SDH Radio: The Technology of Today and Tommorrow
A new generation of synchronous digital heirarchy (SDH) microwave radios that provide an economical solution, fast deployment and security
SDH Radio: The Technology of Today and Tomorrow
Northern Telecom (Nortel)
St. Laurent, Quebec, Canada
Today, telecommunications network planners have a variety of network elements at their disposal to design the solutions to meet their short- and long-term requirements. For high capacity long-haul transport of information, two key technologies are fiber optics and digital microwave radio. Continued progress in the field of fiber optics has translated into larger capacities, improvements in performance and reliability, and longer spans. Therefore, deployment of fiber-optic systems increasingly has become the preferred choice over traditional microwave systems, which have been referred to by some as the weak link in the chain. However, innovations in the area of digital microwave systems have led to significant advances as well, which will ensure the viability of this technology alongside fiber optics in present and future networks.
The new generation of digital microwave systems, based on synchronous digital hierarchy (SDH) standards, is able to meet the requirements of the networks of today and tomorrow. The growing need for SDH radios requires that equipment vendors address the best method to achieve true global intercompatibility between SDH radio and fiber systems in evolving networks.
The Need for SDH Microwave Radio
Although fiber optics increasingly has become the medium of choice for long-haul, high capacity transmission systems, SDH microwave radio is still required by many network planners. The main reasons for this demand are related to economics, speed of deployment and security.
SDH radio provides the most economical solution to network planners when existing infrastructure (for example, towers, shelters, power plants and antenna feeder systems) can be re-used, when rights of way or adverse terrain (such as mountains and bodies of water) make fiber deployment extremely costly, or when initial (lower) capacity requirements do not warrant the capital outlay of implementing a fiber system. In the case of a radio network implementation, it is important to note that capital is expended in proportion to the deployed network capacity, whereas for a fiber network implementation, the bulk of the initial capital outlay is for the installation of the fiber cables, which is independent of capacity. Therefore, radio investment is better matched to revenue generation compared to fiber systems since, with radio, there are lower startup costs for initial lower capacities and incremental cost when additional capacity is required.
To show the cost differences between radio and fiber network builds, the results of a joint economic study conducted by Northern Telecom (Nortel) and Stentor (the alliance of Canadian telephone operating companies) are presented. The objective of the study was to determine the most cost-effective transmission network design for a 2000 km route based on capacity requirements and terrain conditions. The results of the study, shown in Figure 1 , indicate that for difficult terrain conditions, radio is more cost effective for capacities up to 2.5 Gbps (equivalent to 16 × STM-1), where one synchronous transport module (STM) = 155 Mbps). For easy terrain conditions, fiber becomes more cost effective for capacities greater than 310 Mbps (equivalent to 2 × STM-1). Radio costs assume existing infrastructure while fiber costs assume new deployment, including laying the fiber cable. Based on this study, Stentor was able to optimize the design of its 6500 km TransCanadian high capacity synchronous network economically by deploying both fiber and radio.
Fig. 1: Radio vs. fiber construction costs for a 2000 km route.
Note that the cost comparison study was performed based on single radio and fiber routes. However, in order to meet the International Telecommunication Union (ITU)-T long-haul availability objectives, a dual fiber route (that is, route diversity) is required normally to compensate for cable cuts, which on average occur two to three times per year per 1000 km route. Each cable cut could require up to 12 hours to repair.1 In comparison, long-haul radio routes can be designed for 99.98 percent availability over 6500 km, which translates into less than two hours of downtime per year using a single route. Therefore, if this study was carried out comparing a single radio route vs. a dual fiber route, which delivers the same network availability, radio would be cost advantageous in even more capacity/terrain states.
Speed of Deployment
SDH radio offers faster deployment and quicker revenue generation than fiber, especially when existing infrastructure can be re-used. Figure 2 shows the number of years required for deployment, from planning until the system is ready for service.
Fig. 2: Deployment velocity for a 1000 km route.
Consequently, a practical deployment strategy for a network planner could be to deploy an SDH radio network initially. Then, as a revenue stream is established and the capacity requirements increase, a fiber route would be deployed in order to achieve a route- and media-diverse network, as well as to position the fiber for potentially higher capacities (such as 10 Gbps), if applicable.
Radio networks, which consist of sheltered radio sites spaced every 40 to 60 km, are easier to secure than fiber networks. Fiber networks are more difficult to secure because the entire fiber route must be protected to ensure the privacy and availability of the network.
Integration with Fiber Network Elements
To maximize the benefits of SDH radio, the radio must be able to complement a synchronous fiber network. However, in order for SDH radio to interoperate and integrate with fiber-optic network elements, its design must address several parameters, including capacity and growth, network management, maintaining pace with SDH standards evolution, interface and performance.
Capacity and Growth
Today's SDH radio technology is capable of delivering bandwidth efficiencies of over 8 bps for every hertz of bandwidth (8 bps/Hz). In particular, 512-state quadrature amplitude modulation (QAM) technology packs two STM-1s in a single 40 MHz channel using a single carrier. By adding channels in a 1:N frequency diversity configuration, system capacities up to 14 protected STM-1s can be achieved within one frequency band of operation (for example, in the upper 6 GHz band, eight bidirectional channels are available). By deploying a dual-band configuration (such as the lower 4 and 5 GHz bands), system capacities of STM-16 and greater are achievable. In addition, 512-QAM technology permits coexistence with older microwave systems as well as overbuilding of existing routes and infrastructure because it allows for operation within the standard ITU-R 40 MHz interleaved (alternate polarization) channel plan, as shown in Figure 3 .
Fig. 3: An interleaved 512-QAM 40 MHz channel plan.
