PPP connect speed and V.92 support - PPP

This is a discussion on PPP connect speed and V.92 support - PPP ; Using Fedora Core 2 with pppd-2.4.2-2 and a 56k V.90 modem on /dev/ttyS4. Two questions: 1. is it possible to see the connect speed with pppstats or some other command line tool while connected (this system is not running X, ...

+ Reply to Thread
Results 1 to 18 of 18

Thread: PPP connect speed and V.92 support

  1. PPP connect speed and V.92 support

    Using Fedora Core 2 with pppd-2.4.2-2 and a 56k V.90 modem on /dev/ttyS4.

    Two questions: 1. is it possible to see the connect speed with pppstats or
    some other command line tool while connected (this system is not running X,
    kppd, and I set up the ppp and chat scripts by hand)? 2. Does pppd support
    a V.92 modems faster connect speed, along with options of 48k up/down
    instead of 53k down/33k up?



  2. Re: PPP connect speed and V.92 support

    ShadowEyez wrote:
    > Using Fedora Core 2 with pppd-2.4.2-2 and a 56k V.90 modem on /dev/ttyS4.


    > Two questions: 1. is it possible to see the connect speed with pppstats or
    > some other command line tool while connected (this system is not running X,
    > kppd, and I set up the ppp and chat scripts by hand)?


    Initialize the modem with W1s95=47 after any other AT initialization.
    Add the chat -vsS options, and the pppd updetach or nodetach option.
    I *think* that's all you need.

    2. Does pppd support
    > a V.92 modems faster connect speed, along with options of 48k up/down
    > instead of 53k down/33k up?


    Pppd supports speeds up to 11.52 kB/s (92.16 kb/s) given a serial
    device with a 16550A UART, which is much greater than a V92 modem
    supports for transferring data from a uniform distribution, but not
    much greater, or perhaps even less, than the transfer rate for some
    data that can be highly compressed by the modem.

    It's the V92 modem that must support and be configured for 48 kb/s
    up/down, *and* the DCE on the other end of the connection, usually the
    ISP DCE, must also support it. That last part is important since in
    many (most?) instances it's not yet supported in ISP V92 DCE, although
    the 53 kb/s down rate is supported - given a high-quality landline
    connection.

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/
    /* Microsoft is a great marketing organization.
    * It _has_ to be */

  3. Re: PPP connect speed and V.92 support

    On Mon, 26 Jul 2004 16:01:04 -0500, Clifford Kite wrote:
    >ShadowEyez wrote:
    >
    >2. Does pppd support
    >> a V.92 modems faster connect speed, along with options of 48k up/down
    >> instead of 53k down/33k up?

    >
    >It's the V92 modem that must support and be configured for 48 kb/s
    >up/down, *and* the DCE on the other end of the connection, usually the
    >ISP DCE, must also support it. That last part is important since in
    >many (most?) instances it's not yet supported in ISP V92 DCE, although
    >the 53 kb/s down rate is supported - given a high-quality landline
    >connection.
    >


    You also need to find out if a suitable ISP available in your area offers
    V92, ShadowEyez. The ( supposedly ) main promotional website for V92,
    http://www.v92.com has a list of "V92 ISPs" viewable on web-page
    http://www.v92.com/v92isp

    There don't seem many of them, if that list is up-to-date. Probably not so
    bad in the States, as I suspect you may be. There are mixed findings about
    V92 modems. A selection of such comments can be seen at web-page
    http://www.modemsite.com/56k/v92sur.asp

    My own experiences with a "V92-capable" modem have not been happy. It
    wouldn't work for > 5 minutes because of "retrain fails", so I went back to
    a V90 modem. Clifford's comment about the landline needing to be "high
    quality" is so true !

    So don't rush in to anything, ShadowEyez. Find a suitable ISP. Get some
    recommendations about the best V92 modems - indeed, ask the ISP what they
    can recommend - what do their staff use at home ? Make sure you get a full
    receipt when you buy the modem, so you can return it if like mine, it's
    ultra-fussy about the phone line....

    Philip Dodd

  4. Re: PPP connect speed and V.92 support

    "ShadowEyez" writes:

    ]Using Fedora Core 2 with pppd-2.4.2-2 and a 56k V.90 modem on /dev/ttyS4.

