KR20030047591A - Device and method for compensating received signal of orthogonal frequency division multiplexing communication system - Google Patents

Abstract

PURPOSE: An apparatus for compensating a signal of an orthogonal frequency division multiplexing mobile communication system is provided to improve the performance of an orthogonal frequency division multiplexing receiving apparatus by compensating I/Q imbalance by a feed forward method. CONSTITUTION: An analog-to-digital converter(151) converts signals of I and Q channels into digital signals. A fast Fourier transformer(157) fast Fourier transforms the signals of I and Q channels for dividing the channels. An I/Q imbalance detector(202) inputs the fast Fourier transformed signals, and examines patterns in a frame section of the input signals for outputting I/Q imbalance detecting signals. An I/Q imbalance compensator(157) inputs the fast Fourier transformed signals and compensates the input signals with the I/Q imbalance detecting signals. A channel compensator(159) inputs the signals output from the I/Q imbalance compensator(157) and compensates the channels of the input signals.

Description

Signal Compensation Device and Method for Orthogonal Frequency Division Multiplexing Mobile Communication System {DEVICE AND METHOD FOR COMPENSATING RECEIVED SIGNAL OF ORTHOGONAL FREQUENCY DIVISION MULTIPLEXING COMMUNICATION SYSTEM}

The present invention relates to a communication apparatus and method of a communication system, and more particularly, to a communication apparatus and method of a communication system using orthogonal frequency division multiplexing.

Orthogonal Frequency Division Multiplexing (hereinafter, referred to as OFDM), which is recently used as a useful method for high-speed data transmission in wired and wireless channels, is a method of transmitting data using a multi-carrier. It is a kind of Multi Carrier Modulation (MCM) method that converts the symbol strings inputted in parallel to each other and modulates them into a plurality of sub-carriers (Sub-Carriers, Sub-Channels) having mutual orthogonality. .

The system applying the multi-carrier modulation scheme was first applied to military HF radio in the late 1950s. Since the implementation was a difficult problem, there was a limit to the actual system application. However, in 1971, Weinstein et al. Announced that modulation and demodulation using the orthogonal frequency division multiplexing can be efficiently processed using the Discrete Fourier Transform (DFT). In addition, the use of guard intervals and the introduction of cyclic prefix guard intervals have further reduced the negative effects of the system on multipath and delay spread. Thus, this orthogonal frequency division multiplexing technique is used for digital audio broadcasting (DAB), digital television, wireless local area network (WLAN) and wireless asynchronous transfer mode (WATM). It is widely applied to digital transmission technology. In other words, due to hardware complexity, it is not widely used, but it is realized by the development of various digital signal processing technologies including the Fast Fourier Transform (FFT) and the Inverse Fast Fourier Transform (IFFT). It became possible. The orthogonal frequency division multiplexing scheme is similar to the conventional frequency division multiplexing (FDM) scheme, but most of all, an optimal transmission efficiency can be obtained during high-speed data transmission by maintaining orthogonality among a plurality of subcarriers. It also has the characteristics of high frequency usage efficiency and strong multi-path fading, and thus has the characteristics of obtaining optimum transmission efficiency in high-speed data transmission. In addition, because the frequency spectrum is superimposed, frequency use is efficient, strong in frequency selective fading, strong in multipath fading, and protection intervals can be used to reduce the effects of inter symbol interference (ISI). In addition, it is possible to simply design the equalizer structure in terms of hardware and has the advantage of being resistant to impulsive noise, and thus it is being actively used in the communication system structure.

However, when demodulating the in-phase channel signal and the quadrature-phase channel signal in the receiver of the OFDM communication system as described above, the I signal is demodulated. Performance degradation is caused when an imbalance between the amplitude and phase of the channel and Q channel signals occurs. In addition, in the case of Coded Orthogonal Frequency Division Multiplexing (hereinafter referred to as COFDM), if the weight of a branch metric is added according to channel state information in addition to an input signal, the current channel decoder cannot be used. There were problems.

First, the problems caused by the imbalance of the I / Q channel signal are discussed.

The I / Q imbalance problem is a problem not only in a multi-carrier system but also in a system using a single-carrier. In addition, the I / Q imbalance problem may not be a problem in a system having a low data transmission rate, but a problem in a system having a high data transmission rate such as a wireless local area network (WLAN). This is because the C / I characteristic is limited by the I / Q imbalance so that an error flow is formed, so that the Bit Error Rate (BER) characteristic is not improved no matter how large the output of the transmission signal is. 1 is a graph showing the characteristics of BER performance degradation due to I / Q imbalance in an OFDM communication system. In order to solve the above problems, it is possible to improve the characteristics of the transmission and reception device to lower the I / Q imbalance below a certain level or to compensate for the generated I / Q imbalance.

