Exploring the problems and examining the solutions that embedded synchronization brings to integrated voice and data networks.
DAVE FIGGE and NINO DE FALCIS,
Datum Inc.
Network design engineers are in a pinch. Voice and data convergence sounds like a smooth, integrated process, but unfortunately, the reality is that the data communications and telecommunications worlds are actually colliding. Equipment manufacturers from both sides are increasingly driven to design and build networks that can carry the services of the other. Networks that once transmitted data are carrying voice and video as well as Internet traffic. Networks that traditionally carried only voice now have to transport high-speed broadband multimedia services.
Legacy networks built on circuit-switched TDM technology have extended their grasp to include cell-switched ATM over SONET, DWDM, and now the packet-switched option of IP. Not only are traditional telecom manufacturers broadening their products to support IP, datacom manufacturers are crossing over to deliver switching systems to telecom carriers.
Datacom carriers that built architectures with a mix of ATM and IP core functionality have added voice to their array of high-speed bundled options. Supporting this diverse array of traffic and protocols is the new breed of multiservice-network equipment deployed by both telecom and datacom carriers.
One platform led to another, but data-network planners soon discovered something was forgotten in all of the excitement. No matter how many cleverly bundled offerings they began to provide, without some quality-of-service (QoS) guarantee, none of it was going to work effectively or profitably.Extremely precise synchronization, a long-time latency buster or bandwidth manager in the telecom world, continues to be a relatively new concept for most data-network designers. Traditionally, they simply derived a timing signal from the backbone of a network. But now, these data carriers, with their bundled services, are becoming the primary network.
Therefore, they must supply their own synchronization source. As soon as these designers realized that a lack of reliable timing could sabotage even the most carefully planned multimillion-dollar network, they recognized the need to design a quality synchronization solution into their networks.
Synchronization is criticalSynchronization of a network occurs when its elements operate in unison with an internal clock providing a specified timing or rhythm. If a network is in sync, it can receive and read transmitted signals at the same rate they were sent. When a network is out of sync, that transmission often slips and causes static on a voice call. The affect is most dramatic, however, in data transmissions where sync problems can result in lost packets of data. Noticeable screen jumps and garbled text are just a few of the results of a network out of sync.
Network synchronization is critical to deliver high QoS and maintain the integrity of a network transmission. Without it, bottlenecks can occur and phase errors-jitter and wander-can accumulate throughout the network. Data is particularly sensitive to such timing problems, which often require a signal to be retransmitted. New applications, such as voice or video over IP, which require constant-bit-rate services, are creating an even larger need for critical synchronization.
Industry problem
Industry standards for synchronization have yet to catch up with today's intricate level of network interconnection. Designs tend to diverge from the common framework established by legacy carriers, whether or not that is appropriate for a new network. Conversely, legacy designs try to be force-fit into a new network offering new services. Network designers quickly discover they do not have the design skills or resources to account for synchronization.
Purchasing an equipment synchronization clock solution is just the tip of the proverbial iceberg. Synchronization is a very complex component vital to overall network performance and reliability. It is also an exact science that is well understood by a relatively few number of industry experts.
Designers therefore rely on the expertise of equipment manufacturers to properly embed synchronization clocks in their network equipment designs. When designers piece together various options instead of using a complete approach, they risk having to make major redesigns as well as prolonging the development process of non-core technology.
According to telecom analyst Jim Rawitsch, designers that try to build their own synchronization systems have to understand a lot of the subtleties of the technology to ensure it meets all the clocking requirements, particularly things such as wander parameters. Without substantial experience, these designers are facing an uphill battle against time and difficulty in meeting sync requirements.
Embedded clock solutionTwo years ago, a director of advanced technologies for a large synchronization equipment supplier was visiting a carrier office to oversee the installation of synchronization equipment. Coincidentally, a team from one of the most prominent router manufacturers was also onsite, trying to solve a continuing transmission problem involving missing data packets. The director overheard the manufacturer's discussion and suggested synchronizing the system by connecting to a timing signal generator and synchronization supply-unit clock.
The test worked, immediately solving a problem that had perplexed the router team for days. That was a great first step. However, the team said they needed a solution that could reside on a circuit board and be embedded in their own equipment. The engineers quickly went to work and developed a multiservice clock chipset that could satisfy embedded network synchronization requirements for core, edge, or access equipment.
The embedded synchronization technology consists of intelligent plug-and-play clock chipsets that allow designers to customize a sync solution by selecting from a wide variety of basic timing functional pieces to meet specific core, edge, or access application requirements.
The clock chipset can be seamlessly integrated to be fully redundant or not; auto qualify and filter input signals or not; achieve wander filtering based on software selection and oscillator; process input/output Stratum-level traceability data from the network through sync status messages; and achieve holdover characteristics based on a unique oscillator type or software-compensated oscillator. In addition to hardware configuration, there is also the capability to manipulate or customize each chip's performance and possible reaction activity via the network equipment's host system software, allowing performance customization of a standard hardware set.
As the myriad of networks continues to grow in number and complexity, design and equipment manufacturers are expected to gravitate toward the adaptive network-element clocks. Soon, the mystery will be taken out of synchronization, and network design engineers can be assured their customers will continue receiving QoS they've grown to expect, regardless of the application or burden on the network.
Dave Figge is vice president of product marketing and Nino De Falcis is director of marketing communications at Datum Inc. (Austin, TX). They can be reached via the company's Website, www.datum.com.
Lightwave is a monthly international publication focusing on fiber optics and optoelectronics, the technologies driving the growth, convergence, and improved performance of telephony, computer communications, and video. Lightwave provides technology news as well as applications and product information for corporate and technical managers and staff engineers. Lightwave's editors emphasize analysis and interpretation in their reports on the technological impact of fiber-optic components, systems, and networks in these markets.