We see network speed terms (e.g. T1, DS0, OC-192) all over the place and we always get confused (with our modest brain sizes) over what they all mean. So we set out to try and pull all the pieces into one place and this is what we came up with:
First some basic stuff. You will see references to 64K (bits) 'channels' all over the place. This is the basic digital voice signal - called Digital Signal 0 or the infamous DS0 for short. The digital voice signal is encoded using PCM (Pulse Code Modulation) and TDM (Time Division Multiplexing). All other classic copper signal hierarchies, known as PDH - the Plesiochronous Digital Hierarchy, such as T3, are defined as multiples of DS0. Why 64K. Well... to digitize narrowband speech (voice) you take a 4 KHz spectrum (actually 3.1K - see notes below). Normal sampling techniques only give reasonable resolution if sampled at twice the frequency (Nyquist rate) - which gives 2 x 4K(ish) = 8K samples per second. Each sample is 8 bits which gives 8K x 8 = 64K bits per second.
The North American signal hierarchy was created by the old US 'Bell system' (AT&T) in the early 1960's and was the world's first digital voice system. It is based on multiples of the DS0 signal with a little bit of overhead to show its age. The fiendishly cunning Europeans who waited longer to define a digital hierarchy were able to live without the small overhead largely due to improved electronics.
The signal hierarchy defines the levels of multiplexing, that is, the first level of the hierarchy multiplexes (combines) a number of DS0s into a single digital signal (with a DSx designator) which is then placed on a carrier (with a T-x designator). The DSx defines an abstract signal or speed and the T-x defines a physical 'pipe' or format. The DSx and T-x series specs and most other telecom related specifications are standardized by the ANSI accredited Committee T1 (T1E1), now part of Alliance for Telecommunications Industry solutions - ATIS, which in turn represents, via the US State Department, the US at ITU standard sessions.
Remember: a DS0 is 64K or 64,000 bits per second.
|First Level||1.544 Mbit/s||DS1||T-1||24||In ISDN PRI = 23B (user) + 1D (signaling) channels|
|Second Level||6.312 Mbit/s||DS2||T-2||96||4 x DS1|
|Third Level||44.736 Mbit/s||DS3||T-3||672||28 x DS1|
|Intermediate Level||139.264 Mbit/s||DS4NA||?||2016||3 x DS3 Highest designed in ANSI T1.107|
|Fourth Level||274.176 Mbit/s||DS4||T-4||4032||Replaced with OCx|
|Fifth Level||400.352 Mbit/s||DS5||T-5||5760||Replaced with OCx|
BITDROPPING: Now if you have not been sleeping you will have figured out that for a T1 if you multiply 24 x DS0 (64,000) you do NOT get 1.544 Mbit/s instead you get 24 * 64,000 = 1.536 Mbit/s. The extra bits are lost between 'frames' where a frame consists of one 8 bit sample for each of the 24 channels (remember the DS0 basics). So every 192 bits (24 x 8 = 192) we add a 'frame separator' bit to give 193 bits per frame. The final arithmetic is 193 bits x 8K samples = 1.544 Mbit/s. Easy really.
If you do the same arithmetic for DS1C, T2 etc. the above will not give the right answer. In short, above T1 things get really nasty with M-Frames and M-subframes. It's mind numbing stuff and if you really need this information get hold of ANSI T1.107-2002 and lots of coffee or other mind-altering substances.
The fiendish Europeans left the US to blaze the digital voice trail, so when they came standardize things they could forget all this 'frame separator' stuff. Euro Telecom standards are defined by CEPT (a Euro Telecom 'club'). Here in all its glory is the super simple European hierarchy. Again all based on our good friend, the ever popular, 64,000 bit DS0.
|First Level||2.048 Mbit/s||E-1||32||In ISDN PRI = 30B (user) + 2D (signaling) channels|
|Second Level||8.448 Mbit/s||E-2||128||-|
|Third Level||34.368 Mbit/s||E-3||512||-|
|Fourth Level||139.264 Mbit/s||E-4||2048||-|
|Fifth Level||565.148 Mbit/s||E-5||8192||-|
While the table above shows the European carriers as E-1, E-3 etc. in similar format to the American T-1 etc. this terminology is of relatively recent vintage. The original carrier names were CEPT-1, CEPT-3 etc.
