United States Patent
`US 6,490,295 B1
`(10) Patent No.:
`(12)
`’
`Cooklev et al.
`45) Date of Patent:
`Dec. 3, 2002
`
`
`US006490295B1
`
`(54) HIGH-SPEED MODEM OPERATING OVER
`TWO OR MORE TELEPHONE LINES
`
`5,801,695 A
`9/1998 Townshend
`5,815,505 A
`9/1998 Mills
`5,859,872 A *
`1/1999 Townshend ............... 375/242
`eesa0 a : S500 piatsson eseeseseseseseseees sosse
`Inventors: Todor Cooklev, Salt Lake City, UT
`
`ui et ale eee
`5198,
`.
`‘
`.
`
`5/2001 Liu et ale cesssssecsssssseees 370/469
`6,233,250 Bl *
`eS): Kevin Smart, Bountiful, UT
`5/2001 Kato oo... 375/240.26
`6,240,137 Bl *
`(
`)
`* cited by examiner
`(73) Assignee: 3Com Corporation, Santa Clara, CA
`US(US)
`Primary Examiner—Kwang Bin Yao
`Assistant Examiner—Hanh Nguyen
`Subject to any disclaimer, the term ofthis
`(74) Attorney, Agent, or Firm—Workman, Nydegger &
`patent is extended or adjusted under 35
`Seeley
`US.C. 154(b) by 0 days.
`57
`e7
`(21) Appl. No.: 09/304,392
`A communication system configured to transceive a signal
`y
`g
`g
`.
`along multiple communication media of the communication
`(22)
`Filed:
`May4, 1999
`System thereby increasing the rate at which the signal is
`(51) Unt. C1? eects H04J 3/04; HO4J 3/16;
`transceived. The communication system comprising a
`HO4L 5/16
`source configured to transceive a signal. A communication
`370/465; 370/536; 375/222
`(52) US. Cl
`apparatus configured for decomposing the signal
`into a
`(58) Field of Search— ,
`375p42. 285
`plurality of manipulated signals. The numberofthe plurality
`375365,254, 222, 368“ 266. 356, 1.
`of manipulated signals being determined by the number of
`?
`?
`.
`,
`,
`,
`,
`370/535-537, 538, S704 on the multiple communication media in communication with
`ae the source and the maximum transceival rate of each com-
`References Cited
`munication media. In communication with the communica-
`tion apparatus is a reconstructing apparatus that is config-
`U.S. PATENT DOCUMENTS
`ured for reconstructing the plurality of manipulated signals
`into the signal, the signal being capable of being transceived
`by a host.
`
`(75)
`
`(*) Notice:
`
`(56)
`
`ABSTRACT
`
`5/1993 Sridharetal.
`5,214,637 A
`5/1996 Walsh et al.
`5,515,398 A
`6/1997 Saitoh
`5,636,037 A
`5,682,404 A * 10/1997 Miller wo.ee 375/222
`
`39 Claims, 11 Drawing Sheets
`
`
`
`
`
`
`DIGITAL
`
`
` |_|
`TELEPHONE
`ENCODER
` SOURCE
`
`
`NETWORK
`
`
`
`1
`
`Exhibit 1069
`Samsung v. Smart Mobile
`IPR2022-01249
`
`Exhibit 1069
`Samsung v. Smart Mobile
`IPR2022-01249
`
`1
`
`

`

`U.S. Patent
`
`Dec. 3, 2002
`
`Sheet 1 of 11
`
`US 6,490,295 B1
`
`HOLYSANODNOSI
`
`¢Old YOLYSANODNQSI
`
`
`YOLU3ANODNASITNOISyadang
`
`49018°ONISS3004d
`
`JOUNOS
`
`WLI9I0
`
`INOHd3134
`
`YYOMLIN
`
`ONISOdW093030VSUAINI
`
`Y3Q00NI|.|
`
`2
`
`
`
`
`

`

`U.S. Patent
`
`Dec. 3, 2002
`
`Sheet 2 of 11
`
`
`
`US 6,490,295 B1
`
`§|
`
`9000'0r
`
`
`
`3
`
`

