RU2428805C2 - Channel assessment with application of frequency smoothing - Google Patents

Abstract

FIELD: information technologies.

SUBSTANCE: method to calculate channel assessment using frequency smoothing consists in the fact that a tuning signal is received in a quadrature demodulation receiver, besides, the tuning signal represents multiple reference complex values and multiple appropriate mirror complex values, besides, the sum of products of reference complex values and appropriate mirror complex values equals zero; multiple processed signals are generated (y), which represent complex plant information, on the basis of a tuning signal; each processed signal (y) is multiplied by a coupled value of the appropriate reference complex value (p*), and channel assessment (h) is produced on the basis of multiplication results.

EFFECT: elimination of unbalance between I and Q tracks.

Description

This patent application claims priority for provisional application No. 60/896480, filed March 22, 2007, entitled QUADRATURE IMBALANCE MITIGATION USING UNBIASED TRAINING SIGNALS, pending; Patent Attorney's case number 071040P1.

This patent application is a partial continuation of the application for the grant of a US patent under the number 11/684566, filed March 9, 2007, entitled QUADRATURE MODULATION ROTATING TRAINING SEQUENCE, pending, assigned to the assignee of this application and, therefore, incorporated herein by reference .

This patent application is a partial continuation of the application for the grant of US patent No. 11/755719, filed on May 30, 2007, entitled QUADRATURE IMBALANCE MITIGATION USING UNBIASED TRAINING SEQUENCES, patent attorney case number 060395B1, pending, assigned to this assignee as incorporated herein by reference.

This patent application is a related U.S. patent application entitled QUADRATURE IMBALANCE ESTIMATION USING UNBIASED TRAINING SEQUENCES, patent attorney case number 060395B2, which is both filed and assigned to the assignee of this application and incorporated herein by reference in its entirety.

FIELD OF THE INVENTION

The present invention generally relates to channel estimation, in particular to systems and methods for improving the use of unbiased quadrature modulation training sequences when tuning receiver channel estimates by eliminating quadrature discrepancy errors.

State of the art

Figure 1 is a schematic block diagram of a conventional input stage receiver (prior art). A conventional wireless receiver includes an antenna that converts the emitted signal into a conducted signal. After some initial filtering, the conducted signal is amplified. Given a sufficient power level, the frequency of the carrier signal can be converted by mixing the signal (down-conversion) with the local oscillator signal. Since the received signal is subjected to quadrature modulation, the signal is demodulated by separate paths I (in-phase component) and Q (quadrature component) before combining. After frequency conversion, the analog signal can be converted to a digital signal using an analog-to-digital converter (ADC) to process the base band. Processing may include fast Fourier transform (FFT).

There are a number of errors that can be made to the receiver, which adversely affects channel estimates and restoration of a given signal. Errors can be made from mixers, filters, and passive components such as capacitors. Errors are amplified if they cause an imbalance between paths I and Q. In an attempt to evaluate the channel and thus nullify some of these errors, communication systems may use a message format that includes a training sequence, which may be a repeated or predefined data symbol . For example, using an orthogonal frequency division multiplexing (OFDM) system, the same common-mode-quadrature (IQ) constellation point may be retransmitted for each subcarrier.

In an effort to conserve energy in battery-powered devices, some OFDM systems use only a single modulation symbol for tuning. For example, a single direction in the constellation (for example, tract I) is stimulated, while another direction (for example, path Q) is not stimulated. The same type of unidirectional tuning can also be used with pilot tones. Note: scrambling a single modulation channel (for example, channel I) with symbol values of ± 1 does not rotate the constellation point and does not provide stimulation for the quadrature channel.

In the presence of quadrature path imbalance, which is predominant in systems with a large bandwidth, the aforementioned energy-saving training sequence leads to a biased channel estimate. A biased channel estimate may well align IQ constellation in one direction (i.e., path I), but give quadrature imbalance in the orthogonal direction. Preferably, any imbalance is distributed evenly between the two channels.

2 is a block diagram illustrating a quadrature imbalance on the receiver side (prior art). Although not shown, the transmitter side imbalance is similar. Assume that the Q path is a reference signal. The shape of the incident wave is described using cos ( ωt + θ ), where θ is the phase of the channel. The Q path is downconverted with -sin ( ωt ). The path I is converted with decreasing frequency using (1 + 2 ε ) cos ( ωt + 2Δ φ ). 2Δ φ ε and 2 represent hardware imbalance error respectively in phase and amplitude error. Low pass filters H _I and H _Q are different for each path. Filters introduce additional amplitude and phase distortion. However, these additional distortions are concentrated within 2Δ φ 2 and ε. Note: these two filters are real and affect + ω and -ω in the same way.

Provided that the errors are small:

(1 2 + ε) cos (ωt + 2Δ φ ) ≈ ( 1 2 + ε) cos (ωt) -2Δ φ . sin ( ωt )

The first component on the right, cos ( ωt ), is an ideal path I, slightly scaled. The second component, 2Δ φ .sin ( ωt ), is a small leak from the Q path. After conversion with decreasing frequency of the incident signal:

in the path I: (1 + 2 ε ) cos ( θ ) +2 ε .sin ( θ ).

in the path Q: sin ( θ ).

Errors cause an incorrect interpretation of the positions of the characters in the constellation of quadrature modulation, which, in turn, leads to incorrectly demodulated data.

Disclosure of invention

Wireless receivers are prone to errors caused by a lack of tolerance in hardware components associated with mixers, amplifiers, and filters. In quadrature demodulators, these errors can also cause an imbalance between paths I and Q, leading to incorrectly processed data.

The tuning signal can be used to calibrate the receiver channels. However, a tuning signal that does not excite both paths I and Q does not solve the problem of imbalance between the two paths. An unbiased training signal can be used to stimulate both paths I and Q, which leads to a better channel estimate. Typically, channel estimates are derived from predetermined information associated with positive or reference (+ f) subcarriers. Better channel estimates can be obtained if negative or mirror (-f) subcarriers are also used. In addition, by correctly combining adjacent or nearly adjacent reference subcarriers with adjacent or almost adjacent mirror subcarriers, frequency smoothing can be applied to the training signal, which acts as a means of eliminating channel bias errors.

Accordingly, a method is provided for supplying a frequency-smoothed communication tuning signal. The method generates a frequency-smoothed unbiased training signal in a quadrature modulation transmitter. The frequency-smoothed unbiased training signal includes a plurality of pilot signal products, where each pilot signal product includes complex plane information represented by a reference frequency subcarrier multiplied by complex plane information represented by a mirror frequency subcarrier. The sum of the plurality of pilot signal products is zero. The method provides a frequency-smoothed unbiased training signal so that it can be transmitted within a single symbol period.

Typically, a frequency-smoothed unbiased training signal includes a plurality of contiguous reference frequency subcarriers and a plurality of contiguous mirror frequency subcarriers. For example, a frequency-smoothed unbiased training signal may include a group of adjacent reference frequency frequency subcarriers without intermediate subcarriers and a plurality of adjacent mirror frequency frequency subcarriers without intermediate subcarriers.

In one aspect, a frequency-smoothed unbiased training signal may be represented as follows:

Σ p _i p _im = 0, i = from 1 to n;

where p is the reference frequency subcarrier,

where p _m is the mirror frequency subcarrier and

where n is equal to the number of pilot signal products.

A method for calculating a channel estimate using a frequency-smoothed unbiased training signal is also provided. This method receives a frequency-smoothed unbiased training sequence in a quadrature demodulation receiver. The frequency-smoothed unbiased training sequence includes a plurality of pilot signal products, where each pilot signal product includes predetermined complex plane information (p) represented by a reference frequency carrier (f) multiplied by predetermined complex plane information (p _m ) represented by the mirror frequency subcarrier (-f). The sum of the plurality of pilot signal products is zero. The method processes a frequency-smoothed unbiased training signal, generating a plurality of processed symbols (y) representing complex plane information. Each processed symbol (y) is multiplied by the conjugate value of the corresponding reference signal (p *), and a channel-smoothed frequency estimate (h) is obtained.

The following are additional details of the above method, systems for generating frequency-smoothed unbiased training signals, and systems for calculating a channel estimate using a frequency-smoothed unbiased training signal.

Brief Description of the Drawings

Figure 1 - block diagram of the traditional input stages of the receiver (prior art).

2 is a block diagram illustrating a quadrature imbalance on the receiver side (prior art).

3 is a block diagram illustrating an illustrative data transmission system.

4 is a block diagram of a system or device for supplying a frequency-smoothed unbiased training signal.

5 is a diagram illustrating a simple example of a frequency-smoothed unbiased training signal.

6 is a diagram illustrating a second example of a frequency-smoothed unbiased training signal.

7 is a diagram illustrating an unbiased training signal used as a group of pilot symbols accompanying communication symbols.

8 is a diagram illustrating a frequency-smoothed unbiased training signal used as a preamble preceding an undefined communication data.

9 is a block diagram of a system or device for calculating a channel estimate using a frequency-smoothed unbiased training signal.

10 illustrates the performance achieved by applying unbiased training signal algorithms to a standard for ultra-wideband communications (WiMedia, UWB).

11 is a flowchart illustrating a method of supplying a frequency-smoothed communication tuning signal.

12 is a flowchart illustrating a method of calculating a channel estimate using a frequency-smoothed unbiased training signal.

The implementation of the invention

Various embodiments are further described with reference to the drawings. In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of one or more aspects. However, it is obvious that such embodiments may be practiced without these specific details. In other instances, well-known structures and devices are illustrated in block diagram form in order to facilitate describing these embodiments.

The terms “processor”, “processing device”, “component”, “module”, “system” and the like, used in this application, are intended to indicate an entity related to a computer, any of the hardware, hardware implemented software, a combination of hardware tools and software, software or software in progress. For example, a component may be, but not limited to, a process running on a processor, a functional transformation, a processor, an object, an executable, a control thread, a program, and / or a computer. By way of illustration, both an application running on a computing device and a computing device can be a component. One or more components may be within the process and / or control flow, and the component may be localized on one computer and / or distributed between two or more computers. In addition, these components can be executed from various computer-readable media containing various data structures stored on them. Components can communicate through local and / or remote processes, for example, in accordance with a signal containing one or more data packets (for example, data from one component interacting with another component in a local system, distributed system, and / or through a network, such as the Internet, with other systems through a signal).

Various embodiments will be presented in terms of systems, which may include a number of components, modules, and the like. It is understood that various systems may include additional components, modules, and the like. and / or may not include all components, modules, and the like discussed with reference to the drawings. A combination of these approaches can also be used.

