Talk:Synchronous optical networking
From Wikipedia, the free encyclopedia
Contents |
[edit] Digital Hierarchy
The statement "For example, four OC-3 or STM-1 circuits can be agregated to form a 622 Mbit/s circuit designated as OC-12 or STM-4." is a bit misleading. Actually an OC-12 is not an encapsulation of 4 OC-3s as is implied. An OC-12 contains 12 STS-1s (OC-1 at optical level) and as such can carry 4 OC-3s, however each STS-1 is carried without any type of "wrapper" within the OC-12. I'm probably being overly picky but hey, you might as well make it as clear as possible.
"The current state of the art is the OC-192 or STM-64 circuit, which operates at rate of just under 10 Gbit/s." The current state of the art is actually OC-768.
"It is defined by GR-253-CORE from Telcordia." Isn´t it defined by ITU-T´s G707?
- The "agregated" is still there... why didn't you fix it? Fixed. Mga 03:05, 29 May 2005 (UTC)
I think this article lacks a table of the digital hierarchy defined by SONET and SDH, as well as a comparison of the terminolgy. There is more to be done (the article lacks a bit more overview; it did not even mention Time Division Multiplexing, ANSI T.105 or refer to the proper standards. I fixed a few of these, but am not an SONET/SDH expert. For starters, I like to add this subsection in the "The basic unit of transmission" section, but I like someone to verify its correctness. I put question marks next to figures I am not certain about. (Feel free to fix them here, so I can move it to the article, or move it yourself). MacFreek (talk) 15:09, 2 March 2008 (UTC)
[edit] Digital Hierarchy
SONET and SDH define a hierarchy for interweaving multiple channels into a single link. The different channels are interwoven using Time-division multiplexing.
At the highest layer, an Optical Carrier (OC) can carry a large amount of data. For example, a single OC-192 can carry WAN PHY Ethernet. However, it is sometimes convenient to multiplex the signal into smaller channels. An OC-192 can also carry 192 STS-1 SONET Payload Envelopes (STS-1 SPE) or 64 VC-4 channels. The STS-1 SPE or VC-4 can be further split into smaller channels, called Virtual Tributaries (VT) in SONET terminology or Virtual Containers (VC) in SDH terminology. The individual VT or VC can further carry one or more digital signals (DS), such as DS1, the payload of T1 or E1.
The hierarchy is slightly different for SONET and SDH. The following table follows the hierarchy used in GMPLS to describe SONET and SDH links[1].
| SONET layer | SDH layer | Data Rate (kbit/s) | Composition |
|---|---|---|---|
| VT-1.5 SPE | VC-11 | 1728 | Virtual Tributary or Virtual Container, enough to carry a single DS1/T1 of 1,544 kbit/s. |
| VT-2 SPE | VC-12 | 2304 | Virtual Tributary or Virtual Container, enough to carry a single E1 of 2,048 kbit/s. |
| VT-3 SPE | N/A | 3456 | Virtual Tributary or Virtual Container, enough to carry a single DS1C of 3,152 kbit/s. |
| VT-6 SPE | VC-2 | 6912 | Virtual Tributary or Virtual Container, enough to carry a single DS1C of 6,312 kbit/s. |
| VTG SPE | TUG-2 | 6912 ?? | Virtual Tributary Group or Tributary Unit Group. A VTG or TUG-2 can carry one VT-6/VC-2, two VT-3, three VT-2/VC-12 or fout VT-1.5/VC-11. |
| STS-1 SPE | VC-3 | 50,112 | A STS-1 may carry up to seven virtual tributary groups. |
| N/A | TUG-3 | 49,536 ?? | A TUG-3 may carry up to seven TUG-2. |
| STS-3c SPE | VC-4 | 150,336 | Payload of a STS-3. |
| STS-3 | AUG-1 | 155,520 | The analog carrier (?) of an STS-3 SPE, including headers. |
| OC-3 | STM-1 | 155,520 | The optical carrier (?) of an STS-3 SPE, including headers. |
This table does not define lower orders than VT1.5. in particular DS0 is not defined, nor the higher order concatenations of DS0: DS1/T1, DS2/T2, DS3/T3, DS4/T4, DS5/T4, E1, E2, E3, E4 and E5.
