|
|
 |
Understanding SONET Jitter Specifications
Synchronous optical network (SONET) and synchronous digital hierarchy (SDH) are becoming more ubiquitous as the transport layer for wide area network (WAN) connections. The SONET standards define stringent requirements on all aspects of the network to ensure interoperability of equipment made by different manufacturers. This article will explain the jitter specification of the physical interface that connects SONET equipment.
SONET was first defined in 1988 by the Exchange Carriers Standard Association (ECSA) for the American National Standards Institution (ANSI). This standard was also jointly developed by the International Telecommunication Union - Telecommunication Standardization Sector (ITU-T) to set international compliance standards.
Asynchronous vs Synchronous
In order to understand why SONET has such stringent jitter specifications, it is important to understand the difference between an asynchronous and a synchronous network. SONET, as the name implies, is a synchronous network.
Equipment in asynchronous networks may transmit data at different rates. For example, two systems in the network may transmit data at slightly different frequencies. The slower system will loose data (or buffer data) in order to receive information from the system transmitting data at a slightly faster rate. Equipment in synchronous networks must transmit data at the same rate or frequency. However there may be phase differences between equipment due to jitter and propagation delay. SONET jitter specification limits these phase differences.
There are several components that contribute to jitter in SONET equipment. Fig 1 shows the blocks that make up the transport layer for a SONET system.
The transmit function starts with the framer. The framer is responsible for framing data into the SONET payload envelop (SONET SPE), managing the SONET overhead and scrambling the entire frame. The serializer is responsible for converting a wide parallel data stream into a single high-speed data stream. The clock generator provides a reference clock to the transmit phase lock loop (TX PLL). This reference clock is multiplied by TX PLL and is used to clock serial data out of the device to the laser driver. The laser driver converts the voltage from the serializer into a current that is used to drive the laser.
The receive function starts with the photo detector. The photo detector detects light and outputs a small current. The trans-impedance amplifier (TIA) converts this small current into a small voltage. The post-amplifier or limiting amplifier amplifies this small voltage into a larger voltage that drives the clock and data recovery (CDR). The CDR extracts the clock from the data stream. This clock is used to determine the bit boundaries of the incoming data stream and to clock the deserializer. The deserializer converts the serial stream into a wide parallel data stream. The framer then de-scrambles the data, monitors the SONET overhead, and removes data from the SONET SPE.
The SONET jitter specifications are governed by Telcordia Technologies Inc. (formerly Bellcore) GR-253-CORE specification. Telcordia specifies two types of jitter. Category II defines the jitter for OC-N and STS-N electrical and synchronous interfaces. SONET equipment must meet jitter generation, jitter tolerance, and jitter transfer. Jitter is defined as the short-term variation of a digital signal's significant instants from their ideal position in time. These classes of jitter are specified for the system not for the individual components that make up the system. The components that contribute to category II jitter are the reference clock, SERDES (CDR and TX PLL) and the optical module (laser driver, laser, photo detector, TIA, post amplifier).
Jitter Generation
Jitter generation is the amount of jitter added to a data signal by the transmit node. Telcordia specifies limits of 0.1UI peak to peak and 0.01UI rms (root mean square). For measurement purposes, a band-pass filter is used to limit the jitter to the frequencies of interest. The roll-off for each SONET level is 20dB/decade after the cut-off frequency. The band-pass frequency was chosen to measure only the jitter frequencies that impact the SONET network. Jitter, outside this filter, would be added to the measurement of the node but the network would follow the low and attenuate the high frequency jitter. Jitter below the low-pass cut-off is assumed to be wander. The receiver will follow the low frequency wander in the system. Jitter above the high-pass cut-off frequency is assumed to be attenuated by the receive system. Fig 2 shows an eye diagram for an OC-48 output. Peak to peak jitter is measured as the width of the crossing of the rise and fall signals on the eye diagram. RMS jitter measures the distribution of the crossing pattern.
Jitter tolerance is the amount of uncertainty, in the signal (jitter), the receiver can tolerate while preserving the system's ability to recover data. Jitter tolerance is measured by adding jitter to the input signal of the serial interface until a -1dB loss is measured. Normally manufacturers add jitter to the input signal of the serial interface until a BER of 10-12 is measured.
All equipment in the SONET network must meet the jitter tolerance and jitter generation specifications. Network equipment that uses a clock derived from the incoming signal, to clock the output signal, must also meet jitter transfer. This mode of operation is called line-timing and allows the equipment to be clocked from a single incoming OC-N signal. The jitter transfer specification limits how much jitter from the incoming signal may be passed on to the outgoing signal. It is measured as the ratio of the output jitter to the input jitter. This limit is set to allow multiple pieces of equipment in the network to use line-timing mode while insuring a low BER by limiting the amount of jitter that is amplified at each piece of equipment. Fig 3 shows the jitter transfer requirements for OC-N rates.
SONET and SDH standards govern all aspects of the network. Compliance to SONET jitter specification insures physical compatibility between various manufacturers equipment. Understanding the basics of jitter generation, jitter tolerance, and jitter transfer will help system designers understand how SONET equipment interoperates at the physical layer and to insure their systems are complaint.
by Michael Bollesen,
Director of Strategic Marketing,
DataCom Division,
Cypress Semiconductor,
USA
(June 2002 Issue, Nikkei Electronics Asia)
|
