Data/FAXの転送時のエラー訂正と圧縮プロトコル

This is a table of many of the transmission protocols used for data communication over the past twenty years. About half of the protocols are for error-correction or concealment, and the other protocols provide data compression to improve effective transmission rates.

Vendor Name Protocol Type Description
Microcom MNP-2 Data Error-Correction MNP-2 is an error-correction protocol for asynchronous transmissions, with approximately 84% throughput efficiency.  (MNP stands for Microcom Networking Protocol.)
Microcom MNP-3 Data Error-Correction MNP-3 is an error-correction protocol for synchronous transmissions.  By moving bits across the transmission link without start and stop bits on each byte and by combining bytes to be transmitted into blocks, throughput efficiencies of 108% are possible compared to an uncorrected asynchronous transmission. The extra bits used to send the error-correction data is typically fewer than what were consumed by the start and stop bits.

The trade-off for all synchronous error-correction systems is the additional latency in the sending modem while it accumulates enough bytes to transfer as a block and in the receiving modem while waiting for the entire block to arrive so that the data can be checked for errors before any of the bytes in that block are sent on to the DTE.

Microcom MNP-4 Data Error-Correction MNP-4 is an error-correction protocol for synchonous transmissions.  Essentially, MNP-4 is a refinement on MNP-3, allowing for block-sizes that can vary in response to DTE data patterns.  Combined, efficiency can reach 120% compared to an uncorrected asynchronous transmission.
Microcom MNP-5 Data Compression MNP-5 is a data compression protocol that can be used in conjunction with MNP-4 for average compression ratios of 2:1.  However, this compression system does not respond well to data that is already compressed, causing such data transmissions to take longer than if MNP-5 was disabled.
Microcom MNP-10 Data Error-Correction Error-Correction protocol intended for use on Cellular Telephone systems to conceal fades of data and interruptions caused by transitioning from one radio cell to another.
CCITT V.42 Data Error-Correction V.42 is an error-correction protocol for synchronous transmisssions.  It incorporates the LAP-M and MNP2-4 error-correction systems.  LAP-M provides stronger accuracy checks than MNP2-4 and is always used, unless the remote system does not support it. (The CCITT changed its name to ITU in 1994.)
CCITT V.42bis Data Compression V.42bis is a data compression system that can be used in conjunction with the LAP-M mode of V.42.  It provides an average compression ratio of 4:1 and is able to detect uncompressible data and then transmit it uncompressed.  V.42bis uses the BTLZ compression system and is subject to patents held by Unisys.  V.42bis cannot be used with MNP error-correction. (The CCITT changed its name to ITU in 1994.)
ITU V.44
2000
Data Compression Approved in November 2000 along with V.92 and V.59, V.44 provides compression ratios of up to 6:1 and is based on the LZJH compression algorithm developed by Hughes. The older V.42bis compression standard had an average 4:1 compression ratio. Typical web-browsing transfer rates using V.44 are expected to be in excess of 300Kbit/sec, provided customer DTE rates are increased to match. (The typical IBM-PC compatible computer hardware limits serial communications to 115Kbit/sec to maintain compatibility with the IBM PC design that was established in 1981.)
CCITT Group 1
1968
FAX Overall Standard Group 1 provided limited international interoperability. Transmissions are in FSK, using 1300Hz for white and 2100Hz for black, at rates up to six minutes per page. Group 1 is obsolete and few modern devices support this FAX system.
CCITT Group 2
1976
FAX Overall Standard Group 2 provided international interoperability. Transmissions are in FSK, at rates up to three minutes per page. Group 2 uses the T.3 and T.30 protocol standards for operation.
CCITT Group 3
1980
FAX Overall Standard Group 3 builds on top of Group 2 by allowing vendor-unique features to be added and used when the sending and receiving systems agree on their use. Transmission speeds increase to 6 to 30 seconds per page with an initial handshake of 15 seconds. Group 3 used T.4 and T.30 protocol standards for operation. Group 3 is the current FAX standard.
CCITT Group 4
1984
FAX Overall Standard Group 4 is designed for FAX transmission between ISDN 64Kbps FAX systems. Group 4 uses T.6, T.62, T.70, T.72, T.411, T.412, T.414, T.415, T.416, T.417, T.503, T.521, and T.563 protocols.
CCITT T.2
1968
FAX Call Control T.2 is an early standard for FAX transmissions, which is now obsolete. Introduced in 1968, it has been superceded by T.3.
CCITT T.3
1976
FAX Compression T.3 is a standard for the actual transmission of the FAX image. The size, image quality, and compression of the data are handled as part of this standard. Introduced in 1976, it has been superceded by T.4.
CCITT T.4
1980
FAX Compression T.4 is the standard for the actual transmission of the FAX image. The size, image quality, and compression of the data are handled as part of this standard.
CCITT T.30
1980
FAX Call Control T.30 is the standard that describes the negotiation of a FAX transmission. T.30 places and terminates incoming or outgoing FAX calls and negotiates sender and receiver of each FAX. Group 2 and Group 3 FAX systems use T.30.
EIA FAX Class 1 DCE/DTE FAX Task Division When using a data/FAX modem with a DTE device, the DTE is responsible for all T.30 and T.4 operations. The DTE initiates operations via AT commands. (EIA 578)
EIA FAX Class 2 DCE/DTE FAX Task Division When using a data/FAX modem with a DTE device, the DTE is responsible for all T.4 operations while the modem handles the T.30 operations. The DTE initates operations via AT commands.
NONE FAX Class 2.0 DCE/DTE FAX Task Division Similar to Class 2, Class 2.0 was proposed as an improved replacement for plain Class 2, solving some issues that software writers had encountered when developing software for Class 2 modems.

The EIA/TIA voted-down the Class 2.0 proposal, so it has no official standard status. Still, several computer modem makers anticipated that Class 2.0 was going to be approved and they produced modems compatible with "Class 2.0" rather than "Class 2", and some FAX software packages were written expecting Class 2.0 as well, despite its failure to become an official standard.

Software FAX packages expecting Class 2.0 typically do not work properly on plain Class 2 modems, unless the software specifically states that it supports both Class 2 and 2.0.

EIA FAX Class 3 DCE/DTE FAX Task Division When using a data/FAX modem with a DTE device, the DCE is responsible for all T.4 and T.30 operations. The DTE initates operations via AT commands.