Network operators face the difficult task of ensuring their networks remain operational, even during fault conditions such as fiber cuts, equipment failures, and human errors. Very high reliability is achieved by incorporating layers of redundancy and resiliency in modern optical networks. Using advanced protection and restoration techniques, network failures can be instantly detected and services automatically re-routed around the impacted areas.
Advanced restoration techniques that protect modern networks are typically implemented on routers or Optical Transport Network (OTN) switches. Until recently, network restoration options at the optical layer were somewhat limited, compared to the more sophisticated methods incorporated at the higher layers.
Increasing demand for bandwidth and high-speed services is causing ever higher network capacities and wavelength speeds. As network capacities increase, so does the cost of providing restoration using the router or OTN layers, since all services must terminate at every router or OTN switch, at every node location. The transponder and optical interface costs of terminating hundreds of gigabits (or even terabits) of network capacity on both the WDM and router layers, at every location, account for the high expense.
By comparison, optical networks provide efficient, low-cost service transport over metro, regional, and long-haul distances. Until recently, optical networks lacked the sophisticated restoration methods found on routers and OTN switches. But now, more advanced optical layer protection and restoration are being integrated into new optical platforms. These offerings combine the benefits of operating at the optical layer with powerful network restoration features.
Legacy optical protection options
Optical networks have traditionally provided either unprotected or 1+1 protected services, as shown in Figure 1. Dedicated 1+1 optical protection offers a high level of redundancy for mission-critical services with guaranteed, ultra-fast 50-ms switching. However, this level of performance comes with a steep cost, since 50% of the network capacity is reserved and unavailable.
|Figure 1. Traditional optical protection options.|
For services that are truly mission critical, operating networks with 50% idle capacity is an acceptable tradeoff to achieve the highest performance and service availability. However, for general-purpose wavelengths — those that aren't mission critical — reserving 50% of capacity for protection is both costly and inefficient.
Additional protection and restoration options that go beyond the legacy techniques are needed at the optical layer. As networks evolve to higher wavelength speeds and larger capacities, optical layer restoration becomes critical, since relying on the upper layers for restoration becomes increasingly cost prohibitive.
Improved network efficiency with optical layer restoration
Optical layer restoration re-routes and restores traffic around network faults, without the 50% network penalty imposed with dedicated 1+1 optical protection. Operators gain the benefits of automated service restoration with much higher network efficiency and capacity utilization, resulting in lower costs and more room to carry additional revenue-generating services (Figure 2).
|Figure 2. Optical layer restoration.|
When a failure occurs, the network automatically detects the fault, calculates a new restoration path, and re-routes any affected wavelengths over the new path, all without manual intervention. In effect, the network monitors and heals itself.
The ability to share protection resources, as well as re-route wavelengths in real time, is the key to greater network efficiency. It improves network utilization, increases network capacity, and lowers overall costs. In meshed real-world networks, a single shared restoration path can support up to four or five working paths, resulting in as little as 15% to 20% of total capacity needed for network protection.
In addition, carriers can mix 1+1 protection for mission-critical services with wavelengths using optical layer restoration on the same network, achieving the optimal combination of performance and efficiency.
Enhanced service assurance with fast restoration
When fiber cuts occur, it often takes 24 to 72 hours to repair the cable and restore network services. Wavelengths using dedicated 1+1 protection switch in less than 50 ms, but unprotected wavelengths remain out of service until the fiber cut is repaired. For many carriers, an outage lasting this long is simply unacceptable, even on unprotected, best-effort wavelengths.
A key advantage of optical layer restoration is that it automatically detects fault conditions and re-routes wavelengths around the impacted areas. Optical restoration takes a little longer to complete than dedicated 1+1 protection switching, but it offers a vast improvement over three-day repair times on physical cable cuts.
Dynamic, flexible optical networking
Until recently, wavelengths running over WDM networks were "static," with a wavelength route assigned as services were provisioned. Making path changes to these static wavelengths requires substantial time and manual re-provisioning, so has rarely been performed.
Supporting flexible, dynamic wavelength routing and steering, without requiring manual intervention, are key to modern high-speed optical networks. Colorless/directionless (CD) and colorless/directionless/contentionless (CDC) ROADM architectures provide the underlying flexibility needed to automate and simplify network operations and reduce network costs by supporting advanced features, such as optical layer restoration.
Randy Eisenach is responsible for optical networks product marketing at Nokia. He specializes in optical transport technologies, next-generation wavelength routing architectures, and high-speed photonics. Randy has more than 25 years of optical transport and networking experience and has held a wide range of senior level positions in systems engineering, product management, and product marketing. Randy recently joined Nokia after holding similar positions at Fujitsu and Nortel.