`

`
`
`a:9078v8=:OQNILONYISNODIYORVWILS3“¥9010sM3eS
`
`
`U.S. Patent
`
`Ohh
`
`
`
`mere4314
`
`0698
`
`
`OLY3NITISYIANI
`
`+08
`
`
`
`
`
`
`
`US 6,490,295 B1
`
`¢Old
`
`
`
`Uz0}UpOl90)W196
`
`40010
`
`JLVWILSS
`
` ===me|-Yadny«OLY3NI1ISUIANIyo0191p3106ugg199upg2gDye
`
`
`
`4
`
`
`
`
`

`

`U.S. Patent
`
`Dec. 3, 2002
`
`Sheet 4 of 11
`
`US 6,490,295 B1
`
`99
`
`GILONYLSNODIY
`
`QILONYLSNODSY
`
`
`
`NOILONYLSNODSYONISOdW00I0
`
`40018
`
` 40019
`
`5
`
`
`
`

`

`U.S. Patent
`
`Dec. 3, 2002
`
`Sheet 5 of 11
`
`US 6,490,295 B1
`
`Oe}
`
`OVI
`
`8f
`
`OVI
`
`
`
`(pe'A)06°A
`
`Uve'n)06)
`
`§Oldi
`
`re
`
`ne
`
`
`
`0¢|
`
`SIqoPA
`k
`ONILONYLSNODIY
`
`Sf493000|.
`
`(ve'n)o6'n|
` fee
`
`
`
` lve'A)06'H
`
`oft
`
`0¢|
`
`0c|
`
`INISOdN00I0
`
`GGk
`
`6
`
`
`
`
`
`

`

`U.S. Patent
`
`Dec. 3, 2002
`
`Sheet 6 of 11
`
`US 6,490,295 B1
`
`Nygy
`
`CS
`
`Noy
`
`6“Old
`
`Ihmy=|NIWdIgV4d“
`01ha"eg,8
`D2S}odwe)
`comfnofig|.
`
`
`
`Npgy(3910A‘WOOH)N1O¥LNODAV13¥|y|
`
`
`(JO10A‘YOOH)|TONLNODAV134
`
`
`89)
`
`‘091
`
`TL
`
`pl
`
`ah
`
`LSOH
`
`Y3LAdWOd
`
`7
`
`

`

`U.S. Patent
`
`Dec. 3, 2002
`
`Sheet 7 of 11
`
`US 6,490,295 B1
`
`ONIAIIOIY
`
`XV
`
`yagiuosans|,JOTWNY
`
`TWLI9IG
`
`INOHd314L
`
`MYOMLIN
`
`9|
`
`0bOld
`
`gLI
`
`NaS]
`
`UIGlWOSENs
`
`ONIGNIS
`
`Q33d$-HOIH
`
`NOILNGIYLSI0
`
`WALSAS
`
`bhOld
`
`XW
`
`f
`
`8
`
`
`
`
`
`
`

`

`U.S. Patent
`
`Dec. 3, 2002
`
`Y3AYIS Y3IANIS
`1N3I19INOHd3131$$300¥Lay0vd
`
`
`
`
`
`
`YYOMLINAYMALY9YYOMLINJOVIYSINI
`
`00¢
`
`WIGIWILINN
`
`Sheet 8 of 11
`
`US 6,490,295 B1
`
`INIIT9
`
`INOHdITIL
`
`WYOMLIN
`
`Y5AU3S
`
`JOVIYILN
`
`Ge}Old
`
`
`
`Web‘Old
`
`002
`
`VIGSWILINN
`
`YIAY3S
`
`9
`
`
`
`
`
`

`

`U.S. Patent
`
`Dec.3, 2002
`
`Sheet 9 of 11
`
`US 6,490,295 B1
`
`
`
`ew
`
`IN ONE CHANNEL
`
`- 220
`
`YES
`
`10
`
`

`

`
`
`U.S. Patent
`
`270 SIGNALENCRYPTION
`
`Dec.3, 2002
`
`Sheet 10 of 11
`
`US 6,490,295 B1
`
`
`FIG.14
`
`
`
`
`11
`
`

`

`U.S. Patent
`
`Dec.3, 2002
`
`Sheet 11 of 11
`
`US 6,490,295 B1
`
`FIG.16
`
`12
`
`