    ]Two questions: 1. is it possible to see the connect speed with pppstats or
    ]some other command line tool while connected (this system is not running X,

    You can sometimes find the connect speed that was negotiated at the
    beginning of the call (returned with the CONNECt response if the modem
    option to do so is set) You cannot find out what the connect speed is now.

    ]kppd, and I set up the ppp and chat scripts by hand)? 2. Does pppd support
    ]a V.92 modems faster connect speed, along with options of 48k up/down
    ]instead of 53k down/33k up?

    This does not exist. the standard only allows a max of 53K down and 33K up.
    To get more than 33K you MUST have direct digital connection to the phone
    line. The modem does not have that. It only has analog connection. You
    could get a digitial modem which might allow you 56K both up and down.




  5. Re: PPP connect speed and V.92 support

    Bill Unruh wrote:
    > "ShadowEyez" writes:


    > ] .... 2. Does pppd support
    > ]a V.92 modems faster connect speed, along with options of 48k up/down
    > ]instead of 53k down/33k up?


    > This does not exist. the standard only allows a max of 53K down and 33K up.
    > To get more than 33K you MUST have direct digital connection to the phone
    > line. The modem does not have that. It only has analog connection. You
    > could get a digitial modem which might allow you 56K both up and down.


    Sorry, but the standard does exist for V.92 analog modems. It's called
    Pulse Code Modulation (PCM) Upstream. It is supposed to give you 48 kb/s
    upstream and downstream with the right modem configuration.

    It's rather doubtful that there yet exist many, if any, modems that
    implement it. ISP DCE V.92 implementations with it are likely to be
    scarce too.

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/
    /* 97.3% of all statistics are made up. */

  6. Re: PPP connect speed and V.92 support

    Clifford Kite writes:

    ]Bill Unruh wrote:
    ]> "ShadowEyez" writes:

    ]> ] .... 2. Does pppd support
    ]> ]a V.92 modems faster connect speed, along with options of 48k up/down
    ]> ]instead of 53k down/33k up?

    ]> This does not exist. the standard only allows a max of 53K down and 33K up.
    ]> To get more than 33K you MUST have direct digital connection to the phone
    ]> line. The modem does not have that. It only has analog connection. You
    ]> could get a digitial modem which might allow you 56K both up and down.

    ]Sorry, but the standard does exist for V.92 analog modems. It's called
    ]Pulse Code Modulation (PCM) Upstream. It is supposed to give you 48 kb/s
    ]upstream and downstream with the right modem configuration.

    OK, I stand (or sit) corrected. Any idea how it works? The standard line
    was that the 56K downstream connection was only possible because the modem
    connected directly to the digital network, and thus could directly use the
    exact digitial amplitudes to carry information (remember that the bandwidth
    is only 4KHz= 4000 bits per second-- thus one needs to use the amplitude
    bits to push that up to 56K). How does V92 allow the upstream, which does
    not have access to the digitial amplitude levels, but only analog levels
    which may map badly to digitial, to get more than 33K?



    ]It's rather doubtful that there yet exist many, if any, modems that
    ]implement it. ISP DCE V.92 implementations with it are likely to be
    ]scarce too.

    ]--

  7. Re: PPP connect speed and V.92 support

    Bill Unruh wrote:
    > Clifford Kite writes:
    > ]Bill Unruh wrote:


    > ]> To get more than 33K you MUST have direct digital connection to the phone
    > ]> line. The modem does not have that. It only has analog connection. You
    > ]> could get a digitial modem which might allow you 56K both up and down.


    > ]Sorry, but the standard does exist for V.92 analog modems. It's called
    > ]Pulse Code Modulation (PCM) Upstream. It is supposed to give you 48 kb/s
    > ]upstream and downstream with the right modem configuration.


    > OK, I stand (or sit) corrected. Any idea how it works? The standard line
    > was that the 56K downstream connection was only possible because the modem
    > connected directly to the digital network, and thus could directly use the
    > exact digitial amplitudes to carry information (remember that the bandwidth
    > is only 4KHz= 4000 bits per second-- thus one needs to use the amplitude
    > bits to push that up to 56K). How does V92 allow the upstream, which does
    > not have access to the digitial amplitude levels, but only analog levels
    > which may map badly to digitial, to get more than 33K?