The various illustrative logic blocks, modules, and circuits that are described can be implemented or performed using a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof, designed to perform the functions described in the materials of this application ki. A general-purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

The methods or algorithms described in connection with the embodiments disclosed herein may be implemented directly in hardware, in a software module executed by a processor, or in a combination of the two. The software module may reside in RAM, flash memory, ROM, EPROM memory, EEPROM memory, registers, a hard disk, a removable disk, a compact disk, or any other form of storage medium known in the art. The storage medium may be coupled to the processor so that the processor can read information and write information to the storage medium. Alternatively, the storage medium may be an integral part of the processor. The processor and the storage medium may reside in an ASIC. The ASIC may reside on a host or elsewhere. Alternatively, the processor and the storage medium may reside as discrete components on a node or elsewhere in an access network.

Figure 3 is a block diagram depicting an exemplary data transmission system 300. The baseband processor 302 has an input on line 304 to receive digital information from a medium access control (MAC) layer. In one aspect, the baseband processor 302 includes an encoder 306 having an input on line 304 to receive digital (MAC) information, and an output on line 308 to provide encoded digital information in the frequency domain. To interleave the encoded digital information, an interleaver 310 can be used to feed the interleaved information in the frequency domain onto line 312. The interleaver 310 is a device that converts a single high-speed input signal into multiple parallel streams with a lower transmission rate, where each stream with a lower transmission rate is associated with a particular subcarrier. The inverse fast Fourier transform 314 (inverse FFT, IFFT) receives information in the frequency domain, performs an inverse FFT operation on the input information, and provides a digital time-domain signal to line 316. A digital-to-analog converter 318 converts the digital signal on line 316 to an analog baseband signal on line 320. As described in more detail below, the transmitter 322 modulates the baseband signal and provides a modulated carrier signal as an output signal on line 324. Note: alternative circuit configurations m capable of performing the same functions as those described above are probably known to those skilled in the art. Although not explicitly illustrated, the receiver system could consist of a similar set of components for reverse processing information received from the transmitter.

4 is a block diagram of a system or device for supplying a frequency-smoothed unbiased training signal. System 400 comprises a transmitter or signal generator means 402 having an input on line 404 for receiving training information, typically in digital form. For example, information may come from a MAC layer. Transmitter 402 has an output on line 406 for supplying a quadrature modulated and frequency-smoothed unbiased training signal.

Transmitter 402 may include a transmitter subsystem 407, such as a radio frequency (RF) RF transmitter subsystem, that uses an antenna 408 to communicate through an air or vacuum environment. However, it should be understood that the invention is applicable to any communication medium (e.g., wireless, wired, optical) capable of transmitting quadrature modulated information. The transmitter subsystem 407 includes an in-phase (I) modulation path 410, or means for generating the modulation information of modulation I. The transmitter subsystem 407 also includes a quadrature (Q) modulation path 412, or of means for generating the modulation Q modulation information Q. Path information I line 404a is upconverted in mixer 414 with carrier fc, while path information Q on line 404b is upconverted in mixer 416 with a phase-shifted carrier variant (fc + 90 °). Path 410 I and path 412 Q are summed on adder 418 and provided on line 420. In some aspects, the signal is amplified in amplifier 422 and supplied to antenna 408 on line 406 where frequency-smoothed unbiased training signals are emitted. Alternatively, paths I and Q may be referred to as channels I and Q. A frequency-smoothed unbiased training signal may also be referred to as a frequency-balanced training sequence and is part of a larger class of balanced or unbiased training signals described in related applications and in detail below.

The frequency-smoothed unbiased training signal includes a plurality of pilot signal products, where each pilot signal product includes complex plane information represented by a reference frequency subcarrier multiplied by complex plane information represented by a mirror frequency subcarrier. The sum of the plurality of pilot signal products is zero. Transmitter 402 provides a frequency-smoothed unbiased training signal within a single symbol period.

Not shown, but as an alternative, the components of a frequency-smoothed training signal (FSTS), for example, are supplied sequentially or served in batch mode and are collected in memory (not shown). Once complete FSTS is compiled, it can be issued for use within a single character period. In this aspect, the memory / collection and combination means may be considered part of the transmitter 402, even if they are involved in separate modules or devices (not shown). It should be understood that in some aspects, the transmitter 402 acts as a signal conditioning instrument, while the actual sending of the FSTS through the communication medium is performed by other modules or devices.

Typically, transmitter 402 also sends quadrature modulated (undefined) communication data. A frequency-smoothed unbiased training signal is used by a receiver (not shown) to create channel estimates that enable undefined communication data to recover more accurately. In one aspect, quadrature modulated communication data is sent after sending an unbiased training sequence. In another aspect, the unbiased training sequence is sent simultaneously with the communication data in the form of pilot signals. The system is not limited to any particular time relationship between the training signal and quadrature modulated communication data.

A message is a grouping of characters in a predefined format. A message may have a duration of several periods of characters. One or more characters may be transmitted each character period. Some messages include a preamble preceding the main body of the message. For example, a message may be generated as a long packet containing many OFDM, CDMA, or TDMA symbols.

FSTS may consist of 2 or more than 2 pieces of pilot signals. In one aspect, transmitter 402 generates a frequency-smoothed unbiased training signal including a plurality of adjacent reference frequency reference subcarriers and a plurality of adjacent mirror frequency subcarriers. Typically, reference subcarriers and corresponding mirror subcarriers are within a relatively dense (spectrally) proximity. For example, a frequency-smoothed unbiased training signal may include a group of adjacent reference frequency frequency subcarriers without intermediate subcarriers and a plurality of adjacent mirror frequency frequency subcarriers without intermediate subcarriers. The intermediate subcarrier, for example, may be a subcarrier carrying (undefined) communication data or other information not related to the training signal. In one embodiment, the group includes all the reference and mirror subcarriers in the FSTS.

In another aspect, a frequency-smoothed unbiased training signal may be represented as follows:

Σ p _i p _im = 0, i = from 1 to n;

where p is the reference frequency subcarrier,

where p _m is the mirror frequency subcarrier and

where n is equal to the number of pilot signal products.

As noted above, the sum of the products of the pilot signals is zero. However, due to system errors, the amount can sometimes be more accurately indicated as being near zero. In the worst case analysis, L pilot products are integrated as follows:

| sum p _i · p _im | = L.

If L is 100% and if | sum p _i · p _im | = L / 4, then the error is 25%. Frequency-smoothed unbiased training sequence with a 25% error still gives excellent results. If L / 2 is used (50% error), good results are obtained, since the common-mode-quadrature interference from the channel estimate is still attenuated by 6 dB.

As explained in more detail below, a frequency-smoothed unbiased training signal can be represented using weighted products of the pilot signals, as follows:

Σ w _i p _i p _im = 0, i = from 1 to n;

where w is the weight coefficient.

In one aspect, the unique weight associated with some of the pilot products may be unit (normalized). In some aspects, the individual weights are approximately the same and the weight can be removed from the equation as a constant.

5 is a diagram illustrating a simple example of a frequency-smoothed unbiased training signal. The first product of the pilot signals has a reference subcarrier 500 at a frequency + f representing information as the first value of the complex plane, and a mirror subcarrier 502 at a frequency -f representing the first value of the complex plane. Here, the "arrows" of the subcarriers can be represented as phasors having an amplitude of 1 and an angle of 90 degrees. The second pilot product has a reference subcarrier 504 at a frequency (f + 1), an adjacent frequency + f representing the first value of the complex plane, and a mirror subcarrier 506 at a frequency of (f + 1), an adjacent frequency -f representing the first value of the complex plane +180 degrees. To continue this example, the arrow representing the mirror subcarrier 506 would have to have an amplitude of 1 and an angle of 270 degrees.

Although all subcarriers have the same value, normalized to 1, in this example, it should be understood that more complex variations of this example can use uneven amplitudes. Similarly, FSTS is not limited to using only 90 degrees and 270 degrees angles. Furthermore, although an example of only 2 pilot signals is illustrated, the same methodology could be applied to FSTS with more than 2 pilot products. For example, the illustrated FSTS could be modified to add a third product of pilot signals (not illustrated) with a subcarrier at a frequency (f-1), an adjacent frequency + f representing the first value of the complex plane, and a mirror subcarrier at a frequency of (f-1 ) adjacent to the frequency -f, representing the first value of the complex plane +180 degrees.

6 is a diagram illustrating a second example of a frequency-smoothed unbiased training signal. The first pilot product has a reference subcarrier 600 at a frequency + f representing information as the first value of the complex plane, and a mirror subcarrier 602 at a frequency-f representing the first value of the complex plane. Again, the "arrows" of the subcarriers can be represented as phasors having an amplitude of 1 and an angle of 90 degrees. The second product of the pilot signals has a reference subcarrier 604 at a frequency (f + 1), an adjacent frequency + f representing the first value of the complex plane +90 degrees, and a mirror subcarrier 606 at a frequency of (f + 1), an adjacent frequency -f representing the first value of the complex plane is -90 degrees. To continue this example, an arrow representing the reference subcarrier 604 would have an amplitude of 1 and an angle of 180 degrees, and a mirror subcarrier 606 would have an amplitude of 1 and an angle of 0 degrees.

Again, in this example, the subcarriers have the same value, normalized to 1, and it should be clear that more complex variations of this example can use uneven amplitudes. Similarly, FSTS is not limited to using only angles of 0 degrees, 90 degrees, and 180 degrees. Furthermore, although an example of only 2 pilot signals is illustrated, the same methodology could be applied to FSTS with more than 2 pilot products. For example, the illustrated FSTS could be modified to add a third product of pilot signals (not illustrated) with a subcarrier at a frequency (f-1), an adjacent frequency + f representing the first value of the complex plane +90 degrees, and a mirror subcarrier at a frequency of (( f-1), an adjacent frequency of -f, representing the first value of the complex plane of -90 degrees.

7 is a diagram illustrating an unbiased training signal used as a group of pilot symbols accompanying communication symbols. In one aspect, in addition to receiving training information, the transmitter receives (undefined) communication data. Then, the unbiased frequency-smoothed training signal is generated by P products of the pilot signals along with (N-P) symbols (subcarriers) of the communication data. As a result, N subcarriers are served in one symbol period, incl. frequency-smoothed unbiased training signal and quadrature modulated communication data. Many communication systems, such as those compatible with IEEE 802.11 and UWB, use pilot signals for channel control purposes.

Alternatively, components of a frequency-smoothed training signal (FSTS), or communication data symbols, or both, may be supplied sequentially or in batch mode and collected in memory (not illustrated). Once all characters in a character period are collected, they can be issued for use within a single character period. In this aspect, the memory / collection and combination means can be considered part of the transmitter, even if they are involved in separate modules or devices.