[edit] 10 Gigabit Ethernet and OC-192/STM-64
This article used to say they were compatible, but now it doesn't. Could somebody knowledgeable elaborate? - mako 06:29, 26 Apr 2005 (UTC)
The signals are not directly compatible, however many manufacturers produce cards that allow 10GigE signals to be carried over SDH/SONET or STM-64 DWDM signals (also variants that carry 1 or 2x 1GigE signals over STM-16). For example, Marconi MSH2K Optical Core Switch (link to commercial site) - gjk 10:55, 13 July 2005 (UTC)
- Actually, 10GbE has two different types of PHYs. One is the LAN-PHY which looks a lot like the lower speed Ethernet PHYs. The other is the WAN-PHY which was specifically intended for interworking with the SDH/SONET network: it has the same frame format as an STM-64 with VC-4-64c (OC-192 with STS-192c SPE) - bit rate is a bit lower than the LAN-PHY. A WAN-PHY which additionally observes the SDH/SONET clock specifications fits seamlessly into the SDH/SONET network. BertK 22:21, 15 July 2005 (UTC)
- Also note that a single NE's compatibility with 10GbE doesn't mean that intermediate NEs will be able to support it. 10GbE is very close in flavor to STS-192c, which most NEs don't support. Also, 10GbE doesn't really make use of a lot of the SONET overhead and features: These are basically turned off which could drive an actual SONET Network crazy.
Aside from that, many new SONET NEs support it, though frankly when you're concatenating at that level you'll probably have optical networking equipment (as opposed to SONET) do things like protection.
[edit] NE?
Suddenly in the text is the abbreviation "NE". What's that?
- Network Element BertK 18:04, 23 January 2006 (UTC)
A network element contains (is made up of amongst other things) racks/shelfs/cards/slots/ports. —Preceding unsigned comment added by 218.186.14.42 (talk) 23:43, August 27, 2007 (UTC)
[edit] Equipment vendor listing?
An anonymous user that has a decent editing history originally added three links to different Nortel systems that do SONET. I removed those. A different link has now been added back, again by the anonymous user, to a top level for all SONET systems at Nortel. Do we want a list of vendors on the article at all? If so, it will slowly go out-of-date as vendors merge, change names, go out of business, or new ones are created. Having the list also invites more churn on the article. Not only that, but it isn't a short list (Here is an incomplete list):
Alcatel,
Anda Networks,
Appian Communications,
Cisco Systems,
Coriolis Networks,
ECI Telecom Communications,
Fujitsu Network Communications,
Internet Photonics,
Lucent Technologies,
Mahi Networks,
Marconi,
Nortel Networks,
Photonic Bridges,
RAD Data Communications,
Sycamore Networks,
Siemens ICN,
Tejas Networks,
Tellabs,
Turin Networks,
White Rock Networks
Or we could add several links to the ODP (dmoz.org), who actually attempts to be a directory. Mrand 00:40, 24 February 2006 (UTC)
gea-141020: Agreed, the list of SONET equipment vendors doesn't belong in this article. Eventually, however, it might be useful to have an article that shows North American SONET equipment deployment by network and Vendor, although there's a lot of other SONET-related articles that should be written first. In that context, of course, Nortel will seem pretty important, along with Lucent, Alcatel, Fujistu and, possibly, Cisco. Maybe DSC, too. Those vendors probably outweigh the little guys by 1000s to one, so there is merit in mentioning names within the right context.
[edit] SDH != SONET
[edit] Discussion
If someone wants to specifically know about SDH, I think this is the wrong article. There are differences (not only in labels), and there is certainly more (channels, how they are multiplexed, headers, units, pointers etc.).
Therefore I propose that a new article, solely about SDH, is forked from this one.
