18 Sep 2010
Measuring Jitter on Synchronous Ethernet Networks
Andreas Alpert, Senior Product Manager for Optical Transport, JDSU, looks at the way in which jitter can be measured on synchronous Ethernet networks
Synchronous Ethernet (SynchE) has largely arisen as a solution to two challenges faced by telecoms operators. Firstly, as the demand for broadband services has increased operators have had to find a way to ensure they can transport the additional traffic (largely originating from bandwidth-intensive applications such as broadcast TV, Video on Demand and IPTV) at a low enough cost-per-bit to guarantee they are profitable. Secondly the rise of mobile broadband has led to a requirement for a low-cost, scalable backhaul technology that can ensure a high end-user performance.
Ethernet has quickly emerged as the ideal connectivity medium for operators looking to address these challenges. However, it is not without its own problems. Ethernet was originally designed as a Local Area Network technology and as such lacks several key attributes necessary for mobile backhaul. For example, frequency synchronisation and stability are essential at mobile base sites if the radio spectrum is going to be used efficiently and the handover between sites conducted effectively. While this has been available on SDH/SONET networks due to the inclusion of a timing reference in the network architecture, this is missing in the native Ethernet standard. SynchE has been designed to fill in these gaps, providing the timing reference operators require to synchronise with frequency modulations at the mobile base station.
SynchE provides a means of transferring frequency over the Ethernet physical layer, traceable to a network clock. Due to this SynchE is considered a part of the synchronisation network. SynchE must fit within the general architecture of the Ethernet network. Based on the ITU-T standards , SynchE has a clock accuracy of &plumn;4.6 ppm (parts per million) similar to the range found on SDH/SONET, a great improvement on the native (asynchronous) Ethernet limit with a free-running clock of +/- 100 ppm accuracy. In non-synch mode, however, SynchE is identical to IEEE 802.3 and can interwork with conventional Ethernet interfaces. Importantly SynchE has been architected to interoperate with the large base of existing SDH/SONET synchronisation networks in existence today, allowing operators to migrate to Ethernet transport and backhaul at a pace that suits their business requirements.
Synchronisation clocks are part of the network element : SDH Equipment Clock (SEC), SONET Minimum Clock (SMC) and Ethernet Equipment Clock (EEC).
SynchE will, therefore, prove vital in the rapid migration to NGNs based on Ethernet-based packet networks. Ethernet is the technology of choice for both circuit emulation services (CES) and mobile backhaul but will only live up to its potential through the frequency synchronisation offered by SynchE.
What is Jittter and why is it an issue for operators?
The demands of modern networks place great strain on synchronisation. As voice, video and data are transported over a common network, the phase stability of clock and data signals can be impaired as factors such as crosstalk, clock degradation, noise, bandwidth limitations, mapping and framing prevent perfect synchronisation. This is called jitter. Jitter can affect the quality of transmission leading to bit errors, slips and data loss. For end users this can mean a poor experiential quality and disruption to their internet sessions and related services (be that poor voice quality or an uneven viewing experience of video content).
When designing, developing, interconnecting and maintaining a SynchE network it is essential for operators to factor in the detection of jitter if they are going to be able to guarantee the quality of the network. If this is not done correctly, operators risk customer churn as subscribers move to competitors that can offer a superior user experience.
There are three facets of jitter that need to be measured when deploying and maintaining SynchE enabled networks: jitter generation, jitter tolerance and jitter transfer. The jitter limits have been determined by standardization bodies like ITU-T, Telcordia and IEEE. The specifications and test methodologies for jitter on native Ethernet differ from those for SDH/SONET/SyncE:
The SynchE jitter measurements defined by the ITU are comparable to the SDH/SONET measurements. These recommendations outline requirements for network limits (ITU-T G.8261) and SyncE clocks (ITU-T G.8262 ) . In addition it defines the wander performance of SyncE circuits.
Any measurement solution needs to be able to address the essential tasks of characterising both the jitter and wander performance so that critical timing and synchronisation functions can be maintained without signal degradation or network performance problems. As SynchE is based on the ITU-T standards the test environment considerably easier, removing the need for special test equipment with complicated calibration settings as would be the case on the Ethernet native standard.
It is vital that operators deploy test and measurement solutions for intrinsic jitter measurement to get enough margin to the specified jitter limits required by ITU-T standards. These measurements help determine the amount of jitter acceptable at SyncE interfaces. The standard for jitter test equipment has recently been set down and is defined in ITU-T 0.174:
To ensure that Ethernet has the necessary attributes to be truly carrier grade, operators and vendors are introducing several key technologies for the transport of timing and synchronisation over packet networks. Of these Synchronous Ethernet is perhaps the most important. Selection and verification of SynchE however requires careful analysis both in the lab and in the field. A core part of this verification is the measurement and characterisation of jitter performance at the Ethernet interfaces. With the emergence of jitter test and measurement devices for SynchE operators can now leverage the potential of the technology while ensuring that quality levels can be maintained. This will prove vital for operators as the build business cases for implementing SynchE in the WAN and backhaul environments.
Page 1 of 1
About the author
Andreas Alpert, JDSU Product Line Manager for the ONT-5xx family is based in Eningen/Germany. Andreas has been with Wandel & Goltermann, Acterna and JDSU for 33 years in both Engineering and Marketing roles. He has been actively contributing to standards bodies, especially ITU-T.
JDSU is the worldwide provider of broadband test and measurement solutions and optical products for communications, commercial and consumer markets.