A method for compensating for such I / Q imbalance is described in the article "Effects of tuner IQ imbalance on multicarrier-modulation systems" [M. Buchholz, A. Schchuert and R. Hasholzner. IEEE Proc. 2000]. Consider the I / Q imbalance compensation operation disclosed in the paper.

2 is a diagram illustrating a configuration for modeling an I / Q imbalance of a receiver. In FIG. 2, δω represents an offset between the input signal r (t) and the output of the local oscillator, φ represents a phase imbalance of the I channel and Q channel signals, and ε is Gain imbalance of the I and Q channel signals is shown. S _I (t) is transmission data of I channel, and S _Q (t) represents transmission data of Q channel.

Referring to FIG. 2, input signals are applied to the mixers 111 and 113, respectively. Then, the mixer 111 mixes the input signal r (t) and a signal generated by a local oscillator (not shown) to frequency down conversion, and the mixer 121 converts the input signal r (t) and the local oscillation signal. Mix and convert frequency down. At this time, an offset between the input signal and the local oscillation signal is generated in the mixing process, and phase imbalance is also generated. The signals down-converted by the mixers 111 and 121 are applied to the gain controllers 113 and 123, respectively, and the gain regulators 113 and 123 control the gains of the frequency converted signals by corresponding gain control signals, respectively. At this time, gain imbalances 1 + ε and 1-ε are generated in the gain control process. The low pass filters 115 and 125 then low-pass the signals of the gain-adjusted I and Q channels to the baseband and output them as S _I (t) and S _Q (t) signals. The S _I (t) signal and the S _Q (t) signal are given by Equation 1 below, and are converted into digital signals by the A / D converters 117 and 127, respectively.

As shown in FIG. 2, it can be seen that an imbalance between phases and gains is caused between I-channel and Q-channel signals in a process of converting a received signal into digital data by falling frequency converting the received signal. Therefore, the reception apparatus can perform a reliable decoding operation in the latter stage only by solving the imbalance between the I channel and the Q channel signals generated in the reception process as described above.

3 is a diagram illustrating a configuration of a conventional receiver for solving an I / Q imbalance problem in a communication system using an orthogonal frequency division multiplexing scheme.

Referring to FIG. 3, an A / D converter 151 performs a function of converting a received RF signal into digital data. The I / Q imbalance compensator 192 compensates for the imbalance of the I / Q channel of the digital data converted by the I / Q imbalance detection signal described later. The multiplier 155 multiplies the signal compensated for the I / Q imbalance by an automatic frequency control signal to be described later to automatically compensate the frequency of the received data. The Fast Fourier Transform (FFT) 157 performs Fourier transform of data output from the multiplier 155 at high speed. An automatic frequency controller (AFC) 153 generates an automatic frequency control signal of the Fourier transformed signal at high speed and applies it to the multiplier 155. A channel compensator 159 inputs the fast Fourier transformed signal to compensate the channel signal. An I / Q imbalance detector 190 inputs the channel compensated signal to input the I and Q channels. The imbalance between the signals is detected, and the detected imbalance signal is applied to the I / Q imbalance compensator. A soft decision block 161 determines and outputs the channel compensated signal. Decoder block 163 decodes the soft decision signal.

In the configuration of FIG. 3, the thick line represents a complex signal as a data flow, and the thin solid line represents a control signal.

The method for compensating the I / Q imbalance as described above detects an I / Q imbalance value from a signal having fast Fourier transform and channel compensation, and compensates for the I / Q imbalance of the currently input signal using this value. Therefore, the I / Q imbalance compensation method uses a feedback method that compensates for the I / Q imbalance of the next signal by using the I / Q imbalance value detected from the previous signal. That is, in the conventional receiver as shown in FIG. 3, in order to compensate for the I / Q imbalance, the I / Q imbalance detector 190 detects an imbalance between the I and Q channel signals from the output of the channel compensator 150. The feedback signal is applied to the I / Q imbalance compensator 192.

The conventional I / Q imbalance compensation method as described above may be an efficient method for continuous signals such as the European terrestrial digital video broadcasting system (DVB-T), but it is a wireless local area (WLAN) using a wireless packet method. It is not an efficient method for bursty input signals such as a network system. That is, in the WLAN, since an input signal is intermittently present in units of frames and a training signal for channel estimation is present at the front of the frame, feed forward rather than the feedback method as described above. A method of compensating for I / Q imbalance is advantageous.

Second, a conventional method of decoding received according to channel state information in addition to an input signal in coded orthogonal frequency division multiplexing (hereinafter referred to as COFDM) will be described.

12A is a diagram illustrating a configuration of a conventional COFDM decoding apparatus using channel state information (CSI).