`

`US 6,490,295 B1
`
`1
`HIGH-SPEED MODEM OPERATING OVER
`TWO OR MORE TELEPHONE LINES
`
`BACKGROUND OF THE INVENTION
`
`1. The Field of the Invention
`
`The present invention generally relates to communication
`devices used within a communication system or network,
`and, more particularly, to a device for transceiving signals
`over multiple telephone lines or similar transceiving lines.
`2. Present State of the Art
`
`Throughout the ages man has initiated and developed
`numerous methods to communicate information. Commu-
`nication in one form or another is used continuously,
`whetherit be face to face conversation involving both body
`and verbal communication or through pictures, music,orart.
`With the advancesin technology, however, individuals wish
`to spend more time communicating to discuss business,
`entertainment, and other daily events, but wish
`communication, in all its forms, to be more easily accom-
`plished.
`The modern society in almost every respect is crucially
`dependent on its ability to communicate signals or data,
`whetherin digital or analog form, from one pointto another.
`With the advances in technology the Internet has become
`ubiquitous for business and electronic commerce, education,
`entertainment, etc. As such,
`individuals, companies and
`other entities demand faster and faster communication
`speeds to manufacture, distribute and sell their products and
`services. In manysituations, the speed of signal transmission
`or receiving (“transceiving”) directly impacts the quality of
`the services provided via the Internet, for example real-time
`video conferencing requires a minimum transceiving speed
`to be feasible.
`The communication channels over which data is trans-
`ceived is almost always the widespread public switched
`telephone network (PSTN). The core of the PSTN in the
`United States and other industrialized countries is com-
`pletely digital, while the connection to the digital backbone
`is traditionally analog. A digital connection to the PSTN is
`possible through a service such as the Integrated Services
`Digital Network (ISDN). The ISDN provides 2 digital
`channels that are each capable of transceiving signals or data
`at a rate of 64,000 bits per second (“b/s”) and a control
`channel that can transceive signals or data at 16,000 b/s.
`To use the ISDN,a user’s central office (“CO”), such as
`a local
`telephone company’s switching office, must be
`upgraded to provide lines and other equipment capable of
`transceiving signals. Therefore, the user must replace the
`analog on-premises equipment with digital equivalents,
`while the individual lines at the CO must be modified to
`
`carry digital data such as fiber optic cable. The installation
`costs and monthly charges for connectivity through an ISDN
`are significant, such that most users do not have a digital
`connection to the PSTN. Furthermore, ISDN digital con-
`nections are infrequently offered in rural and sparsely popu-
`lated areas since it is difficult for telephone companies to
`recoup their investment in equipment and installation. In
`light of this, most users continue to have an analog connec-
`tion to their CO.
`
`The analog portion of the PSTN was designed to carry
`voice as inexpensively as possible. In particular, most analog
`connections to a CO are bandlimited and carry signals with
`a bandwidth ranging from about 200 Hz to about 3200 Hz
`in the United States and from about 300 Hz to about 3400
`
`10
`
`15
`
`20
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`2
`Hz in some other countries. The band ranges were chosen
`decades ago, as the narrowest possible band which could
`contain specific important characteristics of the human
`voice. Any signals outside these ranges are typically sharply
`attenuated.
`
`To transceive data over an analog connection to the CO
`requires a modem. A modem performs two tasks:
`modulation, which converts the digital signal into an analog
`signal in the upstream direction, and demodulation, which
`converts the analog signal into a digital signal in the down-
`stream direction. Most modemstoday convert a digital data
`stream into an analog signal within the bandwidths refer-
`enced in regard to the PSTN.
`In recent years substantial progress has been achieved in
`modem design. While earlier modems could operate only at
`rates of 2400 b/s, modem speeds haveincreased up to 33,600
`b/s. See International Telecommunications Union, Telecom-
`munication Standardization SectorITU-T) Recommenda-
`tion V.34, Geneva, Switzerland (1994) which is hereby
`incorporated as a reference.
`Unfortunately rates of up to 33,600 b/sare insufficient for
`many of the newer applications envisioned with the advent
`of the Internet, such as video conferencing. While text
`transmission is fast, facsimile and especially still
`image
`transmission is slow. Furthermore, even with current sophis-
`ticated audio compression algorithms only low-quality
`video and audio is possible.
`There are fundamental limitations that reduce the quality
`of data transmission in addition to lowering the maximum
`achievable data rate over the PSTN. The capacity of a
`communication channel on the PSTN, as discussed in C.
`Shannon, “A Mathematical Theory of Communication,”
`Bell System Technical Journal volume 27, pp. 379-423 and
`pp. 623-656, 1948, which is incorporated herein by
`reference, is given by
`
`S
`C= w(t + Loe
`
`Wd)
`
`where C is the maximum achievable data rate in b/s, W is the
`bandwidth of the channel in Hertz, and S/N is the signal to
`noise ratio. For most of the PSTN of the United States S/N
`at present is below 2000 (approximately 30 dB). If we
`substitute these numbers into the above equation we can
`easily find out that C~3000x12=36,000 b/s. Regardless of
`the sophistication of current signal processing algorithms or
`the speed of current processors, the maximum achievable
`data rate remains the same for a single PSTNline. It is clear
`that current modem standards have achieved a rate which is
`
`very close to the maximum possible. Thus the speed of
`modemsis limited not by available technology, but by the
`limited bandwidth of the telephone system.
`The bandwidth limitation becomes more acute when
`
`combined with the changing usage of the PSTN.In the past
`mostof the traffic over the PSTN was voice, with very little
`percentageofthe total traffic being data. At the beginning of
`the next century, however, the ratio of voice to datatraffic is
`expected to become reversed; with more data traffic than
`voice traffic.
`A significant portion of the increase datatraffic is caused
`by the availability of the Internet access. Most users today
`connect to the Internet through their Internet Service Pro-
`vider (“ISP”). ISPs usually have a high-bandwidth direct
`digital connection to the PSTN. Normally high-rate of
`communication is necessary in one direction only, from the
`ISP to the user (the downstream direction). This arrange-
`13
`
`13
`
`