    Heh! I said a standard exists, and there is ample evidence on the web
    that it does, not that I could explain the theory behind it.

    How do you get 9600 b/s, or 14.4 kb/s, or 33.6 kb/s out of an analog
    modem using a signal that is limited to 2400 baud on the wire? The name
    of the game is modulation type; PCM just does better than the V.34bis
    used for 33.6 kb/s.

    Here is a simple description:

    http://www.v92.com/about/pcmupstream.html

    Searching for Pulse Code Modulation with google will bring up some more
    detailed answers.

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/

  8. Re: PPP connect speed and V.92 support

    Clifford Kite writes:

    ]Bill Unruh wrote:
    ]> Clifford Kite writes:
    ]> ]Bill Unruh wrote:

    ]> ]> To get more than 33K you MUST have direct digital connection to the phone
    ]> ]> line. The modem does not have that. It only has analog connection. You
    ]> ]> could get a digitial modem which might allow you 56K both up and down.

    ]> ]Sorry, but the standard does exist for V.92 analog modems. It's called
    ]> ]Pulse Code Modulation (PCM) Upstream. It is supposed to give you 48 kb/s
    ]> ]upstream and downstream with the right modem configuration.

    ]> OK, I stand (or sit) corrected. Any idea how it works? The standard line
    ]> was that the 56K downstream connection was only possible because the modem
    ]> connected directly to the digital network, and thus could directly use the
    ]> exact digitial amplitudes to carry information (remember that the bandwidth
    ]> is only 4KHz= 4000 bits per second-- thus one needs to use the amplitude
    ]> bits to push that up to 56K). How does V92 allow the upstream, which does
    ]> not have access to the digitial amplitude levels, but only analog levels
    ]> which may map badly to digitial, to get more than 33K?

    ]Heh! I said a standard exists, and there is ample evidence on the web
    ]that it does, not that I could explain the theory behind it.

    ]How do you get 9600 b/s, or 14.4 kb/s, or 33.6 kb/s out of an analog
    ]modem using a signal that is limited to 2400 baud on the wire? The name
    ]of the game is modulation type; PCM just does better than the V.34bis
    ]used for 33.6 kb/s.

    Shannon tells us that datarate=freq x ln_2(S/N) where S is signal and N is
    noise. Ie you can use the amplitude above the noise threshold-- for ideal
    digital signals that is the digitization threshold-- to carry more bits.
    Since the telco frequency is about 4KHz, to get to 56Kbit, ln_2(S/N)=14.
    -- ie 14 bits of amplitude discrimination. Now any D-A or A_d conversion
    could easily lead to bit creap-- ie waht was one bit level on the D side,
    becomes another on the A side.
    It also depends on the noise level, etc.

    Now a digital modem can put out exactly the level it wants. An analog has a
    harder time. So I am wondering how they do it.

    Most of the web sites just say they do, not how.

    And the Pulse Code Modulation sites are not very helpful either-most are
    concerned with the situation where the Ln_2(S/N)=1.


    ]Here is a simple description:

    ]http://www.v92.com/about/pcmupstream.html

    ]Searching for Pulse Code Modulation with google will bring up some more
    ]detailed answers.

    ]--
    ]Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    ]PPP-Q&A links, downloads: http://ckite.no-ip.net/


  9. Re: PPP connect speed and V.92 support

    Bill Unruh wrote:

    > Shannon tells us that datarate=freq x ln_2(S/N) where S is signal and N is
    > noise. Ie you can use the amplitude above the noise threshold-- for ideal
    > digital signals that is the digitization threshold-- to carry more bits.
    > Since the telco frequency is about 4KHz, to get to 56Kbit, ln_2(S/N)=14.
    > -- ie 14 bits of amplitude discrimination. Now any D-A or A_d conversion
    > could easily lead to bit creap-- ie waht was one bit level on the D side,
    > becomes another on the A side.
    > It also depends on the noise level, etc.


    > Now a digital modem can put out exactly the level it wants. An analog has a
    > harder time. So I am wondering how they do it.


    > Most of the web sites just say they do, not how.