Referring to FIG. 12A, an analog-to-digital converter 411 performs a function of converting a received RF signal into digital data. An auto frequency controller (FCC) 413 generates an auto frequency control signal for controlling the frequency of the converted digital data to maintain a predetermined frequency. A multiplier 415 multiplies the converted digital data with the automatic frequency control signal to compensate for maintaining a constant frequency of the digital data. A Fast Fourier Transformer (FFT) 417 performs Fourier transform of data output from the multiplier 415 at high speed. A channel estimator 419 inputs the fast Fourier transformed signal to output a channel compensation signal and channel state information CSI for compensating for the channel signal. An equalizer 421 compensates for the fast Fourier transformed signal according to the channel compensation signal and outputs the compensated signal. That is, the equalizer 421 performs a channel compensator function to compensate the channel reception signal by the channel compensation signal.

When transmitting, the transmitter maps the binary data (0 or 1) to a hexadecimal signal (+1 or -1) and outputs the signal mapping. The coded data is interleaved and output in order to prevent burst errors. Therefore, upon reception, the signal-mapped signal should be de-mapped and the interleaved signal should be deinterleaved and converted into the original signal. The demapping block 422 demaps and outputs the signal output from the equalizer 421. The deinterleaver 423 deinterleaves the demapped signal. A decoder 425 decodes the signal output from the deinterleaver 423 and outputs the decoded signal.

In the configuration of FIG. 12A, a thick line represents a complex signal as a data flow, and a thin solid line represents a control signal.

12B is a diagram showing the configuration of the decoder 425 in FIG. 12A.

Referring to FIG. 12B, the buffer 451 temporarily stores deinterleaved data. A branch metric calculation block 453 calculates a branch metric for decoding the buffered signal. ACS455 calculates the branch metric. Using the branch metric calculated value, the state having the shortest distance among the two previous states that can come to each state is selected in the Add Compare and Selection (ACS) 455. The path memory 457 stores the state selected in the ACS455. Trace back 459 outputs the final decoded data using the history values of the states obtained from the path memory 457.

12A and 12B, a conventional COFDM decoding apparatus using CSI having a structure as shown in FIG. (weight) to improve the performance of the decoder. However, the conventional decoding method as described above has a problem that the existing decoder cannot be used as it is because the weight of the branch metric must be weighted according to the CSI information in addition to the input signal of the decoder. Therefore, conventionally, a decoder for using the CSI had to be designed separately. In addition, when a transmitter performs modulation of QPSK or more, a symbol assigned to one subcarrier is composed of a plurality of pairs of data bits. In this case, when the bit interleaving operation is performed, the weighting process is very complicated when obtaining the branch metric. Thus, when the conventional decoding method using the CSI is used, the decoding performance is significantly degraded. Table 1 below shows the performance degradation caused by the bit interleaver / deinterleaver in the case of using the conventional CSI decoding method.

Constellation Full interleaver no bit interleaver QPSK 0.7 dB 2.9 dB 16-QAM 0.65 dB 2.7 dB 64-QAM 0.15 dB 2.2 dB

Accordingly, an object of the present invention is to provide an apparatus and method for improving the performance of a communication system using orthogonal frequency division multiplexing.

Another object of the present invention is to provide an apparatus and method for compensating for an imbalance between an in-phase and quadrature channel signal in a receiver of a system using an orthogonal frequency division multiplexing scheme.

It is still another object of the present invention to provide an apparatus and method for compensating for an imbalance between an in-phase and quadrature channel signal using a frequency domain scheme in a receiver of a system using an orthogonal frequency division multiplexing scheme.

It is still another object of the present invention to provide an apparatus and method for compensating for an imbalance between an in-phase and quadrature channel signal using a time domain scheme in a receiver of a system using an orthogonal frequency division multiplexing scheme.

It is still another object of the present invention to provide an apparatus and method for decoding using received channel state information in a receiving apparatus of a system using an encoded orthogonal frequency division multiplexing scheme.

It is still another object of the present invention to generate soft-determined data by using extracted channel state information before performing a decoding operation in a receiver of a system using an encoded orthogonal frequency division multiplexing scheme, and to decode the soft-determined data. It is to provide an apparatus and method that can be.

It is still another object of the present invention to use a code bit output from an inverse signal converter in a receiver of a system using a coded orthogonal frequency division multiplexing scheme, and the confidence bits are converted to each subcarrier by giving a weighted weight by CSI. After reflecting the characteristics of each channel, it provides an apparatus and method that can be decoded using this.

An embodiment of the present invention for achieving the above object is the frequency drop in the receiving apparatus of the orthogonal frequency division multiple communication system having a frequency down converter for converting the received signal to the baseband to convert the signal of the I channel and Q channel The present invention relates to a device for compensating for imbalance between I and Q channels generated during conversion, and comprising: an analog / digital converter for converting the signals of the I and Q channels into digital signals, and a fast Fourier transform of the I and Q channel signals. An I / Q imbalance detector for inputting the fast Fourier transform signal for distinguishing channels, the fast Fourier transform signal, and outputting an I / Q imbalance detection signal by examining a known pattern in a frame section of the input signal; An I / Q imbalance beam that inputs a fast Fourier transform signal and compensates the input signal with the I / Q imbalance detection signal for output. Group, and a signal output from the I / Q unbalance compensator characterized in that configured by a channel compensator for compensating for the channel of the input signal.