8 is a diagram illustrating a frequency-smoothed unbiased training signal used as a preamble preceding an undefined communication data. As illustrated, the transmitter provides a frequency-smoothed unbiased training signal in a first symbol period using a group of reference frequency subcarriers and corresponding mirror frequency subcarriers. The transmitter receives communication data, generates quadrature modulated communication data in a group of reference frequency subcarriers and corresponding mirror frequency subcarriers, and supplies quadrature modulated communication data in a second symbol period following the first symbol period.

Although only the first and second symbol periods are clearly marked, it should be understood that the preamble may consist of a plurality of symbol periods, with FSTS used in some or all of the symbol periods of the preamble. Similarly, communication data may be issued in a plurality of symbol periods (not illustrated) following the preamble. For example, an ultra-wideband (UWB) communication system uses 6 symbol periods transmitted before transmission of communication data or a beacon. Therefore, one or more of 6 symbol periods may be used to transmit FSTS.

Although not specifically illustrated, the transmitter of FIG. 4 or elements of the transmitter can be used as a processing device to generate a frequency-smoothed unbiased training signal. In this case, the processing device would comprise a signal generator module having an input for receiving training information and an output for supplying a quadrature modulated and frequency-smoothed unbiased training signal. As described above, a frequency-smoothed unbiased training signal would include a plurality of pilot signal products, where each pilot signal product includes complex plane information represented by a reference frequency carrier multiplied by complex plane information represented by a mirror frequency subcarrier. As also shown above, the sum of the plurality of pilot signal products is zero. The signal generator module would provide a frequency-smoothed unbiased training signal within a single symbol period.

9 is a block diagram of a system or device for calculating a channel estimate using a frequency-smoothed unbiased training signal. The system or device 900 comprises a quadrature demodulation receiver or reception means 902 having an input on line 904 for receiving a frequency-smoothed unbiased training signal. Like the transmitter of FIG. 4, receiver 902 may be an RF device coupled to an antenna 905 for receiving radiated information. However, the receiver may alternatively receive an unbiased training sequence through a wired or optical medium (not illustrated).

Receiver 902 has an in-phase (I) demodulation path 906 for receiving demodulation I tuning information. A quadrature (Q) demodulation path 908 receives Q demodulation tuning information. Typically, the receiver 902 includes analog-to-digital converters 909 (ADC), a fast converter 910 Fourier transform (FFT), interleaver 912, and decoder 914. The receiver provides training information in response to receiving FSTS. The frequency-smoothed unbiased training sequence includes a plurality of a plurality of pilot signal products. The pilot product includes the predefined information (p) of the complex plane represented by the reference frequency subcarrier (f) times the predetermined information (p _m ) of the complex plane represented by the mirror frequency subcarrier (-f). The sum of the plurality of pilot signal products is zero.

The processor or processing means 916 has an input on line 918 for receiving training information, the processor generates a plurality of processed symbols (y) representing complex plane information. A processor 916 multiplies each processed symbol (y) by the conjugate value of the corresponding reference signal (p ^* ) and provides a frequency-smoothed channel estimate (h) on the output on line 920. In some aspects, the receiver 902 provides tuning information as an output from the ADC 909. In this aspect, the FFT, interleaver, and decoding processes or their equivalents are performed by processor 916.

FSTS consists of 2 or more pilot products. In one aspect, the receiver receives a frequency-smoothed unbiased training signal including a plurality of adjacent reference frequency reference subcarriers and a plurality of adjacent mirror-frequency frequency subcarriers. The value "adjacent" depends on the arrangement of subcarriers, frequency and other modulation characteristics. In another aspect, the frequency-smoothed unbiased training signal includes a group of adjacent reference frequency frequency subcarriers without intermediate subcarriers and a plurality of adjacent mirror frequency frequency subcarriers without intermediate subcarriers. This group may include all or only a subset of all subcarriers in the FSTS.

A received frequency-smoothed unbiased training signal may be expressed as transmitted by the FSTS, as follows:

Σ p _i p _im = 0, i = from 1 to n;

where p is the reference frequency subcarrier,

where p _m is the mirror frequency subcarrier and

where n is equal to the number of pilot signal products.

Like the transmitted FSTS, the received FSTS may include weighted pilot products, as follows:

Σ w _i p _i p _im = 0, i = from 1 to n;

where w is the weight coefficient.

Two examples of the transmitted FSTS were described in FIGS. 5 and 6. These same examples are valid examples of the received FSTS.

In one aspect, the receiver 902 may receive an unbiased frequency-smoothed training signal with P products of pilot signals and (N - P) symbols of communication data in the same symbol period and provide both training information and communication data (also see Fig.7). In yet another aspect, receiver 902 receives a frequency-smoothed unbiased training signal in a first symbol period with a group of reference frequency subcarriers and corresponding mirror frequency subcarriers. The receiver also receives quadrature modulated communication data in a group of reference frequency subcarriers and corresponding mirror frequency subcarriers in a second symbol period following the first symbol period, and provides communication data (see FIG. 8).

Although not specifically illustrated, the receiver of FIG. 9 can also be used as a processing device for calculating a channel estimate using a frequency-smoothed unbiased training signal. In this case, the processing device comprises a receiver module having an input for receiving a frequency-smoothed unbiased training sequence and an output for supplying training information. As described above, a frequency-smoothed unbiased training sequence includes a plurality of pilot products, where each product of pilot signals includes predetermined complex plane information (p) represented by a reference frequency subcarrier (f) times the predetermined information (p _m ) a complex plane represented by a mirror frequency subcarrier (-f). As also shown above, the sum of the plurality of pilot signal products is zero.

The calculation module has an input for receiving tuning information. The calculation module generates a plurality of processed symbols (y) representing information of the complex plane, multiplies each processed symbol (y) by the conjugate value of the corresponding reference signal (p ^* ) and provides a frequency-smoothed channel estimate (h) at the output.

The training signals used in the preamble or as pilot signals are similar for the reason that the information content of the transmitted data is usually predetermined or “known” data that allows the receiver to calibrate and measure the channel. When receiving (undefined) communication data, there are 3 unknowns: the data itself, the channel and noise. The receiver is not able to calibrate for noise because the noise changes randomly. A channel is a measurement, in most cases, associated with delay and multipath. For relatively short periods of time, the error resulting from multipath propagation can be measured if predetermined data, such as training or pilot signals, are used. Since the channel is known, this measurement can be used to eliminate errors in received (undefined) communication data. Therefore, some systems provide a tuning signal for measuring the channel before decoding of the data begins.

However, the channel may change, for example, as the transmitter or receiver moves in space or the clocks drift. From here, many systems continue to send additional “known” data along with “unknown” data in order to track slow changes in the channel.

Although not specifically illustrated, the transmitter of FIG. 4 and the receiver of FIG. 9 can be combined to form a transceiver. In fact, the transmitter and receiver of such a transceiver can share elements such as an antenna, a baseband processor, and MAC layer circuits. The explanations made above are intended to describe a transceiver that both transmits unbiased training sequences and calculates unbiased channel estimates based on the reception of unbiased training sequences from other transceivers in the device network.

Functional Description

Modern communication systems with high-speed data transmission transmit signals through two separate channels, in-phase and quadrature-phase channels (I and Q). Two channels form a two-dimensional (2D) constellation in the complex plane. QPSK and QAM are examples of groups. Channels I and Q can be carried by RF hardware, which cannot be fully balanced due to variations in RF components, which result in in-phase quadrature imbalance. In increasingly common direct conversion systems, the resulting imbalance is even greater. IQ imbalance distorts the constellation and causes crosstalk between channels I and Q: the signal interferes with itself. An increase in transmit power does not help, since self-generated interference is increased along with the signal power. The signal-to-noise ratio (SINR) reaches an upper limit, which imposes a limit on the highest data rate achieved with the installed RF hardware. An expensive way to solve the problem of increasing the data transfer rate is to use more complex and expensive hardware. A possible less expensive solution is to digitally evaluate and compensate for IQ imbalance. The concepts of digital estimation and compensation algorithms were developed earlier in the art. However, solutions tend to be costly since they do not rely on a special type of training sequence. These decisions often take into account imbalance on only one side, usually in the receiver.

The following are examples that focus on Orthogonal Frequency Division Multiplexing (OFDM) with insights into time-domain systems that consider end-to-end imbalance, from transmitter to receiver. Moreover, in OFDM, imbalance is modeled as a function of frequency, taking into account the spread in the frequency response of the filters.

Two types of enhancements are presented: one at zero cost, which estimates the interference from channel estimation by using an unbiased training sequence. Significant benefits are achieved, as the channel estimation error is often more detrimental to performance than the error in the data itself. The second, at relatively low cost, enhancement compensates for data distortion if a large gain is needed.

The IQ imbalance model is shown below. An analysis is proposed to show how a traditional channel estimate using training sequences can suppress part of the IQ imbalance. Then a direct extension is proposed to calculate the IQ-imbalance parameters, proving that the algorithms are efficient. Using the estimated parameters, a simple compensation algorithm is presented to reduce data distortion. The results of simulation for UWB WiMedia communications, as well as proposals for amending the standard are also presented.

IQ Imbalance Model

IQ imbalance occurs when the (amplitude) power balance or orthogonality (phase) between in-phase (I) and quadrature (Q) channels is not maintained. The IQ imbalance is therefore characterized by an amplitude imbalance of 2ε and a phase imbalance of 2Δφ.

Time Domain Signals

The complex symbol x is transmitted and received through channels I and Q. In an ideal channel without interference, the symbol x is received intact. But in the presence of IQ imbalance, a noisy or distorted version is likely to be adopted.

Y = αx + βx ^* (one)

Where

A = cos (Δφ) + jεsin (Δφ),

B = εcos (Δφ) -j sin (Δφ) (2)

are complex quantities that simulate imbalance, α≈1, and β≈0. Non-linear model ( 1 ) is linearized by means of a vector form

→ Y = BX (3)

B is the unbalance matrix. The second line is rudimentary because it is a copied version of the first line. But it gives the same dimension and type of input and output, so that the unbalance blocks in the transmitter and receiver can be concatenated, as described below. The imbalance matrix in the transmitter is determined by B _t , and in the receiver it is determined by B _r .

Single tap channel

A single tap channel suitable for OFDM is considered. The single tap channel h in a suitable matrix form is defined as

H =

(four)

With an imbalance in the transmitter and receiver and with normal white Gaussian noise (AWGN) n in the vector form N = ( n n ^* ) ^{T, the} received signal is expressed as a combination of linear blocks

Y = B _r HB _t X + N

H'X + N

H'X + N

→ y = h'x + β'x ^* + n (5)

The overall result is that the IQ imbalance and the channel are combined to create a global channel h ', plus unwanted distortion or interference characterized by a global unbalance parameter β'. The global imbalance parameter β ′ changes when the channel changes, and may need to be regularly evaluated.