The following table summarizes a number of digital signal hierarchies currently in operation. We have used the terms J-1 etc. to define the Japanese signal designations for convenience without actually knowing if they are used in practice. Maybe you know...
|7.786 Mbit/s||120||-||-||J-2 (alt)|
The rates above T-3, E-3 etc are normally now optical (see below) and ANSI T1.107-2002 makes no reference to anything above DS4NA. Errata: We previously, and incorrectly, defined E-4 in this table to be 139.268 whereas in the European Hierarchy it was always correctly defined to be 139.264. While the two signal rates (E-4 and DS4NA) are the same the DS0 capacity is different. Apologies.
Optical transmission systems are known as SONET (Synchronous Optical NETwork) in North America and SDH (Synchronous Digital Hierarchy) in the Rest of the World. Optical Carriers are typically known by their OC-x number where x is a multiple of the OC-1 rate of 51.84 Mbps (shades of DS0 but a tad faster). While there is a common world-wide standard for optical systems there are differences but they are accommodated within the standard. North America uses an STS-x (Synchronous Transport Signal) format for frames (packets) and Europe an STM-x (Synchronous Transport Module) format because .... well its obvious really, one is from Europe and the other from North America and even if they were both exactly the same, which they are not, the terms would in any case be different because one is.... One day if we ever understand the differences we will add some more information.
|Level Two||2488.32 Mbit/s||STS-48||STM-16||OC-48|
|Level Three||9953.28 Mbit/s||STS-192||STM-64||OC-192|
|Optical Carrier||Data Rate||Payload-SONET (SPE)||User Data Rate||SONET||SDH|
|OC-1||51.84 Mbit/s||50.112 Mbit/s||49.536||STS-1||--|
|OC-3||155.52 Mbit/s||150.336 Mbit/s||148.608||STS-3||STM-1|
|OC-9||466.56 Mbit/s||451.044 Mbit/s||445.824||STS-9||STM-3|
|OC-12||622.08 Mbit/s||601.344 Mbit/s||594.824||STS-12||STM-4|
|OC-18||933.12 Mbit/s||902.088 Mbit/s||891.648||STS-18||STM-6|
|OC-24||1244.16 Mbit/s||1202.784 Mbit/s||1188.864||STS-24||STM-8|
|OC-36||1866.24 Mbit/s||1804.176 Mbit/s||1783.296||STS-36||STM-12|
|OC-48||2488.32 Mbit/s||2.4 Gbps||2377.728||STS-48||STM-16|
|OC-192||9953.28 Mbit/s||9.6 Gbps||9510.912||STS-192||STM-64|
Just when we thought it was getting simple - they go and make it more complicated. SDH/SONET defines a way or packaging capacity into Virtual Containers (VCs) which may be Higher Order Virtual Container (HVC) or Lower Order Virtual Containers (LVC). The term Tributary Unit (TU - used outside of North America) or Virtual Tributary (VT - North America) describes a method of mapping PDH (Plesiochronous Digital Hierarchy, for example, T1) carriers onto SDH/SONET.
The term BER (Bit Error Rate or Ratio both terms are widely used) defines the number of bit errors that can occur during transmission. It is expressed as a negative power, so that 10-10 indicates that there could be one bit error in 10,000,000,000 bits of transmission - which is a lot of bits in anyone's language. Typical error rates for copper and optical transmissions are in the range 10-10 to 10-14 whereas for wireless networks BER lies in the range 10-3 to 10-6.
Problems, comments, suggestions, corrections (including broken links) or something to add? Please take the time from a busy life to 'mail us' (at top of screen), the webmaster (below) or info-support at zytrax. You will have a warm inner glow for the rest of the day.
If you are happy it's OK - but your browser is giving a less than optimal experience on our site. You could, at no charge, upgrade to a W3C standards compliant browser such as Firefox