Phase and frequency offset compensation in a...

Pulse or digital communications – Transceivers – Modems

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

C375S326000, C375S327000

Reexamination Certificate

active

06370188

ABSTRACT:

CROSS-REFERENCE TO RELATED APPLICATIONS
Not applicable.
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
Not applicable.
BACKGROUND OF THE INVENTION
This invention is in the field of telecommunications, and is more specifically directed to the digital signal processing of frequency multiplexed signals in such telecommunications.
In recent years, the data rates at which communications may be carried out over conventional telephone networks and wiring has greatly increased. These increases are due, in large part, to newly adopted techniques of multiplexing and modulating signals representative of the messages or data being communicated, resulting in greatly improved communication bandwidth. In addition, the carrier frequencis at which such communications are being carried out have also increased in recent years, futher improving the bit rate.
According to one well-known class of multiplexing, digital data are communicated at multiple sub-carrier frequencies, or tones. This class of frequency multiplexing is referred to as Discrete MultiTone (DMT) in wired communication, or alternatively as Orthogonal Frequency Division Multiplexing (OFDM) in wireless communication. In this type of multiplexing, the stream of data symbols are multiplexed into N parallel subchannels, each subchannel being associated with a sub-carrier frequency. After modulation, the sub-carriers are added and transmitted together as an analog signal; at the receiving end, the sub-channels are filtered from one another and the original non-multiplexed data streanm is recovered.
It is important in this type of multiplexing that neighboring sub-carrier frequencies do not interfere with one another. Of course, wide separation of the sub-carrier frequencies would eliminate such interchannel interference (ICI), but at a cost of low spectral density. A well known method of ensuring orthogonality of the sub-carriers, and thus avoiding ICI, is to utilize a rectangular pulse shape as the sub-carrier pulse. According to the theorems of the Fourier Transform, a rectangular pulse in the time-domain transforms into a
sin(x)
/
x
frequency-domain spectrum. In the frequency multiplexing case, this spectrum is centered about the sub-carrier frequency ƒ
0
, and has an argument x=&pgr;NT(ƒ−
0
), where ƒ refers to the actual frequency of the communication, N is the number of parallel subchannels being transmitted, and T is the period of communication of discrete information (i.e., the reciprocal of the symbol communication rate). Proper selection of the sub-carrier frequencies to ensure orthogonality follows the relationship:
ƒ
k
=
k
/
NT
where k is the sub-carrier index (i.e., the “tone” in the multitone set). If this relationship is maintained in assigning the sub-carrier frequencies, each sub-carrier will have a center frequency that is located at a zero crossing of the spectrum of other sub-carriers, and as such each sub-carrier will be orthogonal to the other encoded sub-carriers.
FIG. 1
illustrates the frequency spectrum of a sub-channel in OFDM or DMT transmission. In the frequency response plot of
FIG. 1
, the frequency axis is measured in sub-carrier index values relative to the index of the center frequency. The center frequency (relative index of 0) provides the maximum frequency response, as shown. At relative index values of ±1, for example, the frequency response for the sub-channel is at zero. Accordingly, the sub-channel illustrated in
FIG. 1
will provide no contribution at the center frequencies of the adjacent sub-channels (relative index of ±1); conversely, since the adjacent sub-channels have the same normalized frequency response as shown in
FIG. 1
, they will provide no contribution to the sub-channel of
FIG. 1
(relative index of 0). Furthermore, as shown in
FIG. 1
, the frequency response is zero at each integer value of relative index. As such, the illustrated sub-channel provides no contribution to any other sub-channel, and conversely no other sub-channel contributes to the signal at the center frequency of the illustrated sub-channel of FIG.
1
. The use of a rectangular pulse thus provides orthogonality among the various sub-channels, permitting close spacing of the center frequencies and thus high spectral density.
In order to maintain orthogonality, however, the modulation and demodulation of the signals must be performed at the same precise center frequencies. As evident from
FIG. 1
, if demodulation is performed at a frequency that is slightly offset from the center frequency, not only will the frequency response for the desired sub-channel be less than optimal, but the demodulated signal will also contain contributions from other sub-channels; these contributions amount to interchannel interference (ICI), and greatly reduce the signal quality of the system. It is therefore important to ensure precision in the demodulation frequencies in DMT/OFDM communication systems.
The precise matching of modulation and demodulation frequencies is made difficult in modem communications by the physical separation of communicating modems from one another, where each of the communicating modems is driven by its own local clock. Conventional DMT/OFDM modems typically use expensive and complicated circuitry to ensure such precise frequency mathcing.
FIG. 2
illustrates the construction of the receiver side of conventional DMT modem
10
. As shown in
FIG. 2
, modem
10
receives signals from the telephone network at analog-to-digital converter (A/D)
14
. The signals received by modem
10
include, in addition to the communicated messages, a pilot tone generated by the transmitting modem to communicate the frequency at which it carried out the modulation of the message data. The digital output of A/D
14
is processed by time-domain equalization function
20
, cyclic prefix removal function
22
, Fast Fourier Transform (FFT) function
24
, and frequency domain equalization function
26
(such functions typically performed by digital signal processor, or DSP,
12
), following which the received communicated signals are applied, in digital form, to the host computer of modem
10
. Temporal control of modem
10
is maintained in response to the pilot tone, as recovered from the received communication by FFT function
24
, which generates a digital value corresponding to the instantaneous frequency of this detected tone. The frequency of the pilot tone is filtered by digital filter function
28
(also typically within DSP
12
), converted into an analog signal by digital-to-analog converter (D/A)
18
, and applied to voltage controlled oscillator (VCXO)
16
. VCXO
16
responds to the analog signal corresponding to the pilot tone frequency to control A/D
14
, such that the time-domain sampling and conversion of the incoming received communication is performed at a frequency that precisely matches that of the transmitting modem (as communicated by way of the pilot tone). A phase-locked loop (not shown) may also be implemented in conventional modem
10
, to ensure stable matching of the output of VCXO
16
relative to the incoming signals.
It has been observed, however, that VCXO
16
is typically an expensive function to include in client-side modem systems, Furthermore, fluctuations in the control voltage appluied to VCXO
16
by D/A
18
directly cause frequency jitter at the output of VCXO
16
; such fluctuations are common for modems within electtrically noisy environments such as modern personal computers and workstations. As a result, the conventional modem construction, as shown in
FIG. 2
, includes expensive oscillator circuitry that still does not provide a high degree of precision in its frequency output when implemented in the usual applications.
A relatively new type of current modem communications technology is referred to in the art as digital subscriber line (“DSL”). DSL refers generally to a public network technology that delivers relatively high bandwidth over conventional telephone company copper wiring at limited distances

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Phase and frequency offset compensation in a... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Phase and frequency offset compensation in a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Phase and frequency offset compensation in a... will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-2908376

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.