In addition, an embodiment of the present invention for achieving the above object is a frequency division converter for converting the received signal to the baseband to convert the signal of the I-channel and Q-channel in the receiving apparatus of the orthogonal frequency division multiple communication system A device for compensating for the imbalance between I and Q channels generated during a falling conversion, comprising: an analog / digital converter for converting the signals of the I and Q channels into a digital signal, and the digitally converted I and Q channel signals. And an I / Q imbalance detector for detecting an I / Q imbalance detection signal of amplitude and phase by examining the input signal, and inputting the digitally converted I and Q channel signals, and converting the input signal into the I / Q signal. I / Q imbalance compensator for compensating and outputting by imbalance detection signal and fast Fourier transform for dividing channels by fast Fourier transforming the signal compensated for I / Q imbalance. The group consisting of, a channel compensator for compensating for the channel of the fast Fourier transformed signal is characterized.

1 is a diagram illustrating deterioration characteristics of BER according to I / Q imbalance in an OFDM receiver.

2 is a diagram illustrating an unbalanced component of I and Q channel signals.

3 is a diagram illustrating a configuration for compensating for I / Q imbalance in a conventional OFDM receiver.

4 is a diagram illustrating a configuration for compensating I / Q imbalance in a frequency domain in an OFDM receiver according to an embodiment of the present invention.

FIG. 5 is a flow chart showing operation of the I / Q imbalance detector in FIG.

6 is a flowchart showing the operation of the I / Q imbalance compensator in FIG.

7 is a diagram illustrating a configuration for compensating I / Q imbalance in a time domain in an OFDM receiver according to an embodiment of the present invention.

8 is a flow chart showing operation of the I / Q imbalance detector in FIG.

9 is a flowchart illustrating the operation of the I / Q imbalance compensator in FIG.

FIG. 10A is a diagram showing characteristics when I / Q imbalance compensation is not performed, and FIG. 10B is a diagram showing characteristics when the signal as shown in FIG. 10A is compensated by the frequency domain compensation method.

FIG. 11A is a diagram showing characteristics when I / Q imbalance compensation is not performed, and FIG. 11B is a diagram showing characteristics when the signal as shown in FIG. 10A is compensated by the time domain compensation method.

FIG. 12A is a diagram showing the configuration of a conventional COFDM decoder using CSI, and FIG. 12B is a diagram showing the structure of a Viterbi decoder in FIG. 12A.

FIG. 13A is a diagram showing the configuration of a COFDM decoding apparatus according to an embodiment of the present invention using CSI, and FIG. 13B is a diagram showing the configuration of a data generation unit in FIG. 13A.

14 is a flowchart illustrating a CSI processing procedure in a system for COFDM decoding using CSI according to an embodiment of the present invention.

15A is a diagram illustrating a multipath channel model received by a COFDM decoding apparatus, and FIG. 15B is a diagram illustrating performance improvement characteristics of a COFDM system using CSI according to an embodiment of the present invention.

DETAILED DESCRIPTION Hereinafter, detailed description of preferred embodiments of the present invention will be described with reference to the accompanying drawings. It should be noted that the same components in the drawings represent the same numerals wherever possible.

An apparatus and method for improving the performance of an OFDM receiver of the present invention. Methods for improving the performance of the OFDM receiver include a method for compensating for the I / Q imbalance of the received signal, and a method for improving the performance of the CSI COFDM decoder. In the embodiment of the present invention, the apparatus and method for implementing the former will be referred to as the first embodiment, and the apparatus and method for implementing the latter will be referred to as the second embodiment.

First embodiment

The present invention relates to an apparatus and method for compensating for I / Q imbalance in a receiving apparatus of a mobile communication system using OFDM. The method for compensating the I / Q imbalance is divided into a frequency domain method and a time domain method. The compensation method of the frequency domain is a method of compensating I / Q imbalance after performing the fast Fourier transform, and the compensation method of the time domain is a method of compensating I / Q imbalance before performing the Fast Fourier Transform.

First, we look at the compensation method of the frequency domain.

4 is a diagram illustrating a configuration of an apparatus for compensating I / Q imbalance in the frequency domain method according to the first embodiment of the present invention.

Referring to FIG. 4, the input signal becomes the baseband signals S _I (t) and S _Q (t) of the I and Q channels as shown in FIG. 2, and the A / D converter 151 is Convert an input signal such as to a digital signal. The automatic frequency controller 153 generates an automatic frequency control signal for controlling the frequency of the converted digital data to maintain a predetermined frequency. A multiplier 155 multiplies the converted digital data with the automatic frequency control signal to compensate for maintaining a constant frequency of the digital data. The fast Fourier transformer 157 performs fast Fourier transform on the data output from the multiplier 155.