    Neither do I, but clearly some people are convinced that they do know how.
    Even so working implementations may not be easy, a thesis which seems to
    be supported by the fact that support for 48/48 is still scare several
    years after the idea was incorporated as a part of V.92.

    ISPs may also be reluctant to replace existing V.90 (or early V.92) DCE
    with 48/48 capable V.92, which wouldn't seem to provide much benefit to
    the vast majority of their customers. It also would encourage a small,
    but perhaps significant, group to connect 24/7 and run their own website
    from home - even with ISP policies that impose a connection time limit
    and disconnect links that exceed it.

    In any case I think this thread is now in the off-topic category (not
    to mention that thinking about information theory makes my head hurt),
    so any further comment by me will be very limited.

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/

  10. Re: PPP connect speed and V.92 support

    In article , Bill Unruh wrote:
    >Shannon tells us that datarate=freq x ln_2(S/N) where S is signal and N is
    >noise. Ie you can use the amplitude above the noise threshold-- for ideal
    >digital signals that is the digitization threshold-- to carry more bits.
    >Since the telco frequency is about 4KHz, to get to 56Kbit, ln_2(S/N)=14.
    >-- ie 14 bits of amplitude discrimination. Now any D-A or A_d conversion
    >could easily lead to bit creap-- ie waht was one bit level on the D side,
    >becomes another on the A side.
    >It also depends on the noise level, etc.


    I'm sorry, Bill, you're confusing me here. What are you asking?

    >Now a digital modem can put out exactly the level it wants. An analog has a
    >harder time. So I am wondering how they do it.


    Are you asking how a V90 modem packs 56k into a 3.6k channel? If so, the
    answer is QAM (Quadrature Amplitude Modulation) where the modulation gets
    16 states per baud by using four amplitude levels _AND_ four values of
    phase shift keying. I've also seen this mechanism called "trellis code
    modulation". The detection or demodulation scheme is a bit simpler,
    because you are only looking to check four possible amplitude values,
    and four (or 8) of four (or 8) zero crossing points. The transmission is
    fairly sensitive to echos on the line, but that's why the modem training
    thing takes a bit longer, so each modem can "null" out the echos.

    If you are asking the actual technique to create the analog signal,
    that's just an oscillator with quadraphase outputs (typically, sine and
    cosine out, and one or two inverters), and two attenuator settings, all
    of which are selected with some simple digital logic. There is usually
    a couple of poles of low-pass filtering to knock the state switching
    noise down and that's about it.

    Old guy.

  11. Re: PPP connect speed and V.92 support

    ibuprofin@painkiller.example.tld (Moe Trin) writes:

    ]In article , Bill Unruh wrote:
    ]>Shannon tells us that datarate=freq x ln_2(S/N) where S is signal and N is
    ]>noise. Ie you can use the amplitude above the noise threshold-- for ideal
    ]>digital signals that is the digitization threshold-- to carry more bits.
    ]>Since the telco frequency is about 4KHz, to get to 56Kbit, ln_2(S/N)=14.
    ]>-- ie 14 bits of amplitude discrimination. Now any D-A or A_d conversion
    ]>could easily lead to bit creap-- ie waht was one bit level on the D side,
    ]>becomes another on the A side.
    ]>It also depends on the noise level, etc.

    ]I'm sorry, Bill, you're confusing me here. What are you asking?

    ]>Now a digital modem can put out exactly the level it wants. An analog has a
    ]>harder time. So I am wondering how they do it.

    ]Are you asking how a V90 modem packs 56k into a 3.6k channel? If so, the

    Yes.

    ]answer is QAM (Quadrature Amplitude Modulation) where the modulation gets
    ]16 states per baud by using four amplitude levels _AND_ four values of
    ]phase shift keying. I've also seen this mechanism called "trellis code

    No idea what a "phase shift keying" is. The amplitude is all that you have
    left after you have set the max frequency. You have no phase shifting, as
    that would change the frequency.

    ]modulation". The detection or demodulation scheme is a bit simpler,
    ]because you are only looking to check four possible amplitude values,
    ]and four (or 8) of four (or 8) zero crossing points. The transmission is
    ]fairly sensitive to echos on the line, but that's why the modem training
    ]thing takes a bit longer, so each modem can "null" out the echos.


