U interface
The U interface is a Basic Rate Interface (BRI) in an Integrated Services Digital Network (ISDN) environment. It is characterized by the use of a 2-wire-loop transmission system that
- conveys information between the 4-wire user-to-network interface, i.e., the S/T reference point in the network termination type 1 (NT1), and the line termination (LT) in the local exchange,
- is located in the servicing central office, and
- is not as distance sensitive as a service using a T interface.[1]
In America, the NT1 is customer premises equipment (CPE) which is purchased and maintained by the user, which makes the U interface a user-to-network interface.[2] The American variant is specified by the American National Standards Institute (ANSI) in T1.601.[2] In Europe, the NT1 belongs to the network operator, so the user doesn't have direct access to the U interface.[2] The European variant is specified by the European Telecommunications Standards Institute (ETSI) in recommendation ETR 080.[2] The ITU-T has issued recommendations G.960 and G.961 with world-wide scope, encompassing both the European and American variants of the U interface.[2]
Logical interface
Like all all other ISDN basic rate interfaces, the U interface carries two B (bearer) channels at 64 kbit/s and one D (data) channel at 16 kbit/s for a combined bitrate of 144 kbit/s (2B+D).
Duplex transmission
While in a four-wire interface such as the ISDN S and T-interfaces one wire pair is available for each direction of transmission, a two-wire interface needs to implement both directions on a single wire pair. To that end, ITU-T recommendation G.961 specifies two duplex transmission technologies for the ISDN U interface, either of which shall be used: Echo cancellation (ECH) and Time Compression Multiplex (TCM).[3]
Echo cancellation (ECH)
When a transmitter applies a signal to the wire-pair, parts of the signal will be reflected as a result of imperfect balance of the hybrid and because of impedance discontinuities on the line.[3] These reflections return to the transmitter as an echo and are indistinguishable from a signal transmitted at the far end. In the echo cancellation (ECH) scheme, the transmitter locally simulates the echo it expects to receive, and subtracts it from the received signal.[3]
Time Compression Multiplex (TCM)
The Time Compression Multiplex (TCM) duplex method, also referred to as "burst mode", solves the echo problem indirectly.[3] The line is operated at a rate at least twice the signal rate and both ends of the line take turns transmitting, in a time-division duplex fashion.[3]
Line Systems
ITU-T G.961 specifies four line systems for the ISDN U interface: MMS43, 2B1Q, TCM, and SU32.[3] All line systems except TCM use echo cancellation for duplex operation.[3] The American standard ANSI T1.601 specifies the 2B1Q line system, the European ETSI TR 080 recommendation specifies 2B1Q and MMS43.[2]
MMMS43 (4B3T)
4B3T is a standard used in Europe and elsewhere in the world. 4B3T is a "block code" that uses Return-to-Zero states on the line. 4B3T converts each group of 4 data bits into 3 "ternary" line signal states (3 symbols). The symbol rate is 3/4ths of the data rate (120 kbaud).
4B3T is defined in ETR 080, Annex B and other national standards, like Germany's 1TR220. 4B3T can be transmitted reliably at up to 4.2 km over 0.4 mm cable or up to 8.2 km over 0.6 mm cable. An internal termination impedance of 150ohms is presented to the line at each end of the U-interface. Echo cancellation techniques allow full-duplex operation on the line.
In 4B3T coding, there are three states presented to line: a positive pulse (+), a negative pulse (-), or a zero-state (no pulse: 0). An analogy here is that operation is similar to B8ZS or HDB3 in T1/E1 systems, except that there is an actual gain in the information rate by coding 16 possible binary states to one of 27 ternary states.
There are many ways to perform this conversion, but BRI standards define a code known as MMS43 (Modified Monitoring State 43).
One requirement for line transmission is that there should be no DC build-up on the line, so there are some code substitutions based upon the transmission of the previous bits.