The I / Q imbalance compensator 204 compensates for an imbalance between the I channel and the Q channel signal of the fast Fourier transform signal by the I / Q imbalance detection signal described below. The channel compensator 159 performs channel compensation on the signal for which the I / Q imbalance is compensated. An I / Q imbalance detector 202 inputs the channel compensated signal to detect an imbalance between an I channel and a Q channel signal, and applies the detected imbalance signal to the I / Q imbalance compensator 202. The soft decision unit 161 determines and outputs the channel compensated signal. The decoder 163 decodes the soft decision signal.

In the configuration of FIG. 4, the thick line represents a complex signal as a data flow, and the thin solid line represents a control signal.

FIG. 5 is a flowchart illustrating the operation of the I / Q imbalance detector 202 in the apparatus for compensating the I / Q imbalance by the frequency domain compensation method as shown in FIG. 4, and FIG. 6 is a diagram illustrating the operation of the I / Q imbalance compensator 204. It is a flow chart.

Hereinafter, the I / Q imbalance detection and compensation process in the frequency domain method will be described in detail.

Here, S (t) = S _I (t) + jS _Q (t), and S (t) = S _I (t) + jS _Q (t). t) `is given by Equation 2 below.

In Equation 2, S (t) is a desired signal, S ^* (t) is an interference signal as an image signal of S (t), and * represents a complex conjugate. Α and β in Equation 2 are respectively given by Equation 3 below.

If the frequency offset is removed in Equation 2, S (t) `is given by Equation 4 below.

If the signal obtained by fast Fourier transforming S` (t) is S (k) `, S (k)` from Equation 4 is given by Equation 5 below.

Since S` (k) is a value obtained by fast Fourier transforming an input signal, S (k), which is a desired signal, can be obtained by knowing α and β. Α and β are obtained using the training signal (or pilot signal). In the case of IEEE801.11a, α and β will be obtained. In IEEE802.11a, a short training signal, a long training signal, and a pilot signal have an integer value of (1) or (-1) in a mode frequency domain. In order to calculate α and β, S (k) and S (-k) are divided into groups as shown in Equation 6 below.

In order to reduce the influence of noise when calculating α and β, the received signal when the transmission signal belongs to each group is averaged. This averaged received signal Speaking of which, Is given by Equation 7 below.

In Equation 7, AVG represents an average, and from Equation 7, Is as shown in Equation 8 below.

Α and β are obtained from Equation 8 as shown in Equation 9 below.

Referring to FIG. 5, the I / Q imbalance detector 202 receives an S (k) 'signal as shown in Equation 5 at step 262, which is the start time of I / Q imbalance detection. S (k) 'in Equation 5 is a modeling value of the input signal, and α and β are the same as Equation 3 above. In step 264, the I / Q imbalance detector 202 divides S (k) and S (−k) into four groups as shown in Equation 6 below to obtain α and β. That is, in the frequency domain, the pilot signal has a value of +1 or -1, and this pattern is a known pattern known to the transmitter and the receiver in advance. Therefore, the I / Q imbalance detector 202 groups the received pilot signals according to a known pattern. That is, signals at the time of transmission group signals having the same value, and the result is shown in Equation 6 above. The reason for performing the grouping is to reduce the noise effect of the detected signal.

In operation 266, the I / Q imbalance detector 202 averages the received signal when the transmission signal belongs to each group to reduce the influence of noise when obtaining α and β. That is, in step 266, the grouped signals are averaged for each group to remove noise. When the grouped signals are averaged for each group, noise is eliminated, thereby reducing errors in detecting I / Q imbalance.

In step 268, the I / Q imbalance detector 202 applies the received signal averaged to Equation 8 to the groups shown in Equation 6 and converts it to Equation 8. Therefore, in step 268, I / Q imbalance values of the groups are detected. In operation 270, α and β, as shown in Equation 9, are obtained from Equation 8. That is, in step 270, the I / Q imbalance component is estimated by adding and averaging the I / Q imbalance values of the groups detected in step 268. Where α is an element that distorts the original signal S (t) and β is an element that distorts the complex signal S ^* (t) d, which is an unbalanced component of the I and Q channels.

As described above, the I / Q imbalance detector 202 inputs a signal of the fast Fourier transformer 157 which distinguishes channels in the OFDM receiver. In this case, the input signal is a signal intermittently transmitted in units of frames, and the I / Q imbalance detector 202 may be a training signal for channel estimation located earlier in the frame signal. In this case, the training signal may be a pilot signal, and the pattern of the pilot signal is known in advance by the transmitting side and the receiving side.

Accordingly, the I / Q imbalance detector 202 may include a grouper, an averager, and an estimator to detect the I / Q imbalance component. Here, the grouper groups the input signal into four groups according to a pattern at the time of transmission, an averager removes the noise component by averaging the grouped signals in each group unit, and an estimator removes the signals of the groups. After addition, it is averaged and output as an I / Q imbalance detection signal.