Then, a state is considered where the symbol x, instead of covering the entire complex plane, is limited by a given (one-dimensional, 1D) axis. For example, an axis may be associated with BPSK modulation, a real axis, an imaginary axis, or any axis located between them. In this case, we can write x ^* = kx, where k is the complex constant (rotation), and

Y = (h '+ β'k) x + n

h"x + n

h "x + n (6)

If x is bounded by a single axis, IQ imbalance tends to zero, becoming an integral part of the overall channel response.

Frequency domain signals

Despite the fact that the previous model is applied to signals of the time domain, a modification is considered below, where the signal of interest x is specified in the frequency domain, at a frequency f. In the time domain, this signal is carried by the complex tone xe ^j2πft . Replacing the expressions in equation ( 1 ), we obtain:

αxe ^j2πft + ^{βx *} e ^-j2πft (7)

In OFDM, interference caused by IQ imbalance is not detected at the same frequency f, but rather at a mirror frequency -f and vice versa. What is transmitted at a frequency of -f interferes at a frequency of + f. If the signal x _m is the signal transmitted at the frequency -f, where the index m denotes the parameter at the mirror frequency -f, then at the frequency -f the following

α _m x _m e ^-j2πft + β _m x _m ^* e ^j2πft (8)

A generalization of equations in the time domain was used. The α and β parameters of the IQ imbalance here are a function of frequency. This models the imbalance due to different low-pass filters (main band) or band-pass (IF) filters in the system. The paths I and Q cannot have exactly the same matching filters, and from here the imbalance changes depending on the frequency. In time-domain systems, this kind of imbalance exists, but it is very costly to compensate. Compensator and model expansion are required to deal with different convolutions in different channels. Therefore, in the time domain, volume or medium imbalance is used. Frequency-domain systems are able to take advantage of the conventional compensator design and model imbalance on a frequency basis.

If the output of equations ( 7 ) and ( 8 ) is combined for each subcarrier, the following is observed

Y = (αx + β _m x _m ^* ) e ^j2πft

y _m = (α _m x _m + βx ^* ) e ^-j2πft (9)

By omitting subcarriers (automatically processed by FFT), the linear function of the signal model at + f and -f can be written as

→ Y = BX (10)

In the frequency domain model, the second row is no longer rudimentary. The model, in a single view, deals with a pair of mirror frequencies. The single tap channel h at frequency f and h _m at frequency -f are modeled by the matrix

.H =

. (eleven)

)^Т шума. Сквозной моделью являетсяThe noise n AWGN at the frequency f and n _m at the frequency -f form the vector N = (

) ^T noise. The cross-cutting model is

Y = B _r HB _t X + N

H'X + N

H'X + N

→ y = h'x + β _m 'x _m ^* + n

y _m = h _m 'x _m + β'x ^* + n _m (12)

h ', h _m ' are global channel taps, and β ', β _m ' are global unbalance parameters. The unbalance parameters change when the channels change, and may need regular evaluation.

Since IQ imbalance only interferes with the specular frequency, two interesting cases are noteworthy. If no signals are transmitted at the specular frequency or the channel reaches the highest attenuation, no interference will occur. If, on the other hand, the signal or channel is powerful, the interference can be significant. Hence, in OFDM, the effect of IQ imbalance is more problematic.

Traditional channel rating

Before considering compensation algorithms, it is shown how half the problem can be solved without cost, simply by using an unbiased training sequence. An unbiased training sequence completely eliminates interference from channel estimation, significantly improving performance. In fact, the error in channel estimation is often more pernicious than the error in data, since channel estimation tends to create bias in the group.

Model ( 12 ) is stimulated by pilot signals. At frequency + f, pilot signal p is transmitted, and at frequency-f, pilot signal p _m . Provided that, without loss of generalization, the pilot signals have a unit norm (the channel carries effective power), the traditional channel estimate at frequency f is obtained by the rotation correction by p ^*

= h'pp^* + β'_mp_m ^*p^* + n = h' + β'_mp_m ^*p^* + n

= h'pp ^* + β ' _m p _m ^* p ^* + n = h' + β ' _m p _m ^* p ^* + n (13)

By averaging several channel measurements, the noise is automatically reduced (for clarity, noise rotation correction is omitted). For the β ' _m p _m ^* p ^* member, many OFDM systems (such as WiMedia UWB communication) use a training sequence, which is just a repeated character. Therefore, this term does not fade during averaging. Applying scrambling +1 or -1 to the OFDM symbol taken as a whole does not help, since nothing changes when the sign of both p ^* and p _m ^{* is inverted} . Rather, the following is achieved: after accumulating a certain number of measurements, the sum of the works is reduced to zero

Σ _i p _i p _im = 0 (fourteen)

Often the training sequence consists of the same number of characters, and it is enough to guarantee that each pair is summed to zero

p ₁ p _1m + p ₂ p _2m = 0 (fifteen)

Table 1
Examples of unbiased training sequences P ₂ = jp ₁ The second training symbol is a 90-degree rotation of the first training symbol P ₂ = p ₁ , p _2m = - p _1m For positive frequencies, a constant pilot signal is maintained; for negative frequencies, the sign is constantly inverted

Examples of simple sequences that satisfy the condition are shown in table 1 . These types of training sequences are indicated as unbiased training sequences since, on the one hand, unbiased channel estimates are generated, and on the other hand, training signals equally cover the I and Q dimensions of the complex plane in the time domain. For example, an unbiased training sequence does not concentrate only along the material axis.

a_i exp(-jφ)|² = 0,5 L + 0,5

exp(-2jφ) Σa_ia_i. Эта мощность постоянна в любом направлении φ тогда и только тогда, когда Σ _ia_ia_i ≡ Σ _ip_ip_im = 0. Достигается равномерное покрытие комплексной плоскости.As proof: we consider the complex scalar a _i = p _i e ^jθ = p _im e ^-jθ with the unit norm, located at an equal distance between the points p _i and p _im . In the time domain, the pilot signals are summed in 2a _i cos (2πft + θ). In the time domain and for a given OFDM symbol, 2 mirrored pilot signals overlap the only direction defined by the complex constant a _i . If L characters are transmitted, the total (either average or accumulated) power in the direction φ is Σ _i |

a _i exp (-jφ) | ² = 0.5 L + 0.5

exp (-2jφ) Σa _i a _i . This power is constant in any direction φ if and only if Σ _i a _i a _i ≡ Σ _i p _i p _im = 0. A uniform coverage of the complex plane is achieved.

Estimation of in-phase quadrature discrepancy

After evaluating the global channel h ′, the estimation of the global imbalance parameter β _m ′ is considered. A thorough analysis of equation ( 12 ) shows that this parameter can be obtained in some way, directly similar to the method for traditional channel estimation. That is, β _m ′ may be treated like a “channel” carrying a pilot signal p _m ^* . From here, the discrepancy estimate can be obtained by correcting the rotation by p _m . The condition for the unbiased unbalance estimate is identical to equation ( 14 ).

To summarize the above, using unbiased training sequences and two traditional channel estimates, good estimates of the end-to-end channel and unbalance parameter are obtained (table 2 ).

Table 2: Evaluation Algorithm H ' β ' _m Correct rotation with p ^* Correct rotation with p _m

Adjacent Subcarrier Smoothing

In addition to averaging over adjacent OFDM symbols, the channel estimate may be smoothed over adjacent subcarriers within the same symbol. In OFDM, the cyclic prefix is made short, and it is assumed that the channel should slowly change from tone to tone. Similarly, filters in an RF circuit must have a short time response, and their frequency response also changes slowly, i.e., the IQ imbalance slowly changes in the subcarriers. The same channel smoothing technologies can be used to smooth and improve the estimation of the unbalance parameter. When using unbiased training sequences, there is no relationship between the channel estimate and the unbalance estimate. Each score can be smoothed independently.

If a unique OFDM symbol is used for evaluation, an unbiased training sequence can be found that satisfies equation ( 14 ). In this case, an almost unbiased training sequence can be obtained by applying the summation from equation ( 14 ) to groups of 2 or more adjacent subcarriers. In this case, anti-aliasing automatically dampens all or part of the interference from the mirror frequencies. One solution is to rotate the pilot signal 90 degrees on an adjacent subcarrier (shifting in mirror directions at positive and negative frequencies).

The effect of smoothing the channel estimate by adjacent frequencies can be performed using convolution with a weight vector, rather than simply averaging each pilot signal independently of each other. The same smoothing effect also applies to parameter B of the IQ imbalance.

From the channel estimation equations, it can be seen that after averaging over multiple periods of symbols (e.g., OFDM symbols), the channel estimate (h ^' ) is equal to the original channel h plus some undesirable terms:

h '= h + (1 / P) B _m sum {p _i ^* p _im ^* } + n

where P is the total pilot power, usually assumed to be constant from one symbol to the next. But the value of P should not be constant in the general case, “n” is noise, and p _i and p _im are the pilot signal of interest at the frequency f and the mirror pilot signal at the frequency -f, respectively.

The above equations assume that averaging occurs for each pair of pilot signal and mirror pilot signal, with index i and mirror index im being independent of other pilot signals (e.g., index i 'and i'm). That is, each pair can be processed separately by collecting the observed values for each pair over multiple symbol periods and outputting an average.

However, if adjacent pilots are averaged, a new channel average is obtained, as follows:

h '= h + sum {w _ji B _im p _i ^* p _im ^* } + n,

where the index i now denotes adjacent subcarriers rather than different periods of characters. Index j is the central pilot signal, and w _ji is the weight function. In this case, the formula w _ji B _im p _i ^* p _im ^* can be written as a convolution:

w * (B _m p ^* p _m ^* ) = 0.

Note: the symbol “ * ” indicates a convolution, while “ ^* ” (superscript) indicates complex conjugation.

The weight function w is known and depends on how the channel is averaged. The pilot signals p and p _m are unknown. By optimizing the pilot signals, the value of the equation, which is the unwanted noise added to h, can be minimized unbalance parameters B _m . The parameter B _m is unknown in this equation and depends on physical hardware (IQ imbalance). But since B _{m is} unknown at the time when p and p _{m are} optimized, and if it is assumed that averaging over adjacent frequencies means that B _m does not change significantly, it can be assumed that B _m is a constant independent of index i. From here, the term B _m can be excluded from the equation to obtain

w * (p ^* p _m ^* ) = 0,

whose power is minimized for a given waveform w. For example, by limiting the values of p and p _{m, the} power can remain constant, while the index i changes.

In the absence of restrictions (other than the total power of the pilot signal), this task is not too difficult to solve. The convolution can be written in the form of the matrix W Toeplitz formed by the vector w, which is synchronized with the vector A = (p ^* p _m ^* ). To minimize the power of convolution WA, i.e. the norm of the vector, optimally:

min A ^H W ^H WA.

