Sonet capabilities will dominate planned networks
Donald L. Dittberner
Dittberner Associates Inc.
Synchronous optical network/synchronous digital hierarchy-based transmission systems on optical fiber serve as the principal long-haul and inter-office structures for virtually all new and upgraded telecommunications networks. Inexpensive capabilities such as add/drop, increased supplier competition and self-healing make this possible. Such network installations are being implemented despite the newness of the associated standards and the lack of operational support systems that provide effective system management.
Rapid acceptance of the Sonet/SDH standards has led to diverse products from multiple suppliers with proven mid-span meet capabilities. This equipment availability will soon allow the mixing of products in transmission networks. In support of these trends, many major telecommunications companies have indicated their intention to trade out within the next decade their existing trunk plant for Sonet/SDH-based optical fiber technology.
From an operational standpoint, this conversion will be gradual. It will be led by new deployments of fiber-optic facilities because the sunk-costs of existing facilities for routes still capable of handling the offered traffic can seldom be displaced cost effectively. Sometimes, though, such displacement can be cost-justified where the provisioning costs associated with the churn (user changeover) of private dedicated circuits on the trunk plant are high or where automatic route diversity is a pressing user demand for special service or private dedicated circuits.
The use of Sonet/SDH networks to provide route diversity automatically has been a major driver toward the use of these technologies. In contrast, the test equipment and systems needed to check these networks are costly and complex. These thorny problems remain unresolved. One rectifying approach might involve the purchase of automatic test equipment shared on a regional basis; such equipment would be too expensive for use by individual craft teams.
Price trends
The diverse functions of Sonet/ SDH networks offered by major suppliers of transmission systems makes a direct comparison of prices difficult. However, cost-based price forecasts through the year 2002 for major types of Sonet/SDH equipment indicate a steady decline during this period. In fact, the forecasts show a modest 32% decrease in price during the next eight years.
Detailed analyses of a star configuration feeder plant (typical network configuration) of plesiochronous digital hierarchy technology versus a Sonet/SDH solution disclose a 19.5% cost penalty in 1994 for the ring solution (with full redundancy). This cost penalty is expected to drop to 14.5% in the year 2002.
Nevertheless, many telecommunications operators appear willing to absorb this cost penalty for the improved capability and ease of provisioning offered by the Sonet/SDH ring-network solution. In defense, they cite the multiple attributes of Sonet/SDH networks:
Provide a standard means of supplying transmissions of hundreds of megabits.
Permit rapid nationwide provisioning of circuits or digital virtual private network services for realistic bandwidth.
Provisioning and reconfiguration of circuits can be accomplished remotely and more rapidly than with other available technologies.
Allow add/drop multiplexing at reasonable costs because high-capacity optical fiber can cost-effectively share signal paths.
Implement self-healing functionality in network sections or whole networks at a modest cost penalty--in the range of 20% to 40%.
Network implementation
Sonet/SDH network implementations are satisfactorily progressing, but not at an accelerated pace in most industrialized countries. One impediment is that many telecommunications operators have installed PDH-based optical fiber transmission systems that are not yet fully amortized but that still provide adequate capacity.
Another impediment appears to result from a lack of availability of standard operational support systems to give full remote network management capabilities. Yet another obstacle involves the uncertainty that is perceived by many telecommunications operators- whether Sonet/SDH will operate adequately on fiber-optic cable systems that were installed extensively during the 1980s.
Viable applications
Likely applications for Sonet/SDH networks cover the gamut of existing installations. In North America, point-to-point systems running above the DS-3 (45-Mbit/sec) level are potential applications for upgrading to Sonet/SDH networks because of standard structures and interfaces.
In the regional trunk plant, Sonet/SDH rings and interconnected rings are another anticipated area of beneficial usage. This plant structure supports better management of network capacity and brings superior automatic (self-healing) or semi-automatic (craft-assisted) network restoration.
Installed in the access network (loop plant), Sonet/SDH rings would allow telecommunications operators to cost-effectively increase service availability. If this is not a consideration, then linear topologies that have multiple add/drop multiplexers will compete with multiple DS-3-based feeder route systems by adding flexibility in case of unpredicted traffic growth.
In long-distance networks, configurations with point-to-point links and multiple hubs would allow the creation of intermeshed networks that can support route diversity and semi-automatic or fully automatic restoration in cases of link failure.
If the proper type of optical fiber is initially installed, Sonet/SDH networks would allow capacity upgrading to OC-48 rates (2.5 Gbits/sec) and, in the future, to OC-192 rates (10 Gbits/sec) by changing either the transmission terminal equipment or, in many cases, only selected subsystems.
Another promising technology, wavelength-division multiplexing, makes it possible to upgrade the capacity of a Sonet/SDH system that is operating at its highest level. Two to four different wavelengths are currently used. These numbers will soon increase.
Some pitfalls
The path to Sonet/SDH network acceptance is not smooth; some pitfalls could derail a seemingly easy transition.
For one, comprehensive centralized network management is feasible only in single vendor environments. Full multi-vendor approaches will have to wait--possibly until 1995 or 1996, when supplier equipment interoperability issues are projected to be resolved. Thus, telecommunications operators should take this current limitation into account.
For another, automatic restoration appears to be working well within individual network rings. But this restoration within intermeshed hub/link network sections is still dependent on a central controller installation and has a number of operational limitations.
Automatic restoration in network sections comprising interconnected rings, as well as hub/ring topologies, is still difficult to implement. It requires cooperation between the automatic restoration process built into the ring`s nodes and that of the hub/link sections.
Prices of Sonet/SDH equipment are still higher than those of equivalent PDH equipment. Studies show that telecommunications operators can negotiate lower prices if they commit to volume purchases for retrofitting their networks.
In the public switched network arena, Sonet/SDH implementations are projected to increase the usage of remote subscriber units associated with local digital switching central office equipment. They provide flexibility for growth and capacity allocation among remote subscriber units connected to the Sonet/SDH ring, and they permit telecommunications operators to provide guaranteed service.
However, digital switching suppliers must provide Sonet/SDH interfaces on the subscriber loop side of their switching systems. They must also integrate the operation and maintenance capabilities of Sonet/SDH with those of the operational support system interfaces of their local exchanges to support ring structures in the local plant.
Little impact is seen from Sonet/SDH implementations on the trunk side of local central office switches. Passing the Sonet/SDH overhead through the local digital exchange appears to demonstrate little benefit.
The impact of Sonet/SDH within private enterprise networks should accelerate the trend toward the use of asynchronous transfer mode switches by employing direct interfaces.
The offering of digital crossconnect capabilities by public network carriers to support reconfiguration of private network Sonet/SDH facilities appears attractive to users and telecommunications operators, particularly to avoid dark fiber offerings with little value added.
Donald L. Dittberner is president of Dittberner Associates Inc. in Bethesda, MD.