Next, the operation of compensating for the I / Q imbalance compensator 204 using the detected imbalance signals of the I and Q channels will be described.

The method for obtaining S (k) with I / Q imbalance compensated from Equation 5 is as follows. When S ^* (−k) `is obtained from Equation 5, Equation 10 is obtained.

Using Equation 5 and Equation 10, S (k) having an I / Q imbalance compensated may be obtained. When Equation 5 is multiplied by α ^* and then Equation 10 is multiplied by β, it is as shown in Equation 11 below. From Equation 11, S (k) can be obtained as shown in Equation 12 below.

Referring to FIG. 6, the I / Q imbalance compensator 204 inputs a signal output from the fast Fourier transformer 157 and an I / Q imbalance detection signal output from the I / Q imbalance detector 202 in step 282. That is, the I / Q imbalance compensator 204 obtains S ^* (k) as shown in Equation 5 from Equation 5 in step 282 to obtain S (k) compensated for I / Q imbalance. In this case, the signal output from the fast Fourier transformer 157 is an I / Q imbalance signal, and generates an S (k) signal and an S ^* (-k) signal conjugated to this signal. In other words, Because of, Becomes The I / Q imbalance compensator 204 compensates for the S (k) 'and S ^* (− k) signals using the I / Q imbalance detection signal in steps 284 and 286. That is, in step 284, an imbalance between the signals of the I channel and the Q channel is performed by using Equation 5 and Equation 10 as shown in Equation 11, and according to the result, Equation 12 in Equation 286. After compensating with the desired signal S (k), it is output to the channel compensator 159.

FIG. 10A is a constellation diagram illustrating characteristics when the I channel / Q channel imbalance of the signal received when the NFR is 25 dBc is not compensated for, and FIG. 10 b is the I channel of the signal received when the NFR is 25 dBc. A diagram showing characteristics when the imbalance of the / Q channel is compensated by the frequency domain compensation method. As shown in FIGS. 10A and 10B, it can be seen that by compensating for the I / Q imbalance using the frequency domain compensation method, the costelization destortion can be largely compensated.

Second, we look at the time domain compensation method.

7 is a diagram illustrating a configuration of an apparatus for compensating I / Q imbalance in a time domain method according to the first embodiment of the present invention.

Referring to FIG. 7, the A / D converter 151 performs a function of converting a received RF signal into digital data. The automatic frequency controller 153 generates an automatic frequency control signal for controlling the frequency of the converted digital data to maintain a predetermined frequency. A multiplier 155 multiplies the converted digital data with the automatic frequency control signal to compensate for maintaining a constant frequency of the digital data.

The I / Q imbalance compensator 304 compensates for an imbalance between the I-channel and Q-channel signals output from the multiplier 155 by the I / Q imbalance detection signal described below. The I / Q imbalance detector 302 inputs the signal output from the multiplier 155 to detect an imbalance between the I and Q channel signals, and applies the detected imbalance signal to the I / Q imbalance compensator 202.

The fast Fourier transformer 157 performs fast Fourier transform on the signal output from the I / Q imbalance compensator 304. The channel compensator 159 inputs the Fourier transform signal to compensate for the channel. The soft decision unit 161 determines and outputs the channel compensated signal. The decoder 163 decodes the soft decision signal.

In the configuration of FIG. 7, the thick line represents a complex signal as a data flow, and the thin solid line represents a control signal.

FIG. 8 is a flowchart illustrating an operation of an I / Q imbalance detector 302 in an apparatus for compensating I / Q imbalance with a time domain compensation method as shown in FIG. 7, and FIG. 9 illustrates a structure of an I / Q imbalance compensator 304. Drawing.

Hereinafter, the I / Q imbalance detection and compensation process in the time domain method will be described in detail.

S _I (t) `and S <sub>Q</sub> (t)` whose frequency offset is compensated from Equation 1 are obtained from Equation 13.

In Formula 13, S _I (t) and S _Q (t) are assumed to be mutually independent and uncorrelated, and assuming that the powers are the same, The same equations can be obtained.

here Δ and ε can be obtained as shown in Equation 18 below.

Using Equation 18 in Equation 15, φ is given by Equation 19.

In Equation 19 above Φ of Equation 19 can be approximated by Equation 20 below.

The φ may also be obtained by using Equations 15 and 17, or by using Equations 16 and 17. When ε and φ are obtained from Equations 18 and 19, S _I (t) and S _Q (t) can be obtained from Equation 13.

Referring to FIG. 8, the I / Q imbalance detector 302 inputs a signal output from the multiplier 155 in step 362. At this time, the input signal is S _I (t) `and S <sub>Q</sub> (t)` whose frequency offset is compensated from the signal as shown in Equation (1).