Note: “A” contains dependent members. That is, A (i) and A (im) are equal members. This means that only half of the “A” is known, which creates a slightly more difficult problem to solve. But if w is symmetric about the origin, then the matrix and the vector can be mirrored. What remains is a matrix with a size reduced by half, and the vector W and A '. In this case, the optimal convolution WA is:

min A ^'H W ^{' H} W'A '

provided that the total pilot power is constant. This solution is still not simple, if we do not assume that the total power of the pilot signal is approximately equivalent, so that the norm at A 'is constant. Based on this new assumption, in this case, the solution to the problem is a smaller eigenvector of the matrix W ^'H W'.

Under the many assumptions mentioned above, the problem is now easily solvable. The solution gives an idea of the shape of the pilot signals. However, one of the assumptions is very often invalid. Typically, pilot signals have a very simple construction, such as 1 + j, or 1, or j, etc. Otherwise, the pilot signals have at least a constant rate, hence exp (jθ).

Subject to this new limitation, the problem can be solved using optimization technologies. Since unbiased training signals with an error of 25% or even 50% are useful, excellent performance can be achieved. From here it is enough to find a simple form of pilot signals that tend to be unbiased, being 100% unbiased, which is actually difficult to achieve.

This form can be obtained by assuming that w is slowly changing. Therefore, every 2 adjacent pilots can be merged and the task for these two pilots can be optimized separately under the condition of constant w. In this case, the general shape of the pilot signals can be derived. Combining every two adjacent pilot signals together, averaging them as if there is no weighting w (or there is a constant weight), returns to the original formula for a frequency-smoothed unbiased training signal:

p ₁ p _1m + p ₂ p _2m = 0,

where p ₁ and p ₂ are rather adjacent in frequency than falling into consecutive periods of characters. One simple FSTS involves rotating the pilot signals 90 degrees from one pilot signal to the next in frequency with a mirror-like movement from the center frequency to the cutoff frequency. Alternatively, the sign of the pilot signals of the negative (mirror) frequency can be exchanged along with the same sign for the pilot signals of the positive (reference) frequency (or vice versa). Other combinations are possible that achieve good scrambling.

Simulations for WiMedia UWB communications using all positive (reference frequency) pilot signals set to 1, with the inverted sign of each other negative (mirror) frequency pilot signal (e.g. +1, -1, +1, -1, and etc.) show excellent results, comparable to a time-offset training sequence. Frequency smoothing is performed by the w = sine function with relatively slow changes from one pilot signal to another pilot signal. This weighting function is prescribed by other requirements (such as improving channel estimation even in the absence of IQ imbalance).

Rating

The use of unbiased training sequences and the results of the aforementioned traditional channel estimation are the least squares (LS) estimation. Of all the LS estimates, the perception of minimum mean square error (MMSE) is significant.

Least Squares Estimation

L gears X _i , L noise terms N _i and L observations Y _i can be respectively combined into matrices 2 on L

(16)

In this case, equation ( 12 ) becomes

Y = H ' X + N (17)

Unknown is H '. LS score is

(eighteen)

When condition ( 14 ) is satisfied, it is easy to verify that XX ^H is diagonal (the cross terms tend to zero). It is proportional to the unit matrix, since the pilot signals are normalized to the unit rate. In this case

(19)

are strictly four traditional channel estimates with rotation corrections, respectively, by p _i ^* , p _im , p _im ^* and p _i , as described in the previous section. Two estimates were obtained for frequency f and two estimates for specular frequency -f.

Optimal score

также является оценкой LS. Ниже показано, что использование несмещенных настроечных последовательностей дает в результате отличную оценку. Модель (17) может рассматриваться в качестве неизвестной информации H', отправленной посредством 2 следующих друг за другом передач через 2 вектора (строки Х) в пространстве размерности L. Обозначим посредством X _j, N _j и Y _j соответственно строку j у X, N и Y, где j∈{1,2}. Модели (12) и (17) могут быть записаныUnbiased training sequences and traditional channel estimates are an LS estimate. But any assessment

also an LS score. It is shown below that the use of unbiased training sequences results in an excellent score. Model ( 17 ) can be considered as unknown information H 'sent by 2 successive transmissions through 2 vectors (rows X) in a space of dimension L. Denote by X _j , N _j and Y _j respectively the row j of X , N and Y , where j∈ {1,2}. Models ( 12 ) and ( 17 ) can be written

Y ₁ = h'X ₁ + β ' _m X ₂ + N ₁ Y ₂ = β'X ₁ + h ' _m X ₂ + N ₂ (twenty)

There are 2 transmissions, each including 2 vectors X ₁ and X ₂ , and where each vector carries information about the complex amplitude, which must be estimated. The LS estimate consists of projecting onto each vector parallel to another vector in order to neutralize the interference. A very good result is obtained when 2 vectors are orthogonal, that is, when the scalar product ( 14 ) is equal to zero. Unbiased training sequences are, by definition, training sequences that verify this condition. Other sequences use non-orthogonal vectors and suffer the loss of the characteristic function of the angle between the vectors X ₁ and X ₂ . Many OFDM systems currently use a very poor variety of training sequences, where X ₁ , X _{2 are} collinear, and it is not possible to properly evaluate 4 elements in H '. These training sequences tend to evaluate the noisier channel options h 'and h' _m .

Это матрица 2 на 2, то есть 4 значения ошибок. Каждое значение может быть изолировано умножением слева и справа комбинацией векторов (1 0)^T и (0 1)^T. При условии, что ENN^H - единичная матрица или, в более общем смысле, диагональная матрица с элементами σ² и σ_m ², может быть показано, что MSE у

и

соответственно являются первым и вторым диагональными элементами у σ²(XX^H)^-1. А для

и

MSE соответственно являются первым и вторым диагональным элементом у σ_m ²(XX^H)^-1.To calculate the mean square errors (MSE), the estimation error is

This is a 2 by 2 matrix, i.e. 4 error values. Each value can be isolated by multiplying the left and right vectors with a combination of (1 0) ^T and (0 1) ^T. Provided that ENN ^H is the identity matrix or, in a more general sense, a diagonal matrix with elements σ ² and σ _m ² , it can be shown that MSE

and

respectively, are the first and second diagonal elements of σ ² (XX ^H ) ^-1 . And for

and

MSE are respectively the first and second diagonal element in σ _m ² (XX ^H ) ^-1 .

The total MSE is 2 (σ ² + σ _m ² ) tr (XX ^H ) ^-1 . Further, the task is to find X that minimizes tr (XX ^H ) ^-1 , provided that the total pilot power is constant, that is, tr (XX ^H ) = 2L. Using the expansion in eigenvalues, the problem can be written as the problem of minimizing Σ1 / λ _j provided that Σλ _j is constant. The problem is solved by uncertain Lagrange multipliers and is usually optimal when all eigenvalues are equal. This means that XX ^H = LI is proportional to the identity matrix.

The total MSE has been minimized, and the resulting MSE for each element is σ ² / L or σ _m ² / L. But this MSE for each element is likely to be the best that can be obtained even if a single vector transmission is used. It is unlikely that the MSE will be improved for 2 vector transmissions, because the MSE for each element has been minimized. Unbiased training sequences plus traditional channel rating are MMSEs of all LS ratings.

IQ imbalance compensation

If the benefit from an unbiased channel estimate is insufficient, IQ imbalance parameters can be estimated (as described previously) and applied to compensate for data distortion. H 'is estimated in model ( 12 ), Y = H'X + N. Next, attention is drawn to the unknown X data. The model is the same as for any 2-tap channel with cross-correlations. Any channel compensation algorithm can be adapted. A simple compensation algorithm is presented, suitable for ubiquitous coded QAM channels and bit-wise fading.

One of the problems with the ZF approach, H ' ^-l Y = X + H' ^-1 N, is that it amplifies noise when the mirror channel is weak, if complex colored noise is not taken into account. The presented solution uses ZF, but only when the mirror channel is not weak. In equation ( 12 ), by replacing x _{m with} its value, we obtain the following

y = (h'-β _m 'β' ^* / h _m ' ^* ) x + (β _m ' / h _m ' ^* ) y _m ^* - (β _m ' / h _m ' ^* ) n _m ^* + n
≈ h'x + (β _m ' / h _m ' ^* ) y _m ^* + n '+ n (21)

-(β_m'/h_m'^*)n_m ^* - усиление шума. Примечание: предполагается, что член разбаланса второго порядка β'^*β_m' << h'h_m'^*. Когда это приближение недействительно, рассматривается скорректированный канал h'_c

h'- β_m'β'^*/h_m'^*, который влечет за собой точную оценку канала и параметров разбаланса. В основном технология ZF состоит из вычисленияwhere n '

- (β _m '/ h _m ' ^* ) n _m ^* - noise amplification. Note: it is assumed that the second-order unbalance member is β ' ^* β _m '<<h'h _m ' ^* . When this approximation is not valid, the corrected channel h ' _{c is considered}

h'- β _m 'β' ^* / h _m ' ^* , which entails an accurate estimate of the channel and unbalance parameters. Basically, ZF technology consists of computing

z = y- (β _m '/ h _m ' ^* ) y _m ^* ≈h'x + n '+ n (22)

Subtracting the value (β _m '/ h _m ') y _{m of the} mirror frequency from the received signal y yields a simple channel model without IQ imbalance. The rest of the decoding chain is unchanged.

This solution works well as long as the noise gain is weaker than the original interference from the IQ imbalance, that is, | n '| ² <| β _m 'x _m ^* | ² . If not, it is likely to use the original y rather than the unbalance corrected z. In order to make a decision, it is not necessary to evaluate n '. An error-resistant improvement in the form of averaging can be selected. Thus, expected values are taken into account

E | n '| ² = (| β _m '| ² / | h _m ' | ² ) E | n _m | ² <| β _m '| ² E | x _m ^* | ²

SNR_m>1→ | h _m '| ²

SNR _m > 1 (23)

When the ratio of the SNR _{m of} the mirror-frequency signal to noise is greater than 1, the corrected unbalance term z is used. Otherwise, the original signal y is stored. Due to inaccurate channel estimation and imbalance, it is safer to use a larger SNR, for example, SNR _m > 2 works well for WiMedia UWB communication. Note that SNR _m can usually be obtained from the global SNR using the formula SNR _m = | h _m '| ² SNR.

Table 3 summarizes the ZF algorithm with the exception of noise amplification.

Table 3
Compensation Algorithm SNR _m <1 + δ SNR _m > 1 + δ z = y z = y- (β _m '/ h _m ') y _m

Simulation results

10 illustrates the performance achieved by applying unbiased training signal algorithms to the WiMedia UWB standard. The highest data transfer rate, 480 Mbps, is modeled in the CM2 model of the IEEE 802.15.3 channel (picomedia in the room is about 4 meters). Shielding and hopping are disabled. The IQ imbalance is constant and equal to 2ε = 10% (0.8 dB) in amplitude and 2Δφ = 10 degrees in phase. The same amount of unbalance is present in the transmitter and receiver. The drawing shows the packet error rate (PER) as an Eb / No function. Performance deteriorates rapidly without any form of compensation. Table 4 lists the disadvantages of various algorithms with respect to the ideal case.

