Dr.-Ing. Ziaedin Chahabadi is the chairman and CEO of the Executive Board of KEYMILE International GmbH. Previously, Dr. Chahabadi served as a member of the Executive board of the QUANTE group in Wuppertal. Dr. Chahabadi was also Head of the Technology section, Deputy Managing Director, and one of the founding members of Ke Kommunikations-Elektronik GmbH. Dr. Chahabadi began his career at Kabelmetal electro GmbH in Hanover, a subsidiary of the Alcatel group, as head of the department of transmission technology. Dr.-Ing. Ziaedin Chahabadi, earned his doctorate at the University of Hanover in high-frequency engineering.
Fibre to ‘x’ (FTTx) is the key to meeting the vast demand for higher bandwidth. Transmission of TV signals requires particularly high levels of bandwidth. HDTV and 3D-TV for example need up to 50 Mbps per TV channel, but copper wire DSL transmission has reached its limits. FTTH (fibre-to-the-home), or fibre optic transmission to the customer, has enough bandwidth to cope for the next 20 years. The debate on which infrastructure is best for network operators is still ongoing.
The telecommunications industry has more than ten years experience with optical networks and arguments about their pros and cons have been going on for just as long. Fibre optic networks can be laid right into the home (FTTH) by using both passive optical infrastructures (P2MPP) and point-to-point optical infrastructure (PtP). The key difference between active and passive access technology lies in how the fibre optics are used. Ethernet point-to-point active networks use dedicated fibre optics from the central in-feed point to the customer. In the case of passive optical networks (PON) unpowered optical splitters are used to route signals to customers. Both passive and active optical networks bring the fibre optics as close as possible, ideally into, the subscribers’ homes. In terms of transmission quality and bandwidth, FTTH is technically the best option. Passive optical networks (PONs) The first element in a PON network is the OLT (optical line termination), which provides standard Ethernet interfaces with the core network and PON interfaces with the subscriber. The PONs used today are usually Ethernet-PONs (EPON) or Gigabit-PONs (GPON) and in the future Gigabit-Ethernet-PONs (GEPON) or WDM-PON. Currently, EPON installations are widespread in the Far East, GPON on the other hand is common in the US and Europe. With PONs, the signal on the fibre optic to the subscriber is split by a passive splitter to several optical subscriber lines. The splitter is either located in an outdoor housing, or directly in the cable, for example in a sleeve. The network has a point-to-multi-point-like (P2MP) structure. In FTTH network architecture, an optical network termination (ONT) unit implements the subscriber’s access by changing the optical signal into one, or several, electrical interfaces like Ethernet, POTS and ISDN. ONTs with VDSL interfaces are available for FTTB applications to bridge to the existing subscriber lines in the home. In this case, each subscriber receives a VDSL (very-high-bitrate digital subscriber line) modem to terminate the network. Ethernet point-to-point (PtP) In Ethernet point-to-point network structures, all subscribers receive their own fibre optic line that terminates at an AN (access node) optical concentrator. Therefore, in contrast to PON systems, there is a direct connection from the subscriber’s CPE (customer premises equipment) to the central optical port. The interfaces comply with the Ethernet standard, already used in core networks for years. Usually, each subscriber receives bandwidths of 100 Mbps or 1,000 Mbps. Depending on the technology used, one or two fibre optic lines can be used for data transmission; in today’s FTTH applications, single fibre systems are usually deployed. Due to the dedicated fibre optic line for each subscriber, changes in bandwidth can be customised to suit the functions and services required. In the case of FTTB (fibre to the building) applications, a micro DSLAM (digital subscriber line access multiplexer) distributes the data within the building via the existing telephone cable to the customer’s home using VDSL. Passive optical vs. point-to-point Almost any network topology is feasible using PON and Ethernet PtP. However, a network operator should decide early on which architecture is likely to be able to cope with demands in 15 to 20 years, because an investment in infrastructure with an ROI of around ten years should last and not require changes after five years. A recent study by WIK-Consult, on “Architectures and competitive models in fibre networks”, emphasises the long-term impact of investments1. The authors come to the conclusion that point-to-point fibre optic access networks initially cost about ten percent more than point-to-multipoint networks, but the higher investment costs pay for themselves due to the greater and more long-term usage for private households and business customers. This is of course a crucial factor. In a point-to-multi-point structure, which is required for PON systems, network operators save real cash on the initial investment. The reason for the saving is that they do not have to lay as many fibre optic lines as if they had used a point-to-point structure from the very beginning. PON’s weak point is the optical splitter. Where a high level of bandwidth is required for the customer, this network component might have to be replaced. In the worst-case scenario, additional fibre optic lines might also be necessary to upgrade to a point-to-point structure. In terms of bandwidth per subscriber, PtP is better. The maximum bandwidth per subscriber is much higher. Also, the opportunity PtP gives to allocate heavy users their own, separate, bandwidths provides more flexibility than PON systems are can. Depending on the splitting factor, a PON connection via fibre optic lines provides less bandwidth than a VDSL2 connection via copper wire. When it is a question of increasing the bandwidth, PtP architecture is superior to PON’s point-to-multipoint architecture. Subscribers can easily obtain upgrades without having to change the network architecture or the service of other subscribers. Within a PON tree, since all subscribers share the same optical fibre, if a single termination causes faulty synchronization or there is an undefined optical signal, remote location of the defective ONT may not be possible. With PtP architecture, both the fibre optic path and the end customer’s ONT is clearly identifiable. In the worst-case scenario, the laser to the optical concentrator can be deactivated individually from control centre. As regards availability, PON also does worse than PtP, because redundant connections to customers’ sites from the same PON have not been implemented or even planned. Until now, it was a common belief that GPON systems use much less power than Ethernet point-to-point systems because they divide up the laser’s light on the OLT to several fibre optic lines, while in point-to-point systems one single laser each is used. However, the GPONs optical transmission signal has to be higher because of the optical splitter attenuation. In fact, the power consumption of a control centre with an active optical 100 Mbps system and a GPON system with a 16-fold splitter are about the same. If we compare the systems’ terminal equipment, we will notice significantly higher power consumption in PON terminal equipment (approx. eight W in GPON compared with four W in PtP) because it need more power to transmit upstream in order to overcome the splitter’s attenuation. Overall - considering the control centre and customer equipment - Ethernet point-to-point systems use less power. For years, the role of PON systems was splitting up cable-TV signals coming in simultaneously on the same fibre optic line. As a result, network operators were able to offer traditional cable-TV, Internet, voice and IPTV using the same infrastructure. Now, more modern Ethernet point-to-point systems also offer this functionality, so that external equipment for feeding in cable-TV, or even another fibre optic line is no longer needed. ___________________________________ It should be noted that generic comparisons of technology only serve as initial rough guides. While network operators in Asia prefer passive optical networks, a study by the FTTC Council Europe found that in Europe 73 per cent of FTTH/FTTB installations are based on Ethernet PtP (IDATE Study, FTTB Council Europe 12/2010). The result of this study is not surprising because about 80 per cent of the investment in new fibre optic infrastructure does not depend on the use, or not, of active technology but upon the cost of laying fibre optic lines on a widespread basis. The most important point is that the new infrastructure can keep up with growing customer demand over the next 20 to 30 years. The low incremental cost of a point-to-point infrastructure will enable sustainable use for a long period. Ethernet point-to-point technology is ideal for network operators who want to lay their own fibre optic infrastructure or used debundled fibre optic paths to attract business customers, multi-residence buildings, local authorities and universities. In these cases, PtP easily provides the required flexibility, quality and security that PON networks, due to their structure, struggle to provide.