`

`US 6,490,295 B1
`
`3
`ment allows speeds of up to 56,000 b/s in the downstream
`direction. Currently modems capable of receiving data at
`speeds up to 56,000 b/s are available from several modem
`vendors, such as the 3Com Corporation, Santa Clara, Calif.
`Many 56,000 b/s modems are capable of transceiving
`signals at various rates. Furthermore, the ITU-T V.90 stan-
`dard for modemsthat can operate at rates up to 56,000 b/s
`actually envisions several possible modem data rates that
`vary based on the telephone line conditions, such as the
`effects of signal-to-noise ratio. Thus, unlike previous
`modem standards ITU-TV.90 does not specify a single data
`rate in the downstream direction. The allowed rates in the
`
`downstream direction range from about 28,000 to 56,000 in
`1,333 b/s increments.
`In normal communication sessions, two modemsthat are
`in communication will evaluate the telephone line condi-
`tions according to a line probing technique. Such line
`probing techniques are discussed for example in US. Pat.
`No. 5,515,398 entitled “Modem line probing signal
`techniques,” issued to Walsh et al. which is assigned to the
`assignee of the present
`invention. The superior the line
`conditions, the higher the data rate at which the two modems
`will choose to operate.
`Line characteristics of the PSTN lines can change with
`time, however, and may be varied through influence of
`electric and magnetic fields that are in close proximity to the
`PSTN lines. For example, power lines can induce a 60 Hz
`hum onto an analog telephone line. Furthermore, unwanted
`signals from adjacent telephone lines can induce unwanted
`voltages, called crosswalk. The influence of hum and cross-
`walk decrease the signal-to-noise ratio (S/N) and reduce the
`maximum data rate that can be achieved over the telephone
`line. Each time the line characteristics deteriorate the
`
`modems in communication negotiate to select a lowerrate at
`which to communicate reliably. If the line characteristics
`improve the modems will select a higher rate. Therefore,
`over a single telephoneline it is possible to connect some-
`times at 49,333 b/s, while at another timeis only possible to
`achieve 45,333 b/s.
`Unlike end-to-end digital connections used by an ISP, the
`analog telephone lines making up the PSTN are widely
`available and relatively much more expensive. An increas-
`ing number of businesses and people have two and more
`telephone lines to allow them to perform multiple tasks
`concurrently. Indeed many user add a second telephoneline
`just for occasional use, for example, for facsimile services.
`The precious bandwidth that is offered by the second tele-
`phoneline is wasted mostof time. The productivity of many
`users would be increased if they could use the second
`telephone line to achieve higher-speed access to an ISP,
`other modems, or the like.
`Unfortunately, there are numerous problemswith forming
`a modem that is capable of communicating signals over two
`or more telephone lines. A significant problem is the vari-
`ability of telephone line conditions and characteristics.
`When two modemsoperate over two or more telephone
`lines, if the line conditions on all lines are identical, then
`clearly the aggregate data rate is the sum of the data rates
`that are achieved over the individual phonelines.
`However,the line conditions, will not always be identical.
`As a matter of fact, they are very likely to be different. For
`example, it is clear that the amount of noise induced onto
`two telephone lines will be different. This noise typically
`comes from neighboring telephonelines, powerlines,etc, as
`stated above. According to the Shannon’s limit the maxi-