  12. Re: PPP connect speed and V.92 support

    unruh@string.physics.ubc.ca (Bill Unruh) writes:
    > No idea what a "phase shift keying" is.


    It's a classic modulation technique. As with most topics, a quick
    search through google.com will turn up all sorts of references.
    Here's one:

    http://www.physics.udel.edu/wwwusers...sguys/psk.html

    and, of course, the Wikipedia:

    http://en.wikipedia.org/wiki/Phase-shift_keying

    > The amplitude is all that you have
    > left after you have set the max frequency. You have no phase shifting, as
    > that would change the frequency.


    Really?

    --
    James Carlson, IP Systems Group
    Sun Microsystems / 1 Network Drive 71.234W Vox +1 781 442 2084
    MS UBUR02-212 / Burlington MA 01803-2757 42.497N Fax +1 781 442 1677

  13. Re: PPP connect speed and V.92 support

    Bill Unruh wrote:
    > ibuprofin@painkiller.example.tld (Moe Trin) writes:


    > ]answer is QAM (Quadrature Amplitude Modulation) where the modulation gets
    > ]16 states per baud by using four amplitude levels _AND_ four values of
    > ]phase shift keying. I've also seen this mechanism called "trellis code


    > No idea what a "phase shift keying" is. The amplitude is all that you have
    > left after you have set the max frequency. You have no phase shifting, as
    > that would change the frequency.


    http://www.physics.udel.edu/wwwusers...sguys/qam.html

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/
    /* Those who can't write, write manuals. */

  14. Re: PPP connect speed and V.92 support

    unruh@string.physics.ubc.ca (Bill Unruh) writes:
    >
    >No idea what a "phase shift keying" is. The amplitude is all that you have
    >left after you have set the max frequency. You have no phase shifting, as
    >that would change the frequency.
    >


    Clearly not. It changes the bandwidth consumed by the harmonics of
    the waveform, but does not, in any way, change the carrier frequency.

    PSK and DPSK have been in modem modulations since the 1960s. Have a look
    at the V.22 (1200 bps) modulation spec, for example. 2 Amplitudes and
    2 phase states form 4 possible states per baud, which can encode 2 data
    bits per baud. 600 baud * 2 bits = 1200 bps.

    The faster modulations increase the number of both the amplitude states
    and the phase states per baud. Look it up.

    -Greg
    --
    Do NOT reply via e-mail.
    Reply in the newsgroup.

  15. Re: PPP connect speed and V.92 support

    Clifford Kite writes:

    ]Bill Unruh wrote:
    ]> ibuprofin@painkiller.example.tld (Moe Trin) writes:

    ]> ]answer is QAM (Quadrature Amplitude Modulation) where the modulation gets
    ]> ]16 states per baud by using four amplitude levels _AND_ four values of
    ]> ]phase shift keying. I've also seen this mechanism called "trellis code

    ]> No idea what a "phase shift keying" is. The amplitude is all that you have
    ]> left after you have set the max frequency. You have no phase shifting, as
    ]> that would change the frequency.

    ]http://www.physics.udel.edu/wwwusers/watson/student_projects/scen167/thosguys/qam.html

    Except it is nonesense. That phase shifting of successive waves in
    successive bins will require
    a much much higher frequency than the carrier to preserve it. If you put
    their signal through a band pass filter, all of those discontinuities will
    disappear, and you will just get a continuous, unphaseshifted signal with
    varying amplitudes. To get a feeling for what would happen, sample the wave
    form twice per baud. That gives the effective filtering (well it is
    actually filtering plus folding). Now you just have a varying amplitude
    with no "phase shifting" evident. So I am still confused.


    ]--
    ]Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    ]PPP-Q&A links, downloads: http://ckite.no-ip.net/
    ]/* Those who can't write, write manuals. */


  16. Re: PPP connect speed and V.92 support

    James Carlson writes:

    ]unruh@string.physics.ubc.ca (Bill Unruh) writes:
    ]> No idea what a "phase shift keying" is.

