Is Sonet ready to transport video?
Transporting video services, in addition to voice and data services, on Sonet fiber-optic networks calls for careful planning of timing and synchronization performance, imposes new methods for evaluating video transport performance, and poses new transmission challenges for telecommunications network operators
Mark j. lum
tektronix inc.
In addition to delivering voice and data services, telecommunications networks based on Synchronous Optical Network (Sonet) technology are expected to support future broadband Asynchronous Transfer Mode (atm) applications and broadcast television services. Many telecommunications companies, television broadcasters, and video service providers are building Sonet networks or contracting with local service providers for digital-video transport. These networks can support contribution-quality applications that link studios and post-processing houses, as well as primary distribution applications that link broadcast sites such as terrestrial transmitters, cable-TV headends, and satellite up-link stations. (Contribution-quality video refers to the high quality of video that is suitable for rebroadcast as opposed to "distribution quality," for which a lower quality is acceptable because it will not be rebroadcast.)
However, video services demand a more stringent level of timing and synchronization performance than do voice and data services. Without careful planning, the use of Sonet networks to transport video can visibly degrade the quality of service. This factor places strict requirements on the performance, specification, and evaluation of Sonet networks, Sonet terminal equipment, and DS-3 video codec equipment operating at 44.736 Mbits/sec.
The challenge for Sonet equipment is that various video performance parameters have not yet been defined or standardized. In addition, Sonet networks require proper qualification if operators are to reliably deploy DS-3 video services. Despite the omission of video standards, video test tools are available to evaluate and resolve these timing and synchronization issues.
For some network operators, transporting video is not new. They have worked closely with broadcasters to transport analog television signals over dedicated links. With the advent of digital-video services, however, multiplexing video with other information--voice and data, for example--over the same Sonet network can be beneficial.
Sonet technology offers an ideal mechanism for network operators because they can use the same backbone network for all three services, resulting in integrated network management and easier maintenance. Video-transport services can therefore be provided without the need for dedicated facilities.
Regardless of the digital data-format technology, the use of a Sonet transport network for video encoded at DS-3 rates can introduce unwanted timing impairments. These can result in signal distortion and visible picture degradation. It is, of course, essential to evaluate and control those impairments before service degradation occurs.
The cause of these timing impairments is the Sonet "pointer-justification" process, which is used to maintain the synchronous operation of the Sonet network while accommodating inevitable signal phase variations within the network.
Phase variations among Sonet signals are effectively quantified in 24-bit increments and encoded into the transported signal as pointers. These large phase steps ultimately have to be handled when services are delivered, resulting in phase disturbances at low frequencies (predominantly between 1 and 100 Hz).
Unfortunately, since DS-3 video was not considered to be a primary service driver when Sonet standards were initially developed, significant omissions exist in the video equipment and interface specifications. Consequently, some areas of timing performance are essentially undefined and, therefore, not easily tested. These deficits lead directly to operational problems that are difficult to identify and resolve, yet have a marked impact on the quality of service.
Performance specifications
To understand the new performance requirements that digital-video services place on Sonet networks, we must evaluate the current timing and synchronization specifications through a network connection.
Telecommunications equipment and network timing and synchronization specifications are developed by the International Telecommunication Union (itu), American National Standards Institute (ansi), and Bell Communications Research (Bellcore). For transport networks, their goals are to develop service-independent specifications that can be met under all normal operating conditions.
In a Sonet network model, the transport layer can be formed from the number of network elements between the terminating elements (see Fig. 1). The synchronization layer can be implemented using the same physical Sonet transport layer, or it may use a different network.
Jitter and wander timing generation, accumulation, and consequent effects are controlled by specifying parameters for both the equipment and the network. The term "jitter" is defined as phase variation above a frequency of 10 Hz, and "wander" is phase variation below 10 Hz. Jitter is specified using the unit interval parameter, while wander is specified using the parameters of maximum time interval error and time deviation (see Lightwave, October 1995, page 48).
Historically, DS-3 jitter has been specified only at frequencies above 10 Hz. In pre-Sonet networks, DS-1 (1.544 Mbits/sec) and DS-3 circuits were operated plesiochronously. They did not transport timing, so wander was not an issue and was not specified.
Unfortunately, National Television Standards Committee (ntsc) video signals have a sensitivity to phase modulation below historical jitter frequency limits, and the use of Sonet as a transport layer introduces new phase disturbances due to pointer justification at these lower frequencies.
The Sonet desynchronizer is responsible for re-creating DS-3 services at the terminating end of the Sonet network. As a critical part of the network transport chain, it must accommodate the various impairments that have been added to the Sonet/DS-3 signals on their journey through the network to deliver DS-3 services within network limits. Again, the principal timing impairment is pointer justification.
The Sonet standard specifies DS-3 combined jitter due to mapping and pointer-test sequences. This jitter is tested by applying a series of defined Sonet pointer-test sequences and then measuring the consequent jitter at the DS-3 tributary output.
Note that the standard does not specify output wander at the DS-3 interfaces, and output jitter is specified only above 10 Hz. Therefore, a significant timing omission exists in the Sonet specification.
The video decoder is responsible for re-creating the video services from the DS-3 signals received from the desynchronizer. It must recover the clock signal from the transmitted signal, accommodate the various impairments that have been added to the DS-1 and DS-3 signals, and deliver the video signal within ntsc limits. Once again, the principal timing impairments are due to the phase transients of the pointer justifications.