Table 4
WiMedia UWB: IQ Imbalance Loss at PER of 10 ^-2 Current standard Unbiased setting Compensation 3.1 dB 1.1 dB 0.35 dB

End-to-end IQ imbalance and channel combine to form a global 2 by 2 channel matrix. Using unbiased training sequences achieves significant benefits without cost. Unbiased training sequences automatically neutralize end-to-end self-generated interference from the channel estimate. Moreover, such training sequences are ideal for estimating IQ-imbalance parameters, and a simple algorithm for compensating for data distortion is presented: forcing zero with the exception of noise amplification.

WiMedia UWB communications, in particular, benefit from the following enhancements: the traditional offset training sequence, which consists of 6 characters transmitted exclusively on channel I, can be divided into 2 halves to create an unbiased sequence. The first 3 characters are sent on channel I, and the last 3 characters are sent on channel Q. By uniformly covering the complex plane, an unbiased training sequence is created with great benefits for high data rates. For the sake of backward compatibility, this circuit can be designed for high-speed data transmission modes and can be signaled by beacon signals or the type of training sequence can be blindly detected.

In OFDMA (e.g., WiMAX), the subcarriers f and -f can be assigned to different users. Significant interference can occur if the power control brings one user to a high power level. Therefore, it is a good idea to arrange the pilot signals of different users on the mirror subcarriers. Pilot signals must meet the criteria for an unbiased training sequence. Each user automatically benefits without additional effort. Pilot signals can hop-hop to other locations while maintaining mirroring.

Time domain formulas can be expanded to code division multiple access (CDMA) using a multi-tap equalizer combining multiple single tap channels. Unbiased training sequences automatically improve channel estimation for each tap. A simple unbiased training sequence for CDMA consists of continuously rotated 90 degrees complex characters.

The theory can be extended to other time-domain systems (e.g., TDMA) other than CDMA. The channel estimate is obtained by convolution of the received signal with a matched filter, which is a mirror version of the complex interface FSTS. In other words, ignoring AWGN, it is a convolution of the channel transmitted by the FSTS and the matched filter. It can be shown that the channel estimate contains the term (bias) of the intrinsic interference formed by the IQ imbalance. Considering the equation in the frequency domain, the term of intrinsic interference can be made almost tending to zero if the FSTS is chosen carefully. Indeed, through the use of the above-described FSTSs, the interference itself tends to disappear after summing the values from adjacent tones (subject to slow channel changes). Hence, an unbiased training signal for time-domain systems can be designed with frequency-domain restrictions on adjacent tones.

In fact, GSM training sequences were studied and it was found that some training sequences and certain rotation angles of the MSK modulation (minimal shift manipulation) behave better than others. There is no simple interpretation of why there is a difference in performance. As mentioned above for time-domain systems, the difference in performance can be due to good (unbiased) versus poor coverage of the complex plane in the time domain by each group of adjacent subcarriers and their mirror subcarriers.

11 is a flowchart illustrating a method of supplying a frequency-smoothed communication tuning signal. Although the method is illustrated for clarity as a sequence of numbered steps, numbering does not necessarily dictate the order of steps. It should be understood that some of these steps may be skipped, performed in parallel, or performed without requiring a strict ordering. As used in the materials of this application, the terms “generating,” “outputting,” and “multiplying” indicate a sequence of operations that may be involved through the use of machine-readable software instructions, hardware, or a combination of software and hardware. The method begins at block 1100.

Step 1102 generates a frequency-smoothed unbiased training signal in a quadrature modulation transmitter. The frequency-smoothed unbiased training signal includes a plurality of pilot signal products, where each pilot signal product includes complex plane information represented by a reference frequency subcarrier multiplied by complex plane information represented by a mirror frequency subcarrier. The sum of the plurality of pilot signal products is zero. Step 1104 provides a frequency-smoothed unbiased training signal within a single symbol period. In one aspect, FSTS components can be issued sequentially or in batch mode and stored until the complete FSTS is assembled. In this aspect, the next step (not shown) would transmit the assembled FSTS, which is a single symbol period.

In one aspect, a step is provided in which a frequency-smoothed unbiased training signal is generated in step 1102, including a plurality of adjacent reference frequency frequency subcarriers and a plurality of adjacent mirror frequency subcarriers. FSTS may consist of 2 or more pilot products. In yet another aspect, step 1102 generates a frequency-smoothed unbiased training signal that includes a group of adjacent reference frequency sub-carriers without intermediate sub-carriers and a plurality of adjacent mirror frequency sub-carriers without intermediate sub-carriers.

In one aspect, a frequency-smoothed unbiased training signal may be represented as described below, as illustrated in step 1102a:

Σ p _i p _im = 0, i = from 1 to n;

where p is the reference frequency subcarrier,

where p _m is the mirror frequency subcarrier and

where n is equal to the number of pilot signal products.

In another aspect, a frequency-smoothed unbiased training signal may be represented as set forth below, as illustrated in step 1102b:

Σ w _i p _i p _im = 0, i = from 1 to n;

where w is the weight coefficient.

As an example, a frequency-smoothed unbiased training signal may consist of a first product of pilot signals with a reference subcarrier at a frequency + f representing information as the first value of the complex plane and a mirror subcarrier at a frequency -f representing the first value of the complex plane. FSTS, in addition, includes a second product of pilot signals with a reference subcarrier at a frequency (f + 1), an adjacent frequency + f representing the first value of the complex plane, and a mirror subcarrier at a frequency of (f + 1), an adjacent frequency of f representing the first value of the complex plane +180 degrees.

As another example, a frequency-smoothed unbiased training signal may be generated as a first product of pilot signals with a reference subcarrier at a frequency + f representing information as the first value of the complex plane and a mirror subcarrier at the frequency -f representing the first value of the complex the plane. In this case, the FSTS also includes a second product of the pilot signals with a reference subcarrier at a frequency (f + 1), an adjacent frequency + f representing the first value of the complex plane +90 degrees, and a mirror subcarrier at a frequency of (f + 1), adjacent frequency -f, representing the first value of the complex plane -90 degrees.

In another aspect, generating an unbiased frequency-smoothed training signal in step 1102 includes generating P products of pilot signals. In this aspect, step 1103 generates (N-P) communication data symbols. Typically, step 1104 provides FSTS and communication data symbols and constitutes a single symbol period. Step 1106 transmits N subcarriers in one symbol period, including a frequency-smoothed unbiased training signal and quadrature modulated communication data.

Alternatively, step 1102 generates a frequency-smoothed unbiased training signal using a group of reference frequency subcarriers and corresponding mirror frequency subcarriers, and step 1104 provides a frequency-smoothed training signal in a first symbol period. Step 1108 generates quadrature modulated communication data on a group of reference frequency subcarriers and corresponding mirror frequency subcarriers. Step 1110 provides quadrature modulated communication data in a second symbol period following the first symbol period.

The above-described flowchart may also be interpreted as an expression of a computer-readable medium having instructions stored thereon for a frequency-smoothed communication signal. Teams would correspond to steps 1100 through 1110, as explained above.

12 is a flowchart illustrating a method of calculating a channel estimate using a frequency-smoothed unbiased training signal. The method begins at step 1200. Step 1202 receives a frequency-smoothed unbiased training sequence at a quadrature modulation receiver. The frequency-smoothed unbiased training sequence includes a plurality of pilot signal products, where each pilot signal product includes predetermined complex plane information (p) represented by a reference frequency subcarrier (f) multiplied by predetermined complex plane information (p _m ) represented by the mirror frequency subcarrier (-f). The sum of the plurality of pilot signal products is zero. Step 1204 processes a frequency-smoothed unbiased training signal, generating a plurality of processed symbols (y) representing complex plane information. Step 1206 multiplies each processed symbol (y) by the conjugate value of the corresponding reference signal (p ^* ). Step 1208 obtains a frequency-smoothed channel estimate (h).

Note: processed characters (y) are associated with a reference subcarrier. Alternatively, block 1204 may process a frequency-smoothed unbiased training signal to generate a plurality of processed symbols (y _m ) representing complex plane information associated with the mirror subcarrier. In this case, step 1206 multiplies each processed symbol (y _m ) by the conjugate value of the corresponding reference signal (p _m ^* ), and step 1208 obtains a frequency-smoothed channel estimate (h _m ) associated with the mirror subcarrier. As another alternative, the above steps find both channel estimates, (h) and (h _m ).

In one aspect, block 1202 receives a plurality of contiguous reference frequency subcarriers and a plurality of contiguous mirror frequency subcarriers. FSTS may consist of 2 or more pilot products. In yet another aspect, step 1202 receives a frequency-smoothed unbiased training signal including a group of adjacent reference frequency frequency subcarriers without intermediate subcarriers and a plurality of adjacent mirror frequency frequency subcarriers without intermediate subcarriers.

In one aspect, a frequency-smoothed unbiased training signal may be represented as follows:

Σ p _i p _im = 0, i = from 1 to n;

where p is the reference frequency subcarrier,

where p _m is the mirror frequency subcarrier and

where n is equal to the number of pilot signal products.

In another aspect, a frequency-smoothed unbiased training signal may be represented as follows:

Σ w _i p _i p _im = 0, i = from 1 to n;

where w is the weight coefficient.

As in the FSTS examples of the transmitter, a frequency-smoothed unbiased training signal may consist of a first product of pilot signals with a reference subcarrier at a frequency + f representing information as the first value of the complex plane and a mirror subcarrier at a frequency -f representing the first value of the complex plane . FSTS, in addition, includes a second product of pilot signals with a reference subcarrier at a frequency (f + 1), an adjacent frequency + f representing the first value of the complex plane, and a mirror subcarrier at a frequency of (f + 1), an adjacent frequency of f representing the first value of the complex plane +180 degrees.

As another example, a frequency-smoothed unbiased training signal may be generated as a first product of pilot signals with a reference subcarrier at a frequency + f representing information as the first value of the complex plane and a mirror subcarrier at the frequency -f representing the first value of the complex the plane. Then, the FSTS also includes a second product of the pilot signals with a reference subcarrier at a frequency (f + 1), an adjacent frequency + f representing the first value of the complex plane +90 degrees, and a mirror subcarrier at a frequency of (f + 1), an adjacent frequency -f, representing the first value of the complex plane of -90 degrees.

In one embodiment, step 1202 receives the unbiased frequency-smoothed training signal as P products of pilot signals in a symbol period. Then, step 1203 receives (N-P) symbols of communication data in the (same) symbol period.

