Long-Reach Optical Access Technologies

Abstract


A number of leading long-reach optical access technologies are examined. It gives the economic arguments behind the technologies.

1. Introduction


As broadband uptake increases globally, the services offered are becoming increasingly bandwidth-intensive. A service that continues to increase bandwidth requirements is high-definition television (HDTV), which requires in the region of 8 Mb/s per channel. For telecommunication networks to support a single channel of HDTV plus data and telephony services simultaneously, the minimum bandwidth required is approximately 10 Mb/s. A passive optical network (PON) offers a different network architecture that enables the essentially unlimited bandwidth of fiber to the home (FTTH) to be utilized. PONs use a point-tomultipoint architecture to reduce cost by sharing a significant portion of the network among all customers rather than each customer having a dedicated connection as in the current point-to-point architecture. FTTH was deferred in favour of incremental technologies based on DSL, which provided small increases in bandwidth without the requirement for large capital investment. These technologies reuse the copper infrastructure and use modern encoding and compression techniques, such as those developed by the Moving Pictures Experts Group (MPEG) to provide the required performance and quality of service.

2. Long Reach Technologies


A effective solution for an optical access network; the long-reach optical access network. The strength of optical technology is its ability to displace electronics and simplify the network by combining network tiers. The access and metro networks can be combined into one through the use of an extended backhaul fiber, possibly 100 km in length to incorporate protection paths and mechanisms, used with a PON. A general PON architecture consists of a shared fiber that originates from a local exchange. At a point close to the customer premises, a passive optical splitter is used to connect each customer to the main fiber. Even though optical amplifiers must be used to increase the power budget, the distribution section closest to the customer remains passive. Cost savings are introduced as the synchronous digital hierarchy (SDH) rings are replaced with a single backhaul fiber. The combined access and backhaul network terminates at a core node, possibly resulting in the removal of the local exchange site. The result is a highly simplified network. It is envisioned that only 100 nodes would be required for the entire UK, as opposed to the current 400 nodes. Initially, long-reach optical access networks were single channel, where a single wavelength is shared between all users, using time division multiplexing (TDM). These networks were followed by wavelength division multiplexing systems that shared a number of wavelengths between groups of users. More recently, GPON extension systems have been developed that enable a number of existing GPONs to be grouped and converted into long-reach systems with dense wavelength division multiplexing (DWDM) backhaul systems.

3. PLANET SuperPON


The ACTS (advanced communications and technologies and services) project, photonic local access NETwork (PLANET), was initiated in the mid 1990s. The project consisted of a an association of several companies of telecoms operators and suppliers working in parallel to develop a cost-effective, full service access network. The architecture” developed by investigating possible upgrades to the G.983 broadband PON (BPON) architecture ” includes a large splitting factor and longreach. FTTH was a cheaper option than fiber to the cabinet (FTTC). FTTC requires less fiber to be installed but involves extra cost for the housing of the optical network unit (ONU) at a street location rather than the customer premises. Providing FTTH through SuperPON was more cost-effective than BPON as customers share a large part of the network. Operation and maintenance costs are 70 percent lower for FTTH compared with FTTC. Increasing the BPON split size from 32 to 1024 and range from 20 km to 100 km introduces large amounts of attenuation. System performance was compromised as the received signal power was reduced. To overcome the performance loss, SuperPON introduced optical amplifiers into both the upstream and downstream channels. Downstream, transmission was simplified through the distribution of costs. Because the optical line terminal (OLT) was shared between all users, it was not a cost-sensitive component; therefore an expensive high power transmitter can be used. This was advantageous as a strong signal was launched into the backhaul fiber that was easily amplified by the optical amplifiers in the system ensuring a strong signal was present at the ONU receiver. Upstream transmission was not as straight forward. Optical amplifiers provide the gain required for transmission across SuperPON architecture to be feasible. However, optical amplifiers also have a detrimental effect on performance due to unwanted optical noise, known as amplified spontaneous emission (ASE), a side effect of the amplification mechanism. The system performance can be measured in terms of the signal-to-noise ratio (SNR) which is a direct comparison of the signal power compared with the noise power. The ASE produced by the erbium doped fiber amplifier (EDFA) remains constant; hence the SNR was reduced as the signal was attenuated by the increased loss due to Long-Reach Optical Access Technologies the increased split size which is done by creating amplified splitter. Instead of placing amplifiers after the splitters, optical amplifiers were included in parallel between split stages which reduces the split loss before the optical amplifier. When several optical amplifiers are placed in parallel, the effect of ASE becomes more severe and each optical amplifier produces ASE that then is combined with the ASE contributions from the other amplifiers at the splitter, which acts as a combiner in the upstream direction . This effect is known as noise funnelling and increases the amount of ASE present at the receiver.

