justone,
I understand the direction you are going in, and this is a common perception, but fifty percent of all SONET bandwidth is not wasted. The following points should serve to offset such notions.
While there are some SONET Rings that employ self healing to actual end user (subscriber) locations, most SONET feeds to customers are linear in nature. They are not ring-based, although they may meet up with a ring somewhere along the route to the Central Office, or in a feeder office, at an add-drop multiplexer node on a ring.
But most subscriber feeds are linear and do not contain standby bandwidth. It's possible, however, to implement two linear "spurs" to a customer site and perform automatic protection switching (APS) between them, similar to how one would back up a T1 with an A-B protection switch. In this manner, if one goes out the other one takes over. But this is not what you are referring to. You are referring, instead, to the standby bandwidth that is inherent in the design of self-healing bidirectional ring architectures.
SONET Rings of the latter type "do" contain standby bandwidth. These rings are usually 2-fiber or 4-fiber (although some single fiber rings exist in limited situations, as well, but we'll discount those here).
A further drill down reveals additional subclassifications of 2- and 4-fiber rings, based on what is being switched: you have both line-switched and path switched architectures... but we'll leave these finer points for another time.
For the moment, simply consider that SONET rings may be either 2- or 4- fiber based.
2-fiber rings do contain 50% standby capacity that serve to satisfy "self-healing" during breaks. For example:
A 2-fiber OC-48 SONET ring can only support up to 24 T3s on a guaranteed basis. The other 24 T3s are not protected (they "serve" as the protection for the first 24 in this case), and are referred to as standby capacity.
However, carriers do sell this standby bandwidth on 2-fiber rings to ISPs and other organizations on a more tentative and discounted basis, with the prior knowledge that they may be "preempted" or "bumped" when a break occurs. Such leads to a form of double jeopardy for these subscribers -- who are very often other service providers down the food chain -- if you think about it. They get knocked out of service when fiber fails "anywhere" along the ring (on any arc of the ring), even if their end-point pairs are not on the arc of the ring that was compromised. But cash-strapped ISPs and other smaller service providers will opt for this, especially if their networks are resilient at the upper layers (reroutable), but more often to soften the cost impact of their day to day operations.
During the latter periods (preemption), the standby bandwidth is taken away from the ISP or other organization and it is used to support the guaranteed service. Here the erstwhile user of this bandwidth must be prepared to recover through some other means. ISPs will seek to use the inherent rerouting capabilities of the Internet Protocol (Layer 3). Or, they may have a protection path in place that they can switch to at Layer 1 (such as the APS approach I alluded to above).
It's the user's responsibility, in the end, to manage their own "mesh" network connections when they decide to forego the SLA guarantees that are afforded by self healing ring provisions. And this appears to be gaining a lot of interest lately, both in the use of legacy carrier network facilities, and at higher speeds using WDM and optical switching and routing.
4-fiber rings, on the other hand --and let's use the same OC-48 example again -- employ all 48 T3s for production purposes. And during fiber breaks all forty-eight of these T3s are recovered, unlike the 2-fiber ring's ability to only recover 24 T3s.
Of course, the tradeoff here is the additional two strands of fiber and the additional electronics that are needed to make this work in the SONET ADM nodes.
NOTE: In the above examples you can substitute the terms "lambda" and "wavelength" for "fiber," when the SONET ring uses them. Because a lambda, in this case, is nothing more than virtual fiber, anyway. Although, in either case, i.e., in both the 2-fiber or 4-fiber models I alluded to above, each lambda that is used in the scheme should be derived, theoretically, from a fiber that is distinct from any of the others being used to support the other lambdas.
So that, lambdas from 2 distinct fibers would be used in the 2-fiber ring model, and lambdas from four distinct fibers would be used in the 4-fiber ring model.
Finally, there are large regional networks which are made up of linear SONET overlays that are used for the voice switching network. These usually terminate in digital cross connects (DCSes, or Digital Access-Cross Connects (DACSes), and often do not enjoy the benefits of all of their routes being on self healing rings, at the fiber layer.
Sometimes (as if we needed a way to prove this), not often, but sometimes, it takes a DCS mesh many hours, sometimes almost a full day, to fully rebuild their connection tables and recover, after a fiber or central office failure has occured.
In this respect we see cases where instead of wasting bandwidth for standby purposes, the carrier may actually be too frugal with the stuff. Along these same lines, I am a frequent lurker on the ISP lists, and I often come across messages that speak of major turmoil due to fiber breaks. In some of those cases, the ISPs (many of them mom-and-pop- sized) get killed because they are on the short end of the recovery stick. By choice (or perhaps out of necessity), some of them have placed themselves in the double jeopardy category that I briefly talked about, above, and they must always make provisions to leave themselves a back door out.
FAC |