Alternatively, step 1202 receives a frequency-smoothed unbiased training signal with a group of reference frequency subcarriers and corresponding mirror frequency subcarriers. Then, step 1210 receives quadrature modulated communication data on a group of reference frequency subcarriers and corresponding mirror frequency subcarriers following receipt of a frequency-smoothed unbiased training signal.

The above-described flowchart may also be interpreted as an expression of a computer-readable medium having instructions stored therein for calculating a channel estimate using FSTS. The instructions would correspond to steps 1200 through 1210, as described above.

Above are systems, methods, devices, and processors for enabling FSTS transmission and using FSTS in calculating receiver channel estimates. Examples of specific protocols and communication formats are provided to illustrate the invention. However, the invention is not limited to these examples only. Other varieties and embodiments of the invention will be apparent to those skilled in the art.

Claims (58)

1. A method for transmitting a tuning signal, comprising the steps of generating a tuning signal in a quadrature modulation transmitter, wherein the tuning signal is a plurality of reference complex values that must be respectively transmitted on a plurality of reference frequency subcarriers during a single symbol period, and a plurality of corresponding mirror complex values to be respectively transmitted on a plurality of mirror frequency subcarriers during a single symbol period, n The sum of the products of the reference complex values and the corresponding mirror complex values is zero; and
transmit a training signal within a single symbol period. 2. The method according to claim 1, in which the step of generating the tuning signal includes generating a tuning signal representing a plurality of reference complex values that must be respectively transmitted on a plurality of adjacent reference frequency subcarriers and a plurality of corresponding mirror complex values which should be respectively transmitted on a plurality of adjacent corresponding mirror frequency subcarriers. 3. The method according to claim 1, in which the step of generating a tuning signal includes generating a tuning signal representing a plurality of reference complex values to be respectively transmitted on a plurality of adjacent reference frequency subcarriers without intermediate subcarriers, and a plurality of corresponding mirror complex values to be respectively transmitted on a plurality of adjacent corresponding mirror frequency subcarriers without intermediate subcarriers. 4. The method according to claim 1, in which the stage of formation of the training signal includes a step on which to form the training signal as follows:
Σp _i p _im = 0, i = 1 ... n;
where p is the reference complex value to be transmitted on the reference frequency subcarrier;
p _m is the mirror complex value to be transmitted on the corresponding mirror frequency subcarrier; and
n is equal to the number of reference frequency subcarriers. 5. The method according to claim 1, in which the stage of formation of the training signal includes a step on which to form the training signal as follows:
Σw _i p _i p _im = 0, i = 1 ... n;
where p is the reference complex value to be transmitted on the reference frequency subcarrier;
p _m is the mirror complex value to be transmitted on the corresponding mirror frequency subcarrier;
n is equal to the number of reference frequency subcarriers and
w is the weight coefficient. 6. The method according to claim 1, in which the stage of formation of the training signal includes the stage of forming a training signal, which is at least the first reference complex value that must be transmitted on the first reference subcarrier at a frequency + f1, the first corresponding mirror complex value to be transmitted on the first corresponding mirror subcarrier at a frequency of -f1, the second reference complex value to be transmitted on the second reference subcarrier at a frequency + f2, and the second the corresponding mirror complex value to be transmitted on the second corresponding mirror subcarrier at a frequency of-f2. 7. The method according to claim 1, in which the stage of formation of the training signal includes the stage of forming a training signal representing at least the first reference complex value that must be transmitted on the first reference subcarrier at a frequency + f, the first corresponding mirror complex value to be transmitted on the first corresponding mirror subcarrier at a frequency of -f, the second reference complex value to be transmitted on a second reference subcarrier at a frequency (f + 1) adjacent frequency + f, and the second corresponding mirror complex value, which should be transmitted to the second corresponding mirror subcarrier at the frequency - (f + 1), adjacent frequency -f. 8. The method according to claim 1, further comprising stages in which:
form a plurality of communication data symbols and transmit a training signal and communication data symbols in a single symbol period on N subcarriers, where N is the number of communication data symbols added to the number of reference subcarriers. 9. The method according to claim 1, additionally containing stages in which:
generating quadrature modulated communication data to be transmitted on the reference frequency subcarriers and corresponding mirror frequency subcarriers; and
transmit quadrature modulated communication data in a second symbol period following a single symbol period. 10. A method for calculating a channel estimate, comprising the steps of:
receiving a tuning signal in a quadrature demodulation receiver, wherein the tuning signal is a plurality of reference complex values respectively received on a plurality of reference frequency subcarriers during a single symbol period, and a plurality of corresponding mirror complex values respectively received on a plurality of corresponding mirror frequency subcarriers during a single period symbol, and the sum of the products of the reference complex values and the corresponding mirror ompleksnyh values is equal to zero;
generating a plurality of processed symbols (y) representing complex plane information based on the training signal;
each processed symbol (y) is multiplied by the conjugate value of the corresponding reference complex value (p) and a channel estimate (h) is obtained based on the multiplication results. 11. The method according to claim 10, in which the step of receiving the tuning signal includes receiving a tuning signal representing a plurality of reference complex values respectively received on a plurality of adjacent reference frequency subcarriers, and a plurality of corresponding mirror complex values respectively received on a plurality of adjacent mirror frequency subcarriers. 12. The method of claim 10, wherein the receiving signal of the tuning signal includes receiving a tuning signal representing a plurality of reference complex values respectively received on a plurality of adjacent reference frequency subcarriers without intermediate subcarriers, and a plurality of corresponding mirror complex values respectively received on a plurality of adjacent respective mirror frequency subcarriers without intermediate subcarriers. 13. The method according to claim 10, in which the step of receiving a signal includes the stage of receiving the training signal, presented in the form
Σp _i p _im = 0, i = 1 ... n;
where p is the reference complex value received on the reference frequency subcarrier;
p _m is the mirror complex value received on the corresponding mirror frequency subcarrier; and
n is equal to the number of reference frequency subcarriers. 14. The method according to claim 10, in which the step of receiving the training signal includes the stage of receiving the training signal, presented in the form
Σw _i p _i p _im = 0, i = 1 ... n;
where p is the reference complex value received on the reference frequency subcarrier;
p _m is the mirror complex value received on the corresponding mirror frequency subcarrier;
n is equal to the number of reference frequency subcarriers and
w is the weight coefficient. 15. The method according to claim 10, in which the step of receiving the tuning signal includes the stage of receiving the tuning signal, representing at least the first reference complex value received on the first reference subcarrier at a frequency + f1, the first corresponding mirror the complex value received on the first corresponding mirror subcarrier at a frequency of -f1, the second reference complex value received on the second reference subcarrier at a frequency + f2, and the second corresponding mirror complex value, accepted May of the second on the corresponding mirror sub-carrier frequency -f2. 16. The method according to claim 10, in which the step of receiving the tuning signal includes receiving a tuning signal representing at least a first reference complex value received on the first reference subcarrier at a frequency + f, the first corresponding mirror the complex value received on the first corresponding mirror subcarrier at a frequency of -f, the second reference complex value received on the second reference subcarrier at a frequency (f + 1), an adjacent frequency + f, and the second corresponding mirror complex the adopted value on the second corresponding mirror subcarrier at the frequency is (f + 1), the adjacent frequency is -f. 17. The method of claim 10, further comprising receiving a plurality of symbols of communication data in a single symbol period. 18. The method of claim 10, further comprising the step of: receiving quadrature modulated communication data on the reference frequency subcarriers and corresponding mirror frequency subcarriers after receiving the training signal. 19. A system for transmitting a tuning communication signal, comprising:
a signal generator configured to generate a tuning signal representing a plurality of reference complex values that should be respectively transmitted on a plurality of reference frequency subcarriers during a single symbol period, and a plurality of corresponding mirror complex values that should be respectively transmitted on a plurality of corresponding mirror frequency subcarriers during a single symbol period, the sum of the products of the reference complex values and the corresponding the existing mirror complex values is zero; and
a transmitter configured to transmit a training signal within a single symbol period. 20. The system according to claim 19, in which the signal generator generates a tuning signal representing a plurality of reference complex values that must be respectively transmitted on a plurality of adjacent reference frequency subcarriers and a plurality of corresponding mirror complex values that must be transmitted respectively on a plurality of adjacent mirror frequency subcarriers. 21. The system according to claim 19, in which the signal generator generates a tuning signal representing a plurality of reference complex values that should be respectively transmitted on a plurality of adjacent reference frequency subcarriers without intermediate subcarriers, and a plurality of corresponding mirror complex values, which should be respectively transmitted on a plurality of adjacent mirror frequency subcarriers without intermediate subcarriers. 22. The system according to claim 19, in which the signal generator generates a tuning signal as follows:
Σp _i p _im = 0, i = 1 ... n;
where p is the reference complex value to be transmitted on the reference frequency subcarrier;
p _m is the mirror complex value to be transmitted on the corresponding mirror frequency subcarrier; and
n is equal to the number of reference frequency subcarriers. 23. The system according to claim 19, in which the signal generator generates a tuning signal as follows:
Σw _i p _i p _im = 0, i = 1 ... n;
where p is the reference complex value to be transmitted on the reference frequency subcarrier;
p _m is the mirror complex value to be transmitted on the corresponding mirror frequency subcarrier;
n is equal to the number of reference frequency subcarriers and
w is the weight coefficient. 24. The system according to claim 19, in which the signal generator generates a tuning signal representing at least the first reference complex value that must be transmitted on the first reference subcarrier at a frequency + f1, the first corresponding mirror complex value, which should be transmitted on the first corresponding mirror subcarrier at a frequency of −f1, a second reference complex value to be transmitted on the second reference subcarrier at a frequency + f2, and a second corresponding mirror complex value, which should be transmitted on the second corresponding mirror subcarrier at a frequency of -f2. 25. The system according to claim 19, in which the signal generator generates a tuning signal, which is at least the first reference complex value that must be transmitted on the first reference subcarrier at a frequency + f, the first corresponding mirror complex value, which should be transmitted on the first corresponding mirror subcarrier at a frequency of -f, a second reference complex value to be transmitted on a second reference subcarrier at a frequency (f + 1), an adjacent frequency + f, and a second corresponding mirror com the plex value to be transmitted on the second corresponding mirror subcarrier at a frequency of (f + 1), an adjacent frequency of -f. 26. The system of claim 19, wherein the transmitter transmits a training signal and a plurality of communication data symbols in a single symbol period to N subcarriers, where N is the number of communication data symbols added to the number of reference subcarriers. 27. The system of claim 19, wherein the transmitter transmits a plurality of symbols of communication data on reference frequency subcarriers and corresponding mirror frequency subcarriers in a second symbol period following a single symbol period. 28. A system for calculating a channel estimate using a tuning signal, comprising