Network interfaces
Important testing has to be performed at the termination interfaces of the network, because that is where services are re-created after any network impairments have been added. Sonet and synchronization interfaces can, of course, be tested throughout the transport network to aid diagnosis. When commissioning, or when service problems are apparent on the downstream video signal, each desynchronizer interface should be evaluated.
To evaluate the DS-3 service interface according to standardized parameter limits, the frequency, frequency offset, and interface noise (jitter) must be evaluated. In addition, the jitter spectrum must be analyzed to identify the presence of pointer activity (phase noise that predominantly exists from 1 to 100 Hz).
The video service interface can be measured using traditional video measurement sets, after it is reconstructed by the video decoder. Signal parameters need to be checked against ntsc specifications to ensure that subsequent video equipment will operate correctly. The key ntsc timing specifications are those for
color subcarrier (Fsc) offset (<䔮 Hz = ۬.8 ppm),
Fsc drift-rate (<۪.1 Hz/sec = 0.028 ppm/sec),
horizontal sync (H-sync) jitter (<۫.0 nsec peak-to-peak).
Test challenges
A number of test challenges must be resolved to ensure that Sonet networks can reliably support video services. First, specifications must be developed to ensure that individual equipment such as desynchronizers and decoders will perform appropriately in a Sonet network, and that end-to-end services are possible using multiple vendors` equipment. In addition, sufficient operating margins must be ensured under all normal (that is, non-fault) network conditions. Lastly, efficient troubleshooting and diagnostic tools must be made available to handle such networking problems.
Fortunately, several solutions are available to overcome these video service problems. One approach is to deploy a special Sonet desynchronizer whose phase control performance greatly exceeds that of normally specified Sonet equipment. Therefore, a video decoder with even transparent phase noise transfer would work satisfactorily.
Another approach is to deploy a video decoder that can accept the maximum network limit of DS-3 wander and jitter (including pointer-induced degradation) without causing the video timing parameters to exceed their specifications. This decoder would then work satisfactorily with any standard Sonet terminal equipment. Or, as a last resort, the video decoder could be followed by a stand-alone frame resynchronizer or time-base correction device to additionally filter low-frequency phase variations.
Which approach is best depends on the needs of the network user, service-provider, or network operator, taking into account interworking constraints, economic factors, and other equipment capabilities. The available solutions require careful performance evaluation under controlled test conditions and comparison against specifications to ensure correct network operation. However, standard, open-interface specifications, performance metrics, and test methods are still needed to ensure reliable operation and interworking of different Sonet equipment. ansi and itu are expected to initiate work on such specifications.
Sonet video performance
Existing metrics do not fully specify equipment or timing/synchronization performance for transporting video services over Sonet networks in a robust way. Needed improvements can be categorized into two areas: service- independent specifications and video-specific specifications (see Fig. 2).
Sonet/DS-3 desynchronizer--When a desynchronizer is used with a video decoder, video timing quality should be specified, again using standard Sonet pointer-test sequences to stress the desynchronizer. A video-measurement test set can measure the equivalent video Fsc offset and drift-rate, thereby providing information that can help select and characterize a decoder. Parameters can be derived and estimated directly from the DS-3 service.
These new video timing quality metrics allow fast assessment of a DS-3 interface. If a high Fsc drift-rate is measured from the desynchronizer, then a high-performance decoder is needed. Conversely, if a low Fsc drift-rate is measured, a decoder with lower performance can be used.
The DS-3 tributary phase response of the desynchronizer should be evaluated in the presence of Sonet pointer justifications. The amplitude, frequency content, and duration of the phase transient provide insight into desynchronizer behavior.
DS-3 video decoder--The decoder should produce an output video signal within ntsc timing-specification limits when its input DS-3 digital video signal has the maximum wander, jitter, and phase transients allowed at the DS-3 network interface.
The phase transfer function or bandwidth specification describes how incoming jitter and wander are filtered by the decoder timebase recovery function. It places a minimum limit on the filtering required.
DS-3 service interface--During network operation, the DS-3 interface wander must be evaluated according to parameter limits. Again, a new specification is required; Bellcore and ansi have recently proposed such a specification.
During network operation, video signal parameters can be estimated from parameters measured directly on the DS-3 service. Fsc offset (phase velocity), Fsc drift-rate (phase acceleration), and horizontal sync jitter (jitter above 10 Hz) can be estimated. A fast comparison can be made between these values and the maximum equivalent values allowed by the network interface. If the equivalent network limit is exceeded, there is likely a video problem downstream.
During network operation, DS-3 tributary phase transients or "pointer hits" can be analyzed to identify the presence and measure the magnitude of pointer-induced timing degradation. Synchronization problems in the upstream Sonet network can be easily flagged.
Video service interface--At the final video service level, new measurement definitions and tests are required for video H-sync wander. These definitions include H-sync drift-limit (Hz), H-sync drift-rate (Hz/sec), and H-sync jitter (nsec). They can then be directly compared with the appropriate ntsc performance limits.
Sonet is ready to transport video services. However, Sonet equipment requires careful evaluation against various performance parameters, many of which are not yet defined or standardized. Sonet networks also require proper qualification if operators are to reliably deploy DS-3 video services. New test tools meet these needs and will allow the full benefits of Sonet to be realized by video service providers. u
Mark J. Lum is telecommunications market development manager, Europe, at Tektronix Inc., in Marlow, Bucks, UK, e-mail: [email protected]. The author acknowledges the contributions of the following Tektronix experts: Dan Baker, video-service performance requirements; Steve Blazo, derived video timing quality measurements; and Dan Wolaver, network wander specifications.