4. Long reach PON


The long-reach PON was a 1024-way split, 100 km reach 10 Gb/s, optical access network solution. It was a simple solution that comes from modern optical technologies. The optical split size was 1024, which is half of that of SuperPON, but it only requires six optical amplifiers for both upstream and downstream operation as compared to 39 required by SuperPON. The longreach PON was completely passive in the access network section between the customer premises and the local exchange site. An intermediate amplification site was positioned at the local exchange site immediately after the 1024-way split in the distribution section. The local exchange site was an ideal location for the intermediate amplification stage as the site already contains electrical power, and hence no electrical power must be installed in the distribution section. Because the optical amplification was performed after the split, no optical amplifiers are in parallel. Hence, noise funneling was not an issue, and therefore no complex optical gating system was required as in SuperPON. Similar to the SuperPON, the distribution of cost means that transmission in the upstream is more difficult than the downstream. To counter the attenuation of the distribution section, a dual stage intermediate amplification stage was used. The first stage consists of a low noise pre-amplifier. As the signal power was so low, the SNR is determined by the noise produced by the optical amplifier, and hence the SNR can be maximized by minimizing the amplifier ASE. The second amplification stage provides the signal with enough power to overcome the attenuation of the backhaul section. Forward error correction (FEC) is a technique whereby transmission errors can be detected and corrected by encoding the data and including a number of parity bits. An optical filter limits the optical frequencies that the receiver can see. Therefore, an optical filter reduces the amount of ASE present at the receiver, which increases the SNR, and therefore the performance of the system. However, the downside to using an optical filter is that the wavelength of the transmitter must be strictly defined to ensure that it can pass through the pass bandwidth of the optical filter. Unfortunately, the wavelength of a laser is temperature dependant, and therefore a temperature controller must be included to ensure the wavelength does not drift out of the filter pass band. This is not desirable as it adds additional cost to the ONU. A cost-effective solution was found that involved using a 10G Ethernet transmitter. These transmitters include a thermo electric cooler (TEC) but are cost-effective due to volume production for 10 Gb/s Ethernet applications. The downside to this transmitter was that it was designed for short-reach application of less than 40 km. A number of techniques exist to counter dispersion penalties. With the use of appropriate optical technologies, it was possible to achieve 10 Gb/s transmission in the downstream and upstream channels across 100 km to 1024 customers using a low cost optical transceiver in the ONU situated in the customer premises.

5.Long Reach GPON


A major obstacle for optical access networks was the cost associated with installing an optical transmitter and receiver in the ONU at the customer premises . Standard PON reduces this burden by using lower cost uncooled transmitters in the ONU. However, this results in the transmission wavelength being temperature dependant with a possible drift of 20 nm. As no component in a standard PON is wavelength critical, the performance was unaffected. The detrimental consequences of wavelength drift only become crucial when using narrow optical filters for ASE reduction. In these systems, more expensive cooled transmitters are used to ensure a stable wavelength.