- I disagree that a forked article is needed. Here's my opinion: SDH is so close to SONET that a forked article would result in massive amounts of duplicated information. Units (are you referring to Mbps?), and pointers are all the same. Headers (you referring to overhead?) are almost identical. Multiplex schemes are somewhat different, but have massive overlap. I'm not sure what you are referring to by channels, but if it's the naming difference (which I actually assume is what you meant by "labels"), then yes, they are named different. But a (single) well written article would be able to handle the multiple names. At most, I think a second article would only need to explain SDH naming and multiplexing, while referring back to the main SONET/SDH article for everything else. Regards, Mrand 05:50, 29 March 2006 (UTC)
MW: I just came to this page for the first time. It was immediately apparent that the article contains information on two different things. SDH and SONET are two different standards. Because they are two completely different standards, there should be two completely different articles. If one wishes to summarize Synchronous networks in one page, that would be fine, and that article could link to SONET and SDH. But similarity is not enough reason for one article. There are two seperate articles for IPv4 and IPv6. And those are just varients of one standard. There are probably 1000s of examples of why there should be 2 articles on these 2 different standards. MatthewWilder 20:13, 25 May 2006 (UTC)
- They're not two different standards, they're the same standard separated by the Atlantic. You can literally plug SDH equipment into SONET, and provided the equipment breaks the standard in trivial ways (one standard I think it was SONET says that some overhead bits have to be zero, whereas SDH says it can be one or zero), then the two standards interwork incredibly well.- (User) WolfKeeper (Talk) 21:23, 21 April 2008 (UTC)
agree with u on that one MW: I thought of another example. This article is akin to having an article on Sonny and Cher but naming it "Sonny". And an article on Sonny and Cher is reasonable, since they were compatible (at least at times). Anyway, the names are so close that I couldn't resist this comparison. MatthewWilder 20:21, 25 May 2006 (UTC)
CT: I agree that an explanation of SONET and SDH can co-exist on the same page, but the differences should be clear. The payload data rate gets tricky when you start dealing with virtual concatenation (VCAT) and virtual containers and tributaries, e.g. STS-3c/VC-4 SPE is 149.76Mbit/s whereas STS-3 SPE is 150.336Mbit/s. For more information, see http://www.zytrax.com/tech/data_rates.htm or RFC 3255 at http://www.ietf.org/rfc/rfc3255.txt. ChrisTracy 01:30, 16 June 2006 (UTC)
GB: Sonet and SDH are two seperate, distinct entities. They have different overheads, support different line rates as basic units (VC4 vs. STS.) Frankly, the VC3 had to be invented to create a comparison for the two standards. SDH has LO and HO overheads, where as Sonet has a single overhead. Yes, they both use the same methods for OAMP and mostly the same bytes, but SDH has two levels, and they support different timing sources (same idea coming origianlly from a Sonet clock, but different formats.) Having these in the same article should change the title to Optical Network Protocols or something along those lines. GregBouchillon 09:51, 20 September 2006 (UTC)
-
- Actually, SONET and SDH are intimately related standards: they were developped jointly and published within months of each other. In fact SONET is a so-called 'regional subset' of the international SDH standard, so everything in SONET is also in ITU-T SDH (be it with a different name) - so the above claims must be the result of some confusion. (Also note that currently maintenance of SONET standards is done by changing the SDH standards and then refering to those.) Other regions like Europe or Japan also specify their own subset, but keep calling this subset SDH (like all PDH is called PDH, no matter if its DS1, E1 or J1 based). This seems to be causing confusion, most often between the European (ETSI) SDH and the international (ITU-T) SDH.BertK 20:46, 4 October 2006 (UTC)
- I proposed a split. My guess is that someone smarter than me can make the differences clearer. User:Luis F. Gonzalez
-
- There are no differences per se: there is an international superset specified by ITU-T (formerly CCITT) called SDH and a number of regional subset specified by regional standards bodies like ATIS OPTXS (formerly ANSI T1X1), ETSI, TTC,... For historical reasons, the North American regional subset happens to be called SONET and uses different terminology, but it's still the same [1].
- That doesn't mean a split article is not a good idea: I'd like to see a main article on international SDH, including the joint history, and multiple articles on the regional subsets (ETSI SDH, SONET, others?). The SONET article would be special in that it would have pre-standardization history [2] and a terminology cross-reference.BertK 20:46, 4 October 2006 (UTC)
SDH and Sonet Difference
SDH is used in European Contries Sonet is used in American Conuntries —Preceding unsigned comment added by S.santosh2k (talk • contribs) 12:16, 3 October 2006 (UTC)
- To be more precise: SDH is deployed all over the world, except for North America where SONET is deployed. Huub v H 20:08, 12 March 2007 (UTC)
And both of them have a significant amount of diffrence in Data rates. —Preceding unsigned comment added by S.santosh2k (talk • contribs) 12:16, 3 October 2006 (UTC)
- Also the processing of the overhead information is different between SDH and SONET (see http://www.van-helvoort.eu/ and selct standards work) Huub v H 20:08, 12 March 2007 (UTC)
You definitely have two (or three) separate animals here, although they basically started with the same parent. SONET as a North American standard follows ANSI T1.105-2001. SDH is the European standard and falls under the ITU-T G.707 & G.708 standards. Additionally we can further extend the article if we discuss Optical Transport Networks (OTN) and the ITU-T G.709/Y.1993 (and replace the 10 Gbps discussion). A comparison is required between the three to show their similarities and highlight their differences. As a start, why don't we concentrate on making three separate sections (one each on SONET, SDH, and OTN)? JEdwards47 19:48, 3 October 2006 (UTC)
- SONET and SDH are basically the same or at least closely related animals. OTN is a completely separate standard: SDH/SONET signals are clients to OTN signals, like PDH signals were clients to SDH/SONET networks.BertK 20:46, 4 October 2006 (UTC)
gea-141020: I wrote most of this article and worked at Bellcore/Telcordia in the group responsible for GR-253 (I also co-wrote a book on SONET with the inventor of SONET). In my opinion, one could modify this article to better include the differences between SDH and SONET, which are minimal at the OC-N and STS-N levels. Where you DO need an additional article(s), however, is in handling the subrate packaging of PDH signals. In other words, everything that happens below the OC-3 level, including how you build up from the E1 level to an OC-3. For that matter, another article is probably needed describing the same process for North America. But once you getup to the true SDH/SONET transport containers at the line and section llayer (which have different names in SDH, sorry) then t's all basically the same and can be handled by one article.
