specific requirements with regard to power, functionality or uptime? How flexible does your network need to be? The choice between FTTO and POL depends on numerous variables. However, it’s also possible to simply combine the benefits of POL and FTTO. To do this, fibre infrastructure must be planned with single-mode fibres. as POL is based on a bidirectional communication via one fibre. Because most PON ONTs are still connected via an APC connection all connectors should be designed as LC/APC or SC/APC connectors. Infrastructure topology needs to be planned as a logical point-to-point infrastructure between central building or campus distributor and end user (see topology example Fig. 8). Multifibre cabling containing 12 or 24 fibres, for example, acts as vertical cabling and travels from a central optical distribution frame (ODF) to the floors where ‘Zone Distribution Boxes’ (ZD Boxes) are located. The single fibres are terminated and spliced in the ODF and ZD Boxes. From the ZD Boxes, pre-terminated fibre cabling or fibre patch cords connect ONTs or FTTO switches

Figure 8: Fibre infrastructure for POL and FTTO

All active equipment - the OLT and Ethernet Switch - is installed in the central distribution room. The passive optical splitter required for a POL installation also needs to be installed in the central building distributor between OLT and the ODF to archive a maximum of flexibility. This allows the connection of ONTs installed on different floors. Another option could be to install a compact splitter in the ZD Box. That could make sense in big installations where all connected ONTs are on the same floor.

FTTO switches are directly connected to the ZD Boxes and ONTs usually via a fibre outlet or dedicated mounting boxes. ONTs are already equipped with an optical interface (usually SC/APC) whereas FTTO switches need a SFP transceiver with the corresponding optical interface. In this case, it needs to be a bidirectional SFP transceiver providing LC/PC or SC/PC connection. Finally, a hybrid optical patch cord from SC/APC to LC/PC or SC/PC connection is used to connect the FTTO switches to the APC connector.

To meet today’s usage demands, previously disparate platforms and components should be able to communicate in a unified manner and act as an integrated whole. That requires a common language and an integrated approach to structured cabling and devices. In this respect, POL and FTTO each have their pros and cons. However, both are based on a full fibre based infrastructure with all its advantages. So why not plan a fibre based infrastructure that can provide both, keeping the flexibility to move between technologies or running both in parallel?

You can register now for our upcoming webinar on POL versus FTTO on Thursday 1st December presented by Jan Middeldorf.