6. Wavelength converting PON


The wavelength converting (WC) PON uses optical wavelength conversion in the upstream to transfer the transmitted data from the ONU wavelength to a standard DWDM wavelength. In the wavelength conversion process, the network operator has complete control over the wavelength to which the original wavelength was converted. A number of PONs can be grouped together over the same backhaul fiber, with each PON being converted to a separate wavelength. This reduces the cost of the backhaul fiber on a per customer basis as the utilization and efficiency of the backhaul fiber is increased, enabling the cost to be spread over a large number of customers. Therefore, with the inclusion of a wavelength converter, a high split, long-reach optical access network that combines both access and metro networks can be created from standard PON architectures. In the distribution section, an infrastructure that follows the GPON standard is represented by 20 km of standard single-mode fiber and a 64-way split. The ONU transmitter consists of an uncooled transmitter with possible wavelength drift over a coarse wavelength division multiplexing channel (CWDM), that is, 18 nm. In this implementation, a simple cross-gain modulation (XGM) wavelength converting technique was used due to its simplicity. The XGM technique places certain limits on the conversion bandwidth due to physical effects within the semiconductor optical amplifier(SOA). Therefore, the ONU wavelength was centered on 1550 nm and was free to drift within ±9 nm of the center wavelength. A dual fiber backhaul was used to avoid any wavelength clashes between the upstream and downstream channels. The XGM wavelength converter consists of two stages: a SOA, where the XGM conversion was performed and a pre-amplifier stage. As the XGM wavelength conversion process relies on SOA gain saturation, an EDFA was used as a pre-amplifier stage to ensure there was sufficient optical signal power. Ideally, an SOA pre-amplifier would be used to aid integration. To improve the performance of the XGM wavelength conversion process, a CWDM filter (18 nm, l = 1550 nm) optical filter was used. The wavelength that the signal was converted to (probe wavelength) was supplied by a continuous wave laser, injected into the SOA in the counter propagation direction. The signal on the converted wavelength was coupled out of the wavelength converter through a 2×2 coupler and multiplexed into a 100 km standard single mode fiber backhaul through an AWG. A second AWG was positioned immediately after the backhaul fiber to filter each channel to the correct OLT. The AWG had 20 channels with 100 GHz bandwidth, enabling DWDM to be implemented. Therefore, it would be possible for the system to support 1280 users, over 120 km. Each user would have access to a minimum symmetrical bandwidth of 38.8 Mb/s.

7. PIEMAN


Research is continuing into long-reach optical access as demonstrated by the Information Society Technologies (IST) sixth framework project (FP6), photonic integrated extended metro and access network (PIEMAN). The network has a 100-km reach with a 32 wavelength DWDM backhaul. Each 10 Gb/s wavelength is uniquely allocated to a PON with a 512-way split, enabling the network to support (32 × 512) 16,384 users with an average bandwidth of ~20 Mb/s. By using dynamic bandwidth allocation and 10 Gb/s components in the ONU, it is possible for each user to burst at 10 Gb/s. Colorless ONU will be provided for upstream transmission to avoid the requirement for each customer to buy an ONU with specific wavelength. Two components were identified as having the potential to provide a colorless ONU. The first is a tunable ONU that can select from 32 wavelengths with 50-GHz spacing. To reduce the cost of the devices, a set and forget approach is taken for the tunable ONUs. Instead of tuning the ONU to a specific wavelength before transmitting, which requires the device to be wavelength agile, the wavelength of the device is set once when the ONU is added to the network, and no further changes to the ONU wavelength are made. A second approach under consideration is to use reflective SOAs. This approach is similar to the hybrid WDM/TDMA PON, where a wavelength is supplied to the ONU by the network. The burst mode nature of the TDMA traffic in the upstream path of PONs causes a number of problems. Standard EDFAs cannot be used due to the slow gain dynamics. That is, the gain of EDFA cannot change fast enough to ensure that the entire packet receives constant gain. The gain changes as the packet propagates through the EDFA causing the amplitude of the packet to be non-uniform. The selected solution uses an auxiliary wavelength that is adjusted relative to the transmitted upstream packet so that the optical power through the EDFA remains constant. Hence, the gain of the EDFA remains constant for the duration of the burst. Upstream burst-mode traffic also means that a standard continuous mode receiver cannot be used. In current PONs, a DC coupled receiver is used, where the receiver must determine the correct threshold on a burst-by-burst basis due to path attenuation differences for different ONUs. Problems occur when scaling these techniques to 10 Gb/s. Hence, the PIEMAN project is developing a new 10 Gb/s burst mode receiver that uses a multistage feed-forward architecture, reducing DC offsets through a long chain of electrical amplifiers.

Conclusion


A number of promising technologies have been developed that enable long-reach access to become a reality. SuperPON provided the foundation that proved the feasibility of providing FTTH along with network simplification through a combined access and metro network. Long-reach PON improved SuperPON by using modern components and techniques to provide 10 Gb/s symmetrical bandwidth, but the choice of singlewavelength TDM/TDMA operation results in low efficiency in terms of bandwidth allocation and fiber utilization. The PIEMAN project is advancing the WDM/TDM architecture and is researching technologies such as burst-mode receivers that will be vital when implementing optical access at 10 Gb/s. However, we believe that the most suitable systems for deployment of long-reach optical access will be the long-reach GPON and WC-PON. This is because these systems provide an intermediate step; enabling operators to deploy standard PONs now and then upgrade to long-reach in a simple manner.