Appendix I of G.961 specifies a line system based on echo cancellation and a Modified Monitoring State Code mapping 4 bits into 3 ternary symbols (MMS43), which is also referred to as 4B3T (four binary, three ternary).[3] The ternary alphabet consists of positive voltage (+), zero voltage (0), and negative voltage (-).[3] A 1 ms frame carrying 144 bits of 2B+D data is mapped to 108 ternary symbols.[3] These symbols from the 2B+D channels are scrambled, with different codes for the two transmission directions, in order reduce correlation between transmitted and received signal.[3] To this frame, an 11-symbol preamble and a symbol from the CL channel are added, yielding a frame size of 120 ternary symbols and a symbol rate of 120 kBaud.[3] The CL channel is used to request activation or deactivation of a loopback in either the NT1 or a line regenerator.[3] The redundancy introduced by mapping 24=16 into 33=27 symbols is used to generate a zero DC-bias signal.[3] Of the 16 binary information words, some are always mapped to a DC-component free (ternary) code word, while others can be mapped to either one of two code words, one with a positive and the other with a negative DC-component.[3] In the latter case, the transmitter chooses whether to send the code-word with negative or positive DC-component based on the accumulated DC-offset.[3]
2B1Q
2B1Q coding is the standard used in North America and in Italy. 2B1Q means that two bits are combined to form a single Quaternary line state (symbol). 2B1Q combines two bits at a time to be represented by one of four signal levels on the line. The signal rate, therefore, is 80 kilobaud and the line rate 160 kbit/s. It operates with a maximum frequency range (bandwidth) of 40 kHz.
2B1Q coding is defined in ANSI T1.601 and ETR 080, Annex A. It can operate at distances up to about 18,000 feet (5.5 km) with loss up to 42dB. An internal termination impedance of 135 ohms is presented to the line at each end of the U-interface. Echo cancellation techniques allow full-duplex operation on the line.
Appendix II of G.961 specifies a line system based on echo cancellation and a line code mapping two binary into one quaternary symbol (2B1Q).[3] A 1.5 ms frame carrying 216 scrambled bits of 2B+D data is mapped to 108 quaternary symbols.[3] To this frame, a 9-symbol preamble and 3 symbols from the CL channel are added, yielding a frame size of 120 quaternary symbols and a symbol rate of 80 kBaud.[3] The CL channel is used for communication between LT and NT1, a 12-bit cyclic redundancy check (CRC), and various other physical layer functions.[3] The CRC covers one 12 ms multiframe (8×1.5 ms frames).[3]
TCM / AMI
Appendix III of G.961 specifies a line system based on the Time Compression Multiplex (TCM) duplex method and an alternate mark inversion (AMI) line code.[3] The AMI line code maps one input bit to one ternary symbol.[3] Like with MMS43, the ternary symbol can either be a positive (+), zero (0), or negative (-) voltage.[3] A 0 bit is represented by a zero voltage, while a 1 bit is alternatingly represented by a positive and a negative voltage, resulting in a DC-bias free signal.[3] In a 2.5 ms interval, each side can send a 1.178 ms frame representing 360 bits of 2B+D data.[3] To the 2B+D data, an 8-bit preamble, 8 bits from the CL channel, as well as a parity bit are added, yielding a frame size of 377 bits and a baud rate of 320 kBaud.[3] The CL channel is used for operations and maintenance, as well transmitting a 12-bit CRC covering 4 frames.[3]
SU32
Appendix IV of G.961 specifies a line system based on echo cancellation and a substitutional 3B2T (SU32) line code, which maps three bits into 2 ternary symbols.[3] As with MMS43 and AMI, the ternary symbol can either be a positive (+), zero (0), or negative (-) voltage.[3] The mapping from 23=8 to 32=9 symbols leaves one unused symbol.[3] When two subsequent input (binary) information words are identical, the (ternary) code word is substituted by the unused code word.[3] A 0.75 ms frame carrying 108 bits of 2B+D data is mapped to 72 ternary symbols.[3] To this frame, a 6-symbol preamble, one CRC symbol, and 2 symbols from the CL channel are added, yielding a frame size of 81 ternary symbols and a symbol rate of 108 kBaud.[3] The CL channel is used for supervisory and maintenance functions between the LT and NT1.[3] The 15-bit CRC covers 16 frames.[3]
References
- ^
This article incorporates public domain material from Federal Standard 1037C. General Services Administration. Archived from the original on 2022-01-22.
- ^ a b c d e f Burd, Nick (1997). "Section 4.5: U interface standards". ISDN Subscriber Loop. London: Chapman & Hall.
{{cite book}}: External link in|chapterurl=|chapterurl=ignored (|chapter-url=suggested) (help) - ^ a b c d e f g h i j k l m n o p q r s t u v w x y z aa ab ac ad ae af ag ah ai aj "G.961: Digital transmission system on metallic local lines for ISDN basic rate access". ITU-T. March 1993. Retrieved 2014-01-06.
See also
Further reading
- Lechleider, J.W. (1989). "Line codes for digital subscriber lines (ISDN basic access)" (fee required). IEEE Communications Magazine. 27 (9): 25–32. doi:10.1109/35.35509. ISSN 0163-6804.