A separate article for SDH wouldn't be the end of the world, however.
[edit] Survey
The discussion may continue above.
I vote for the split on the grounds that many readers want a clear definition of what makes SONET SONET. However, I think for those looking for the "Big Picture" and trying to develop a context, a very high-level comparison to SDH is helpful, if only because so many texts lump SONET and SDH into their three-page writeup of "high speed optical networks". Here's what I mean by "High Level":
SONET and SDH
SONET and SDH are both high-speed networks that are most commonly implemented on fiber. The North American and European approaches are similar enough that data can be exchanged between these different network technologies with suitable equipment.
| SONET | SDH | |
|---|---|---|
| Created By | ANSI | ITU |
| Used | North America | Europe |
| Name of Data Unit | OC (Optical Carrier), STS (rare, for copper implemenations) | STM (Synchronous Transport Module) |
| Raw Bitrate of Lowest Level | OC-1 = 51.84 | STM-1 = 155.52 |
—Preceding unsigned comment added by 199.0.156.254 (talk) 22:08, 2 May 2007
- What most people care about is the same in both SONET and SDH. Why not retitle this article "SONET and SDH" (or "SONET/SDH") with both SONET and SDH redirecting there, and then discuss their differences explicitly in sections, which could then be split out into seperate articles? jhawkinson 00:25, 4 May 2007 (UTC)
I vote for a split too. Technologies are getting advanced at a pace faster than blink of an eye.. Though initially it was not wrong to integrate what we have, but looking at the developements & at tha pace at which they are changing, I think its now high time that a step should be taken.. We need to be more focussed on them individually so as not to defeat the purpose of this information's existence.. If we split, we'll be able to focus more clearly on the concepts.. I agree there may be some duplication, but I dont see sumthing wrong in it.. But whats more important is the missing information, that we'll be able to focus upon, when we'll opt for split..
The article needs to be split because SONET and SDH are two different technologies and two different implementations. While they may be close to the same idea, they are NOT the same thing. Not to say they shouldn't be on the same page, but then the article needs to be renamed appropriately. --Chris.Seedyk 11:43, 24 July 2007 (UTC)
Strongly oppose split. In this world we are dealing with explosion of standards serving the same goals. We have at least those two that closely overlap and that's great! Keep them together, do not make three (nor two) redundant articles. Three articles means three times more contributions is needed, plus they will never be in-sync again.
Article grows, if the text will approach 32 kB, some sections can be split. Those split-off articles should be either SONET-dedicated or SDH-dedicated. However, the main article should definately describe both technologies, clearly stating minor differences. That is what an inexperienced user coming here intuitively expects. If needed, rename to SONET and SDH per jhawkinson. --Kubanczyk 14:14, 16 September 2007 (UTC)
I vote for split! Or at least make a section explaining the differences, and try to make that at the beggining. According to my textbook, Tanenbaum's Computer Network 3rd ed, SONET is the result of a join work between BELLCORE (the research division of the U.S. RBOCS) and CCITT. Strangely, 2 results came out of this cooperation: a standard called SONET and a series of recommendations (maybe non binding) published by CCITT (G.707 to G.709). They have some differences and those should be mentioned.--Dhcpy 20:56, 5 November 2007 (UTC)
I vote for split. It is stupid that if you want to know about _exactly_ one technology, you have to digg your way through the whole article, just to find what is important to you. And there is a way more information about SONET here, than SDH, which is bad. --83.7.68.115 (talk) 20:43, 21 April 2008 (UTC)
Don't ever split The SDH standard is a slight superset of the SONET standard. There's nothing in SONET that isn't in SDH, and the two standards interwork perfectly well- you can plug STS-48 into STM-16 equipment, and provided the optics are compatible (and a few other minor things), it just works. They're really the same thing with a few minor terminology differences and a few bits in some overhead that tell you whether it's STS or STM.- (User) WolfKeeper (Talk) 21:16, 21 April 2008 (UTC)
I strongly vote against a split, and suggest renaming the article to SONET/SDH Two separate articles would over and over again refer to each other, since a comparison is interesting. The differences are readily presented in numerical comparison tables. In course literature and telecommunication lectures, SONET and SDH are typically presented together. (I'm a lecturer.) The notation SONET/SDH occurs as book title and chapter title, and gives 750.000 hits in Google, 693 hits at Google books and 206 hits at wikipedia. Only a small amount of advanced level readers are interested in the details that are specific to only one of the standards, and not the other one, and they can just as well see browse down in the article to the the comparison tables. Perhaps, a SDH section and a SONET section might be added that summarize what is unique to each of the standards. Mange01 (talk) 19:03, 24 April 2008 (UTC)
[edit] Putting "optical" on SDH is a misnomer
For crying out loud, the SDH hierarchy was defined for use over *copper* wire, not optical!