This interleaved channel technology is in contrast with co-channel cross-polarization technologies, which also provide a means to double a channel's capacity using a lower order modulation scheme but with the need for tighter intrasystem interference planning considerations. With co-channel cross-polarization systems, frequencies and polarizations are re-used every two hops vs. every four hops for interleaved adjacent-channel systems, as shown in Figure 4 .
Fig. 4: Frequency and polarization re-use;
(a) co-channel and (b) interleaved 512-QAM channel.
It must be emphasized that 512-QAM technology is now considered a mature technology. The first technical trials of Nortel's North American version of its 512-QAM synchronous radio system, SONET Radio 4/40, took place in the summer of 1990, followed by field trials in 1991 and in the first half of 1992. Since then, over 1100 transmit/receive (TR) pairs have been manufactured and shipped, including the international SDH version, to Canada, the Far East and the Caribbean. Currently, nearly 900 TR pairs are in service. The expertise gained in the design and manufacturing process over the six-year period has enabled Nortel to produce modems exhibiting better than 10-20 residual bit error ratio (RBER) consistently.
Approximately 800 in-service TRs are in use by nearly 60 hops of radio covering 2600 km of the 6500 km hybrid synchronous route across Canada. (The other 3900 km are based on the OC-48 2.5 Gbps optical fiber system.) The radio system was implemented initially as a 1:3 capacity system (equivalent to six protected STM-1s), and has grown to 1:7 and 1:8 (to match STM-16 capacity) in most regions.
Network Management and Keeping Pace
with SDH Standards Evolution
It is important that the SDH radio software components be implemented with certain key functionality. Interfacing in accordance with the open systems interconnection seven-layer stack to facilitate evolution with Telecommunications Management Network standards is essential. As shown in Figure 5 , the integrated management of radio and fiber elements from the same platform is important as well, thus allowing truly seamless radio-fiber networks. Also, simple, nonservice-affecting, remote control software upgradeability becomes vital with increased software dependency. It should be possible to upgrade all radio network element processors remotely from a single point, without affecting service, through the use of the SDH embedded communication channel (ECC). This use of the SDH ECC allows fast and easy upgrades to SDH radio networks in compliance with standards evolution.
Fig. 5: A hybrid radio/fiber SDH network.
Software-based network management as well as operation, administration, maintenance and provisioning have made software a key component of SDH radio. Common software capabilities include alarm, performance, configuration, payload and inventory management, as well as built-in diagnostics and testing.
SDH radio must provide an interface that offers the most flexibility at the lowest cost, particularly for interconnection with the fiber-optic network elements. This interface is achieved by providing the radio as a transport pipe with a direct optical (STM-1 or STM-4) interface, which eliminates the need for multiplex equipment when only a radio-fiber transition is needed, without any drop or insert of traffic. Also, an optical interface reduces the cost and complexity of a fiber extension from the SDH radio, which can be located on a remote hilltop, to a fiber multiplex/transport terminal located several kilometers away in the city. Finally, an STM-4 optical interface has the added benefit of allowing for SDH radio operation within a 622 Mbps bidirectional line-switched (shared protection) ring configuration.
Leading-edge technologies are also used to ensure that SDH radios meet the stringent error performance objectives defined in ITU-T Rec. G.826 (and subsequently in ITU-R Rec. F.1092). Sophisticated and powerful countermeasures are available to combat propagation anomalies including high gain triple forward error correction (FEC) at each hop to provide the most effective means of coping with bursty errors and enhance switching protection, thus improving RBER. Figure 6 shows the triple FEC results, which can correct up to three errors per frame with three percent FEC parity bits and 5 dB coding gain at 10-6 . Multitap surface acoustic wave-based and application-specific integrated circuit-controlled adaptive time domain equalizers are deployed as a standard part of the demodulator circuitry to maximize the correction of intersymbol interference due to multipath fading and to achieve dispersive fade margin specifications equal to and, in many cases, better than systems using lower modulation schemes. In addition, two- and three-input space-diversity receivers, which improve effective fade margin and enhance severely errored seconds performance (seconds with BER greater than or equal to 10-3 ), are deployed optionally to overbuild existing routes where some hops may be very long or have exhibited difficult propagation conditions historically, as shown in Figure 7 . FEC-based errorless protection switching is integrated into radio systems to provide fast switch initiation and release times and to maximize frequency diversity improvements obtained from the system. Fully solid-state automatic transmit power control (ATPC)-based amplifiers are deployed, which vary the transmitter power in accordance with path loss changes. As a result, the power amplifiers operate optimally in linear regions for the majority of time when path loss is normal, and at maximum power only under fading conditions. Adaptive predistortion circuitry is used to maximize the transmitter performance dynamically at high output levels. ATPC operation translates into optimized system performance, simplified frequency coordination, reduced power consumption and improved overall equipment reliability, as shown in Figure 8 . State-of-the-art low noise amplifiers are incorporated into receivers to achieve noise figures of less than 1.7 dB, including the branching filter and circulators.
Fig. 6: The triple FEC.
Fig. 7: Space diversity receiver performance.
Fig. 8: ATPC operation.
Using these advanced techniques and the progress in technology, today's SDH microwave radio systems are more robust and reliable than older generation systems. More importantly, today's SDH microwave radio systems are able to meet ITU-T G.826 requirements for SDH transport systems.
The new generation of SDH microwave radios provide a good alternative for network planners. Because of SDH radio's economics, speed of deployment and security, it is the ideal vehicle for lower capacity spur routes, traversing difficult terrain or providing route/media diversity. By addressing the issue of capacity and growth, network management, standards evolution, interface and performance, new-generation SDH radios are interchangeable with fiber in high capacity synchronous transmission networks without concerns of impacting the strength of the chain.
Any republication of this article must acknowledge Nortel.
- Bellcore (T1X1.5 documents) network design parameters.