`mum data rates that can be achieved over the two lines will
`be different, as the maximum achievable data rate over each
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`4
`line will directly depend on the signal-to-noise ratio (SIN)
`over that line. If the rates are different, however, it is not
`obvious how can we achieve an aggregate data stream equal
`to the sum of the data rates achieved over the individual
`lines. One obviouspossibility is to select the lowest data rate
`that the telephone lines can work at and use this rate on all
`telephone lines. For example, if it is found that one of the
`lines supports 49,333 b/s and the other supports 45,333 b/s,
`assuming that we have two telephone lines, we mightselect
`to operate at 45,333 b/s over the two telephone lines,
`achieving an aggregate data stream of 90,666 b/s. Clearly
`this is not the optimum solution. It
`is very desirable to
`achievea data rate of 94,666 b/s, whichis the sum of the two
`data rates in this example.
`Thus, the presentstate of the art dictates that if the two or
`more data rates achievable overthe different telephonelines
`are not the same negotiation is performed, followed by a
`fallback on all
`lines onto a data rate equal
`to the rate
`achieved by the slowest line. It is clear that the aggregate
`data rate would be only the data rate achieved on the slowest
`line times the number of channels, but not the sum of the
`maximum data rates on every line. Furthermore the process
`of negotiating different rates on the lines is slow. It is also
`very inefficient
`to require the modems to negotiate new
`communication rates each time the minimum data rate
`changes.
`These disadvantages can have a significant and negative
`effect on a modem’s performance and might make it less
`commercially viable for sale.
`SUMMARY AND OBJECTS OF THE
`INVENTION
`
`invention to provide a
`is an object of the present
`It
`communication devicethat is capable of transceiving signals
`along two or more communication lines.
`It is another object of the present invention to provide a
`communication device that is capable of transceiving data at
`a rate corresponding to the maximum aggregate communi-
`cation rate of two or more communication lines.
`
`Another object of the present invention is to provide a
`communication device that achieves an aggregate data rate
`that is the sum of the maximum data rates achievable on the
`individual telephone lines.
`It is another object of the present invention to provide a
`communication device that is cheap and inexpensive.
`Still yet another object of the present invention is to
`provide a communication device that is capable of secure
`communication over an unsecured communication network.
`
`Yet another object of the present invention is to provide a
`communication device that is capable of multiplexing sig-
`nals along multiple telephone lines at a communication rate
`substantially similar to the aggregate communicationrate of
`the multiple telephone lines.
`It is another object of the present invention to provide a
`method of manipulating signals to be communicated along
`multiple telephonelines into a form that allow maximization
`of the communication rate of the telephone lines.
`It is another object of the present invention to provide a
`method and system that achieves signal communication
`rates that are substantially equal to the aggregate of the
`telephone lines used.
`Still yet another object of the present invention is to
`provide a modem device that is capable of secure commu-
`nication over an unsecured communication channel.
`
`Additional objects and advantagesof the invention will be
`set forth in the description which follows,and in part will be
`14
`
`14
`
`