    ]It's a classic modulation technique. As with most topics, a quick
    ]search through google.com will turn up all sorts of references.
    ]Here's one:

    ]http://www.physics.udel.edu/wwwusers/watson/student_projects/scen167/thosguys/psk.html

    ]and, of course, the Wikipedia:

    ]http://en.wikipedia.org/wiki/Phase-shift_keying

    ]> The amplitude is all that you have
    ]> left after you have set the max frequency. You have no phase shifting, as
    ]> that would change the frequency.

    ]Really?


    Look at www.theory.physics.ubc.ca/fsk.gif and
    www.theory.physics.ubc.ca/fskbp.gif
    The first is the six numbers 1 2 3 0 1 2 encoded with the simply phase
    shift keying. (the bin boundaries occur at each multiple of 100)
    Note that very sharp transitions in the first graph when the
    successive bins are phase shifted. The second graph is the first graph put
    through a simple fourth order filter with the cutoff being the bin
    frequency ( ie 6 cycles over the range of the graph). Note that all
    evidence of the phase shifting has dissappeared and all that is left is
    variations in amplitude of the signal No sharp discontinuities (they
    require much higher frequencies to be represented.). Ie, the phase shift
    keying is just the same as amplitude keying. and if one now sticks this
    into a time and amplitude digitization circuit all you have left is
    variations in amplitude.


  17. Re: PPP connect speed and V.92 support

    unruh@string.physics.ubc.ca (Bill Unruh) writes:
    > Look at www.theory.physics.ubc.ca/fsk.gif and
    > www.theory.physics.ubc.ca/fskbp.gif
    > The first is the six numbers 1 2 3 0 1 2 encoded with the simply phase
    > shift keying. (the bin boundaries occur at each multiple of 100)
    > Note that very sharp transitions in the first graph when the
    > successive bins are phase shifted. The second graph is the first graph put
    > through a simple fourth order filter with the cutoff being the bin
    > frequency ( ie 6 cycles over the range of the graph). Note that all
    > evidence of the phase shifting has dissappeared and all that is left is
    > variations in amplitude of the signal No sharp discontinuities (they
    > require much higher frequencies to be represented.). Ie, the phase shift
    > keying is just the same as amplitude keying.


    I'm not sure what that's supposed to prove. Sure; if you put a signal
    through a filter, you can lose data. So ... ?

    > and if one now sticks this
    > into a time and amplitude digitization circuit all you have left is
    > variations in amplitude.


    Phase angle is neither frequency nor amplitude.

    I think you're looking at bandwidth rather than carrier frequency.
    The two are *quite* a bit different. Even plain old amplitude
    modulation produces a change in bandwidth -- which is noted in the
    frequency domain as sidebands on the carrier. In fact, every kind of
    modulation produces an effect in the frequency domain, and that change
    can be blunted or erased by filtering. That doesn't mean that it's a
    change of carrier frequency.

    --
    James Carlson, IP Systems Group
    Sun Microsystems / 1 Network Drive 71.234W Vox +1 781 442 2084
    MS UBUR02-212 / Burlington MA 01803-2757 42.497N Fax +1 781 442 1677

  18. Re: PPP connect speed and V.92 support

    In article , Greg Andrews wrote:
    >PSK and DPSK have been in modem modulations since the 1960s.


    Bell 212A DPSK at 600 baud, giving 1200 BPS. I still have several
    modems that know how to talk Bell 103 (300 BPS at 300 baud using
    frequency shift keying), though I have no idea why one would want to.
    There was also a V.21 standard for 300 BPS/300 baud, which was
    incompatible with Bell 103. I'm pretty sure it was also FSK.

    >Have a look at the V.22 (1200 bps) modulation spec, for example. 2
    >Amplitudes and 2 phase states form 4 possible states per baud, which
    >can encode 2 data bits per baud. 600 baud * 2 bits = 1200 bps.


    4 states per baud transmitting 2 bits per baud ???

    Actually, 'V.22' was 2 bits per baud using DPSK like (but incompatible
    with) Bell 212A. 'V.22.bis' used QAM with two amplitude levels and two
    phase shifts to acheive 2400 BPS. Things got more complex from there,
    through V.32, V.32.bis, V.34, and then V,90. I'm not up on that
    new-fangled V.92 jazz.

    Old guy

+ Reply to Thread