In step 364, the I / Q imbalance detector 302 detects the expected value according to each condition. In other words, assuming that S _I (t) and S _Q (t) are independent of each other, are uncorrelated, and have the same power, the I / Q imbalance detector 302 is expressed in Equation 14 at step 364. You can get it as That is, assuming that the powers of the transmission signals of the I channel and the Q channel are the same, an expectation value can be obtained as shown in Equation (14). In addition, assuming that the signals of the I channel and the Q channel are independent from each other and uncorrelated, an average value obtained by multiplying and multiplying the I channel received signal S _I (t) `(distorted signal) can be obtained. By multiplying and multiplying the Q channel received signal S _Q (t) &quot; However, since the received signals of the I channel and the Q channel are correlated with each other and are not independent, an expectation value as shown in Equation 17 can be obtained.

Thereafter, the I / Q imbalance detector 302 detects I / Q imbalance values for amplitude ε and phase φ by using Equation 18 in Equation 14. First, in step 366, the I / Q imbalance detector detects an I / Q imbalance for the amplitude ε. In this case, the I / Q imbalance detector 302 detects an I / Q imbalance with respect to amplitude? As shown in Equation 18 using the expected values such as Equations 15 and 16 obtained in step 364 in step 366. do.

The I / Q imbalance detector 302 calculates an I / Q imbalance for the phase φ while performing steps 368 and 370 using Equation 15-18. The I / Q imbalance for the phase φ may be obtained using Equation 15 and Equation 17 or Equation 16 and Equation 17.

As described above, the I / Q imbalance detector 302 that compensates for the I / Q imbalance in the time domain may include an expectation detector, an amplitude imbalance detector, and a phase imbalance detector. Here, the expectation detector obtains a first expectation value as shown in Equation 14 on the assumption that the I and Q channel transmission signals have power, and assumes that the I and Q channel signals are independent of each other and are uncorrelated. The third expected value as shown in Equation 16 is obtained. Since the I and Q channel signals are correlated and independent of each other, a fourth expected value is obtained. The amplitude imbalance detector then detects I / Q imbalances for amplitude using the second and third expected values, and the phase imbalance detector uses the second and fourth expected and amplitude imbalance values for phase I. / Q Detect imbalance.

When the I / Q imbalance component for amplitude ε and phase φ is detected in the time domain as described above, the I / Q imbalance compensator 304 having the structure as shown in FIG. 9 is an imbalance component of the I channel and the Q channel included in the received signal. To compensate.

Referring to FIG. 9, first, multipliers 332 and 334 respectively correspond to gain imbalance components 1 / (1 + ε) and 1 / (1- corresponding to S _I (t) and S _Q (t) signals, respectively. multiply by ε). Therefore, the gain imbalance component included in the received signal is compensated for the signals output from the multipliers 332 and 334.

Thereafter, the signals output from the multipliers 332 and 334 are applied to the multipliers 340 and 342, and the multipliers 340 and 342 multiply the input signals by the phase imbalance component sin (φ / 2), respectively. In addition, the signals output from the multipliers 332 and 334 are applied to the multipliers 336 and 338, and the multipliers 336 and 338 multiply the input signals by the phase imbalance component cos (φ / 2), respectively. A subtractor 344 subtracts the output of the multiplier 342 from the output of the multiplier 336 to generate an aS _I (t) signal that compensates for the phase imbalance component. The subtractor 346 subtracts the output of the multiplier 340 from the output of the multiplier 338 to generate an aS _Q (t) signal that compensates for the phase imbalance component. Where a is a constant in the signal where the I / Q imbalance component is compensated.

Accordingly, the I / Q imbalance compensator 304 having the configuration as shown in FIG. 9 is configured to multiply the amplitude imbalance detection signals corresponding to the I channel signal and the Q channel signal by the first multipliers 332 and 334, respectively. It performs the function of compensating the amplitude respectively. The second multipliers 340 and 342 multiply the amplitude compensation signals of the I and Q channels by the phase imbalance detection signal and perform a first compensation of the phases of the I and Q channel signals, respectively. In addition, the third multipliers 336 and 338 perform a second compensation on the phases of the I-channel and Q-channel signals by multiplying the amplitude compensation signals of the I-channel and the Q-channel by the phase imbalance detection signal shifted 90 degrees. And subtractors 344 and 346 subtract the first phase compensated Q channel signal from the second phase compensated I channel signal and subtract the first phase compensated I channel signal from the second phase compensated Q channel signal. Perform the function.

FIG. 11A is a constellation diagram illustrating characteristics when the I channel / Q channel imbalance of the signal received when the NFR is 25 dBc is not compensated for, and FIG. 11B is the I channel of the signal received when the NFR is 25 dBc. A diagram showing characteristics when the imbalance of the / Q channel is compensated by the time domain compensation method. As shown in FIGS. 11A and 11B, it can be seen that by compensating for I / Q imbalance using a time domain compensation method, a costelation destortion can be largely compensated.