When SDH was defined, real use of optical fiber for communication was still the lab rats' wet dream, and Real Men used coax and microwave relays. Article needs splitting, badly. --Alvestrand 19:19, 4 February 2007 (UTC)
- When SDH was developped E4 (140 Mbit/s) was already deployed over (optical) fiber, and SDH was based on this experience. SDH has only one line signal defined for electrical transport: The STM-1 signal ES1, its companion optical is STM-1 OS1 and up STM-4, STM-16, STM-64 STM-256. The Electrical interface is used intraoffice and the optiacl interfaces are deployed interoffice. Huub v H 20:17, 12 March 2007 (UTC)
[edit] ICMP
ICMP is not part of the network layer but transport layer according to Network layer —Preceding unsigned comment added by 193.226.228.16 (talk) 16:22, 16 April 2007 (UTC)
- It is not in Network layer article, and ICMP is a network layer protocol. Slidersv (talk) 21:11, 12 April 2008 (UTC)
[edit] Naive User
If it helps at all I came to this article to do some research and was struck by the number of times that the terms SONET and SDH needed to be qualified. This does indicate to me the need for a split. If SONET and Euro SDH are subsets of a single SDH spec then it would seem logical to describe the top level and refer to subset differences as a separate section somehow; "compare and contrast" maybe. —The preceding unsigned comment was added by RobGi (talk • contribs) 12:40, 26 April 2007 (UTC).
- The problem is that two different things are being described and it needlessly confuses things. For example, the framing section seems to make it sound like SONET and SDH are completely different, when in reality, it has almost nothing to do with SONET vs. SDH. What is being described is the difference between STS-1 and STS-3c (or STM-0 vs. STM-1). Describing it in this way (in long-form text) doesn't do anyone any favors - it should be done in a completely different way - because the "framing" is identical between the two. 216.141.109.132 17:55, 26 April 2007 (UTC)
- No it is not "needless". Maybe just this section needs re-phrasing if it is confusing? Nothing is more scientific/encyclopedic than synthesis of knowledge. We need a few general articles, not a zillion articles about a separate tiny pieces of knowledge. --Kubanczyk 14:21, 16 September 2007 (UTC)
[edit] SSM
I think all modern SDH equipment has SSM. So timing loops are not so badly sounds, as it described in article.
Not sure if this is the proper format, but I would like to link and maybe add content to the transmux article. This is a persistent item of confusion when working with SONET to DS3 handoff between carriers. Here is the link: http://en.wikipedia.org/wiki/Transmux Thanks! Burningclod 21:17, 30 July 2007 (UTC)
[edit] basic units
I've never heard STS-3c called a basic unit in SONET, since it's really 3 STS-1s with concatenation indicators in the 2nd and 3rd STS1s. Also, I'd say VC4 is the SDH basic unit, equivalent to an STS-3c, not STM-1. —Preceding unsigned comment added by Mam711 (talk • contribs) 21:38, 7 September 2007 (UTC)
And herein lies a good reason to do the split. Reading the article, you think that there is actually something called STM-32, STM-2, etc, ignoring the fact that SDH is always multiplexed by four. There is of course room for concatination (like VCAT and GFP), but that is called something a bit different, like VC-4-2c (two concatinated VC-4 containters). Compared to most other telecommunication articles on WikiPedia, this one is one of the least educational. AllanEising 13:47, 22 September 2007 (UTC)
[edit] Certification
are there any exams to be certified in designing the sdh networks...... —Preceding unsigned comment added by 212.62.97.23 (talk) 17:09, 29 December 2007 (UTC)