`

`US 6,490,295 B1
`
`5
`obvious from the description, or may be learned by the
`practice of the invention. The objects and advantages of the
`invention may be realized and obtained by means of the
`instruments and combinationsparticularly pointed out in the
`appended claims. To achieve the foregoing objects, and in
`accordance with the invention as embodied and broadly
`described herein, a communication system configured to
`transceive a signal along multiple communication media of
`the communication system to thereby increase the rate at
`which the signal is transceived is disclosed. The communi-
`cation system comprising a source configured to transceive
`a signal. Acommunication apparatus configured for decom-
`posing the signal into a plurality of manipulated signals. The
`numberof the plurality of manipulated signals being deter-
`mined by the numberof the multiple communication media
`in communication with the source and the maximum trans-
`ceival rate of each communication media. In communication
`
`with the communication apparatus is a reconstructing appa-
`ratus that is configured for reconstructing the plurality of
`manipulated signals into the signal, the signal being capable
`of being transceived by a host.
`In general, the present invention allows a users to reap
`maximum benefits of a “bandwidth-on-demand” policy
`where users can access any type of digital signal (high-
`quality audio, video, etc.) at speeds which are maximum for
`their available telephone lines. By allowing increase signal
`transceival rates, the present invention makes several new
`applications possible, such as for example videophone,
`teleconferencing, high-quality video and audio, etc.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`In order that the manner in which the above-recited and
`
`other advantages and objects of the invention are obtained,
`a more particular description of the invention briefly
`described above will be rendered by reference to specific
`embodiments thereof which are illustrated in the appended
`drawings. Understanding that these drawings depict only
`typical embodiments of the invention and are not therefore
`to be consideredto be limiting of its scope, the invention will
`be described and explained with additional specificity and
`detail through the use of the accompanying drawings in
`which:
`
`FIG. 1 is a block diagram representing one embodiment
`of a communication system of the present invention.
`FIG. 2 is a block diagram representing an encoderof the
`communication system in FIG. 1.
`FIG. 3 is a block diagram representing a decomposing
`block and a reconstructing block of the communication
`system in FIG. 1.
`FIG. 4 is a low level block diagram representing a
`decomposing block of the communication system in FIG. 1.
`FIG. 5 is a block diagram representing a decoder of the
`communication system in FIG. 1.
`FIG. 6 is a block diagram representing the operation of a
`decomposing block and a reconstructing block of the com-
`munication system in FIG. 1.
`FIG. 7 is a low level block diagram representing a decoder
`of the communication system in FIG. 1.
`FIG. 8 is a block diagram representing a high speed two
`line modem utilizing the principal of the present invention.
`FIG. 9 is a low level block diagram representing the
`components necessary to perform the operation of the high
`speed two line modem ofFIG. 8.
`FIG. 10 is a block diagram representing a digital tele-
`phony relay in accordance with the present invention.
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`6
`FIG. 11 is a block diagram representing a high speed
`facsimile system in accordance with the present invention.
`FIG. 12 is a block diagram representing a multimedia
`distribution server in accordance with the present invention.
`FIG. 13 is a flow diagram representing the operation of a
`communication protocol utilizing the principals of the
`present invention.
`FIG. 14 is block diagram representing a signal encryptor
`in accordance with the present invention.
`FIG. 15 is a block diagram of a analyzing stage of the
`signal encryptor of FIG. 14.
`DETAILED DESCRIPTION OF THE
`PREFERRED EMBODIMENTS
`
`The present invention is directed to systems and devices
`that are capable of manipulating multiple signals such that
`they may be transceived along multiple communication
`channels at a rate substantially equal to the sum of the
`maximum communication rate achievable on each indi-
`
`the
`vidual communication channel. More specifically,
`present invention is capable of decomposing multiple input
`signals to be transceived within a telephone network at a
`data communication rate equal to the sum of the maximum
`communication rate of each individual telephone line con-
`nected to the communication device.
`
`Discussion herein will be made in relation to an asym-
`metric communication system 10 whenthe beneficial effects
`of the principals of the present invention are best shown. It
`can be appreciated that the beneficial effects may also be
`seen as an asymmetric communication system. Referring
`now to FIG. 1, a communication system 10 of the present
`invention is depicted. Communication system 10,
`in this
`particular configuration, utilizes an encoder 12 and a
`decoder 14 in communication by way of a telephone net-
`work 16. A plurality of data streams from a source (not
`shown) are applied to encoder 12. This particular configu-
`ration source 18 is an internet service provider (ISP), how-
`ever various other sources are applicable. Encoder 12 is
`formed to convert the digital data streams from ISP 18 into
`a numberof data streams that correspond to the number of
`communication channels or telephone lines whichare avail-
`able to encoder 12. Furthermore, encoder 12 manipulates the
`digital signal into a format suitable for transceival through
`telephone network 16. The connection between encoder 12
`and the telephone network 16 is digital thereby allowing a
`fast communication rate. Analog connectors may be used
`when communication system 10 is asymmetric.
`The manipulated digital signal
`is output
`to telephone
`network 16 and more specifically to the digital backbone 20
`of the telephone network 16. Through a direct digital-to-
`digital connection the manipulated digital data is incident to
`appear undistorted at a client’s central office (“CCO”) 26.
`The CCO 26 includes a plurality of line interfaces 28 and
`hybrid circuitry 30 that converts the digital data stream from
`digital backbone 20 to analog signals. More specifically, line
`interfaces 28 connects the input digital signal into an analog
`signal which hybrid circuitry 30 connects the two wire
`bi-directional analog signal to a pair of one waysignals, 1.e.,
`4 wire to 2 wire connectors.
`
`The analog digital output for CCO 26 are capable of being
`transceived along the analog lines the CCO 26 to a decoder
`14. Decoder 14 uses the transceived analog signals to
`compensate for any distortion introduced by the conversion
`from digital
`to analog by line interfaces 28 and hybrid
`circuitry 30 while constructing a plurality of data streams
`that are sent
`to a user’s host (not shown).
`In general,
`15
`
`15
`
`