a receiver configured to receive a tuning signal representing a plurality of reference complex values respectively received on a plurality of reference frequency subcarriers during a single symbol period, and a plurality of corresponding mirror complex values respectively received on a plurality of corresponding mirror frequency subcarriers during a single symbol period, moreover, the sum of the products of the reference complex values and the corresponding mirror complex values is and zero; and
a processor configured to generate a plurality of processed symbols (y) representing complex plane information based on the training signal, multiply each processed symbol (y) by the conjugate value of the corresponding reference signal (p ^* ) and determine the channel estimate (h) based on multiplication results. 29. The system of claim 28, wherein the receiver is configured to receive a tuning signal representing a plurality of reference complex values respectively received on a plurality of adjacent reference frequency subcarriers and a plurality of corresponding mirror complex values respectively received on a plurality of adjacent mirror frequency subcarriers. 30. The system of claim 28, wherein the receiver is configured to receive a tuning signal representing a plurality of reference complex values respectively received on a plurality of adjacent reference frequency subcarriers without intermediate subcarriers and a plurality of corresponding mirror complex values respectively received on a plurality of adjacent corresponding mirror frequency subcarriers without intermediate subcarriers. 31. The system of claim 28, wherein the receiver is configured to receive a tuning signal, presented as:
Σp _i p _im = 0, i = 1 ... n;
where p is the reference complex value received on the reference frequency subcarrier;
p _m is the mirror complex value taken on the corresponding mirror frequency subcarrier and
n is equal to the number of reference frequency subcarriers. 32. The system of claim 28, wherein the receiver is configured to receive a tuning signal, presented as:
Σw _i p _i p _im = 0, i = 1 ... n;
where p is the reference complex value received on the reference frequency subcarrier;
p _m is the mirror complex value received on the corresponding mirror frequency subcarrier;
n is equal to the number of reference frequency subcarriers and
w is the weight coefficient. 33. The system of claim 28, wherein the receiver is configured to receive a tuning signal representing at least a first reference complex value received at a first reference subcarrier at a frequency + f1, a first corresponding mirror complex value received at a first corresponding mirror subcarrier at a frequency of-f1, the second reference complex value received at the second reference subcarrier at a frequency + f2, and the second corresponding mirror complex value received at the second corresponds her mirror -f2 subcarrier frequency. 34. The system of claim 28, wherein the receiver is configured to receive a tuning signal representing at least a first reference complex value received at a first reference subcarrier at a frequency + f, a first corresponding mirror complex value received at a first corresponding a mirror subcarrier at a frequency of -f, a second reference complex value received at a second reference subcarrier at a frequency (f + 1), an adjacent frequency + f, and a second corresponding mirror complex value taken at W with the corresponding mirror subcarrier at the frequency - (f + 1), the adjacent frequency -f. 35. The system of claim 28, wherein the receiver is configured to receive a plurality of symbols of communication data in a single symbol period. 36. The system of claim 28, wherein the receiver is configured to receive quadrature modulated communication data on the reference frequency subcarriers and corresponding mirror frequency subcarriers in a second symbol period following a single symbol period. 37. A computer-readable storage medium that stores instructions that, when executed by a processor, cause the processor to implement the method of claim 1. 38. A computer-readable storage medium that stores instructions that, when executed by a processor, cause the processor to implement the method of claim 10. 39. A device for transmitting a tuning signal, comprising:
signal generating means for generating a tuning signal representing a plurality of reference complex values that should be respectively transmitted on a plurality of reference frequency subcarriers during a single symbol period, and a plurality of corresponding mirror complex values that should be respectively transmitted on a plurality of corresponding mirror frequency subcarriers in the course of a single period of the symbol, the sum of the products of the reference complex values and the corresponding x mirror complex values equal to zero; and
transmission means for transmitting a training signal within a single symbol period. 40. The device according to § 39, in which the means of generating signals includes means for generating a tuning signal, which is a set of reference complex values that must be respectively transmitted on a set of adjacent reference frequency subcarriers, and many corresponding mirror frequency values, which should be respectively transmitted on a plurality of adjacent mirror frequency subcarriers. 41. The device according to § 39, in which the signal generating means includes a means for generating a tuning signal, which is a set of reference complex values that must be respectively transmitted on a set of adjacent reference frequency frequency subcarriers without intermediate subcarriers, and a plurality of corresponding mirror complex values which should be respectively transmitted on a plurality of adjacent mirror frequency subcarriers without intermediate subcarriers. 42. The device according to § 39, in which the means for generating signals includes means for generating a tuning signal as follows:
Σp _i p _im = 0, i = 1 ... n;
where p is the reference complex value to be transmitted on the reference frequency subcarrier;
p _m is the mirror complex value to be transmitted on the corresponding mirror frequency subcarrier; and
n is equal to the number of reference frequency subcarriers. 43. The device according to § 39, in which the signal generation means includes means for generating a tuning signal as follows:
Σw _i p _i p _im = 0, i = from 1 to n;
where p is the reference complex value to be transmitted on the reference frequency subcarrier;
p _m is the mirror complex value to be transmitted on the corresponding mirror frequency subcarrier;
n is equal to the number of reference frequency subcarriers and
w is the weight coefficient. 44. The device according to § 39, in which the signal generating means includes means for generating a tuning signal, which is at least the first reference complex value to be transmitted on the first reference subcarrier at a frequency + f1, the first corresponding mirror the complex value to be transmitted on the first corresponding mirror subcarrier at a frequency of -f1, the second reference complex value to be transmitted on the second reference subcarrier at a frequency + f2, and second corresponding mirror complex value to be transmitted on the second subcarrier on the corresponding mirror frequency -f2. 45. The device according to § 39, in which the signal generation means includes means for generating a tuning signal, which is at least a first reference complex value to be transmitted on the first reference subcarrier at a frequency + f, the first corresponding mirror the complex value to be transmitted on the first corresponding mirror subcarrier at a frequency of -f, the second reference complex value to be transmitted on the second reference subcarrier at a frequency (f + 1) adjacent frequency + f, and the second corresponding mirror complex value, which should be transmitted to the second corresponding mirror subcarrier at the frequency - (f + 1), adjacent frequency -f. 46. The device according to § 39, in which the transmission means includes means for transmitting a training signal and a plurality of symbols of communication data in a single symbol period on N subcarriers, where N is the number of symbols of communication data added to the number of reference subcarriers. 47. The apparatus of claim 39, wherein the transmission means includes means for transmitting a plurality of symbols of communication data on reference frequency subcarriers and corresponding mirror frequency subcarriers in a second symbol period following a single symbol period. 48. A device for calculating a channel estimate using a tuning signal, the device comprising:
reception means having an input for receiving a tuning signal representing a plurality of reference complex values respectively received on a plurality of reference frequency subcarriers during a single symbol period and a plurality of corresponding mirror complex values respectively received on a plurality of corresponding mirror frequency subcarriers during a single symbol period , and the sum of the products of the reference complex values and the corresponding mirror complex values is equal to n Liu; and;
processing means for generating a plurality of processed symbols (y) representing complex plane information based on the training signal, multiplying each processed symbol (y) by the conjugate value of the corresponding reference signal (p ^* ), and determining the channel estimate (h) based on the multiplication results . 49. The device according to § 48, in which the receiving means includes means for receiving a tuning signal, which is a set of reference complex values, respectively, received on a set of adjacent reference frequency subcarriers, and a set of corresponding mirror complex values, respectively, received on a set of adjacent mirror frequency subcarriers. 50. The apparatus of claim 48, wherein the reception means includes means for receiving a tuning signal, which is a plurality of reference complex values respectively received on a plurality of adjacent reference frequency subcarriers without intermediate subcarriers, and a plurality of corresponding mirror complex values respectively received on a plurality of adjacent respective mirror frequency subcarriers without intermediate subcarriers. 51. The device according to p, in which the receiving means includes a means for receiving a tuning signal, presented in the form:
Σp _i p _im = 0, i = 1 ... n;
where p is the reference complex value received on the reference frequency subcarrier;
p _m is the mirror complex value received on the corresponding mirror frequency subcarrier; and
n is equal to the number of reference frequency subcarriers. 52. The device according to p, in which the receiving means includes a means for receiving a tuning signal, presented in the form:
Σw _i p _i p _im = 0, i = from 1 to n;
where p is the reference complex value received on the reference frequency subcarrier;
p _m is the mirror complex value received on the corresponding mirror frequency subcarrier;
n is equal to the number of reference frequency subcarriers and
w is the weight coefficient. 53. The device according to p, in which the receiving means includes means for receiving a tuning signal, representing at least a first reference complex value received at the first reference subcarrier at a frequency + f1, the first corresponding mirror complex value received on the first corresponding mirror subcarrier at a frequency of -f1, a second reference complex value received at the second reference subcarrier at a frequency + f2, and a second corresponding mirror complex subcarrier at a second corresponding mirror sub-carrier frequency to -f2. 54. The device according to p, in which the receiving means includes means for receiving a tuning signal, representing at least a first reference complex value received on the first reference subcarrier at a frequency + f, the first corresponding mirror complex value received on the first corresponding mirror subcarrier at a frequency of -f, the second reference complex value received at the second reference subcarrier at a frequency (f + 1), an adjacent frequency + f, and the second corresponding mirror complex value, pr Captured mirror corresponding to the second subcarrier at a frequency - (f + 1), adjacent frequency -f. 55. The apparatus of claim 48, wherein the reception means includes means for receiving a plurality of symbols of communication data in a single symbol period. 56. The apparatus of claim 48, wherein the receiving means includes means for receiving quadrature modulated communication data on the reference frequency subcarriers and corresponding mirror frequency subcarriers in a second symbol period following a single symbol period. 57. A processing device for generating a tuning signal, the processing device comprising
a signal generator module configured to generate a tuning signal representing a plurality of reference complex values that should be respectively transmitted on a plurality of reference frequency subcarriers during a single symbol period, and a plurality of corresponding mirror complex values that should be respectively transmitted on a plurality of corresponding mirror frequency subcarriers during a single symbol period, the sum of the products of the reference complex values and corresponding mirror complex values is zero. 58. A processing device for calculating a channel estimate using a training signal, the processing device comprising
a receiver module configured to receive a tuning signal representing a plurality of reference complex values respectively received on a plurality of reference frequency subcarriers during a single symbol period, and a plurality of corresponding mirror complex values respectively received on a plurality of corresponding mirror frequency subcarriers during a single symbol period , and the sum of the products of the reference complex values and the corresponding mirror complex values Nij is zero; and
a calculation module configured to generate a plurality of processed symbols (y) representing complex plane information based on the training signal, multiply each processed symbol (y) by the conjugate value of the corresponding reference signal (p ^* ) and determine the channel estimate (h) by based on the results of the multiplication.

Images