`

`US 6,490,295 B1
`
`7
`telephone network 16 comprises the digital backbone, the
`CCO 26, and the analog connection to the user and may be
`considered the PSTN,
`the components thereof are well
`hence known to one skilled in the art and need not be
`discussed herein. It can be appreciated that the telephone
`network 16 may incorporate various other forms as known
`by one skilled in the art such as fiber optics, copper wire,
`PSTN,orthe like.
`Referring now to FIG. 2, a functional block diagram of
`one possible realization of encoder 12 according to the
`present invention is depicted. Encoder 12 comprises as an
`interface buffer, a decomposing block 38, a DC eliminator
`40 and a ISDN convertor 42 each of which will be discussed
`
`in detail hereinafter. The plurality of data streams from
`different sources are input to an interface buffer 34. Interface
`buffer 34 receives information from a rate control block and
`detaines the characteristics of the channels until the data is
`
`to be transceived along. Interface buffer 34 converts the
`individual data bit streams into sequencesof eight-bit words
`in preparation for transceiving through the communication
`channels. The eight-bit words are sampled at any appropriate
`time depending on the raw data rate and the channel or
`telephone line characteristics of telephone network 16.
`Furthermore, interface buffer 34 implements functions such
`as flow control (not explicitly shown) to maintain a smooth
`aggregate data rate as transceived along the communication
`channels.
`
`Interface buffer 34 is necessary, because the aggregate
`data rate of transmission varies with time and can changeat
`any moment. Thus,it is necessary to maintain a pool of data
`that may be drawn uponto create an optimized data stream
`output along the communication channels.
`In communication with interface buffer 34 is a decom-
`
`posing signal processing block or decomposing block 38.
`The purpose of decomposing block 38 is to transform the
`received plurality of data streams into a second plurality of
`data streams equal to the number of communication chan-
`nels available to encoder 12. The transceiving rate of each
`data stream leaving decomposing block 38 is the maximum
`rate for each individual communication channel and the
`
`aggregate of all available communication channels cooper-
`ating with encoder 12. Since the characteristics of the
`individual communication channels vary with time,
`the
`characteristics of decomposing block 38 vary with time
`under the control of rate control block 36.
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`8
`preparation for transceival to the telephone network 16. The
`functionality of DC eliminator 40 and ISDN conversion 42
`is known to one skilled in the art. One example of the
`operation and functionality of DC eliminator 40 and ISDN
`conversion block 42 is described in U.S. Pat. No. 5,801,695
`entitled “High speed communications system for analog
`subscriber connections” issued to Townshend, which is
`herein incorporated by this reference.
`Referring again to FIG. 3A, a configuration of decom-
`posing block 38 is depicted. If the data rates on all individual
`connections to encoder 12 are equal, then it would be a
`relatively simple matter to split the aggregate data stream
`into individual data streams. For example, the samples with
`an even index can be assigned to one of the data streams and
`the samples with an odd index can be assigned to the other
`stream. A more sophisticated scheme is necessary when the
`data rates are not
`the same as in many situations.
`Furthermore, the data rates vary with time, so decomposing
`block 38 must also vary with time, tracking the character-
`istics of the communication channel,
`to ensure that
`the
`aggregate data rate is optimal at all times.
`Referring now to FIG. 3A,in the simplest case, decom-
`posing block 38 receives one input signal 50 and outputs two
`signals 52 and 54 along a first channel 56 and a second
`channel 58, respectively. Therefore, in the preferred imple-
`mentation of the present invention an input data stream S,is
`transceived by encoder 12 which is to be sampled at the
`highest rate. When the required sample rate is higher than
`the maximum communication rate of any of the communi-
`cation channels, S, is decomposedinto S,,, and S,. such that
`the rate of S,,, is equal to the maximum rate of one of output
`channels 56 and 58, say channel Y,. The signal S,, becomes
`Y, and is transceived within the telephone network 16. If the
`communication rate of S,. exceeds the maximum rate of any
`channel, besides the channel already used by Y,, then the
`same procedureis repeated with respectto S,,, 1. S,, is split
`into S,.,, and S,,>5. If the rate of S,,
`is less than the
`maximum rate of any channel, say Y,,, then S,, is trans-
`ceived along S, to the telephone network 16. In many
`situations the maximum rate of Y,, is greater than S,, and
`therefore another input signal S$; (not shown) or a part
`thereof may be transceived along Y,,,. The configuration will
`be discussed in detail hereinafter.
`
`Therefore, the maximum rate overfirst channel 56 is S,,
`and the maximum rate over second channel 58is S,,. Then
`the downsampling and upsampling ratios must be chosen so
`that:
`
`Po_Si Q)
`
`Shown in FIG. 2, decomposing block 38 has M input
`signals and N output signals. The total number of samples is
`preserved during the sampling operation such that the total
`
`— = — and
`50
`samplingrate of all input signals M is equal to the sampling
`go
`Si
`rate of all output signals N.
`There are various configurations to provide the function-
`ality of decomposing block 38. In one configuration, a
`multirate digital signal processor (“multirate DSP”) changes
`the data rates of the one or more data streamsthat are applied
`to decomposing block 38. The theory of multirate DSP is
`well-known and is described for example in the book
`Mutltirate Digital Signal Processing, Prentice Hall, Engle-
`wood Cliffs, N.J., 1983, by R. E. Crochiere and L. R.
`Rabiner, and is considered in J. Kovacevic and M. Vetterli,
`“Perfect Reconstruction Filter Banks with Rational Sam-
`
`bi _ Si
`a
`Si
`
`hold, subject to the constraint
`
`Po Pl _y
`Go
`4
`
`55
`
`60
`
`a
`
`)
`
`pling Rate Changes,” Proc. IEEE ICASSP, 1991, Toronto,
`ON, Canada, pp. 1788-1795, which are incorporated there
`by these references. The functionality of decomposing block
`38, will be described in greater detail hereafter.
`The data streams from decomposing block 38 are trans-
`ceived to a DC eliminator 40 and an ISDN conversion 42 in
`
`65
`
`where, Po, Py, Io, and q, are constants derived during and
`after the negotiation process between ISP 18 and the user or
`the users modem (not shown). Therefore the establishment
`of a connection encoder 12 and decoder 14 negotiate the data
`communication rates for the two telephone lines. These
`communication rates in turn uniquely determine the param-
`eters Po, Pp, and qo=q,=q.
`16
`
`16
`
`

`

`US 6,490,295 B1
`
`9
`This arrangement will guarantee that:
`
`S=S8;145j2
`
`(5)
`
`at all times. In other words the total communication rate is
`equal to the sum of the two rates, as long as (2), (3), (4) are
`fulfilled. Typically the downsampling and upsampling pp,
`Py Io. and q, constants are not uniquely determined by these
`equations but the smallest values for which the equations
`hold will be chosen.
`
`The implementation of all components of FIG. 3 is
`well-kn