Second embodiment

In the encoding orthogonal frequency division multiplexing decoding apparatus according to the second embodiment of the present invention, the conventional decoder structure is used as it is, and the CSI information characteristics are reflected in soft decided data which is an input signal of the decoder. Used as input signal of decoder.

13A is a diagram illustrating a structure of a decoding apparatus of a coding orthogonal frequency division multiplexing according to an embodiment of the present invention.

Referring to FIG. 13A, the A / D converter 411 performs a function of converting a received RF signal into digital data. The automatic frequency controller 413 generates an automatic frequency control signal for controlling the frequency of the converted digital data to maintain a predetermined frequency. A multiplier 415 multiplies the converted digital data with the automatic frequency control signal to compensate for maintaining a constant frequency of the digital data. The fast Fourier transformer 417 performs Fourier transform of the data output from the multiplier 415 at high speed. The channel estimator 419 inputs the fast Fourier-transformed signal to output a channel compensation signal and channel state information CSI for compensating for the channel signal. The equalizer 421 compensates and outputs the fast Fourier transformed signal according to the channel compensation signal. The equalizer 421 performs a channel compensator function to compensate the channel signal received by the channel compensation signal as described above.

The inverse signal converter demapping block 422 inversely converts the signal output from the equalizer 421 into the original signal form and outputs the inverse signal. The soft decided data generator 500 generates a soft decision signal that reflects the channel characteristics of each subcarrier using the CSI output from the inverse signal converter 422 and the channel estimator 419. The deinterleaver 423 deinterleaves the signal output from the soft decision signal generator 500 to the state at the time of transmission. A decoder 425 decodes the signal output from the deinterleaver 423 and outputs the decoded signal.

In the configuration of FIG. 12A, a thick line represents a complex signal as a data flow, and a thin solid line represents a control signal.

FIG. 13B is a diagram showing the configuration of the soft decision signal generator 500 of FIG. 13A. Referring to FIG. 13B, the soft decision signal generator 500 outputs a sign bit output from the inverse signal converter 422 as it is, and the confidence bits output from the soft signal converter 422 are the same. The confidence level is converted by the CSI output from the channel estimator 419 to convert the confidence level. Accordingly, the soft decision signal generator 500 transmits the code bits output from the inverse signal converter 422 to the deinterleaver 423 as they are, and the confidence bits reflect the channel characteristics of each subcarrier using the CSI to the deinterleaver 423. To pass.

As described above, the decoding apparatus of the coded orthogonal frequency division multiplexing method according to the second embodiment of the present invention does not use the method of decoding the received signal using the CSI in the decoder 425, and the soft decision signal in front of the decoder 425. A generator 500 is configured, and the soft decision signal generator 500 generates a soft decision signal of a signal received using the CSI, and applies it to the decoder 425. When the above decoding method is used, the decoding performance is not affected by the interleaver / deinterleaver. The soft decision signal generator 500 receives the sign bits and the confidence bits from the inverse signal converter 422, wherein the sign bits are applied to the deinterleaver 423 as they are, and the confidence bits are generated by the CSI output from the channel estimator 419. The weight level is added to transform the confidence level. Accordingly, the signals output from the soft decision signal generator 500 are in a state in which channel characteristics of each subcarrier are reflected, and these signals are deinterleaved by the deinterleaver 423 and then decoded by the decoder 425.

For example, among the channel characteristics of each subcarrier, the subcarrier channel better than the AWGN channel is left as it is, and the subcarrier channel worse than the AWGN channel can be used as the confidence bits by multiplying the CSI and taking 2-3 bits. have.

14 is a flowchart illustrating a procedure of generating the CSI in the channel estimator 419 in the second embodiment of the present invention.

The CSI is defined as a signal to noise ratio (SNR) in a carrier. Therefore, in order to extract the CSI information, the SNR of each sub-carrier should be obtained. In the OFDM system, channel estimation is performed in the frequency domain, and the power of the channel estimation coefficient H has an inverse relationship with the SNR. This is expressed by Equation 21 below.

In step 611 of FIG. 14, an inverse power of the channel estimated coefficient H is added in OFDM symbol units to obtain an average power P _avg of the signal. Equation 22 below is an equation for calculating an average power P _avg of the signal, where k represents an index of a subcarrier and N represents a summated number.

On each subcarrier If the average power is greater than P _avg , the soft decision signal is decoded, and if small, the confidence bit of the soft decision signal is changed by multiplying by a weight value less than one. In the case of modulation above QPSK, a symbol assigned to a subcarrier is composed of a plurality of pairs of data bits and must pass through a deinterleaver 423. Therefore, the weight value of the CSI is determined by the link between the inverse signal conversion process and the deinterleaving process. Used in the process of generating the judgment signal. The code bit uses the value determined in the soft decision signal generation process.

Accordingly, in step 613 of FIG. 14, weights of respective subcarriers are calculated using Equation 23 using the CSI information.

In Equation 23, l denotes how many levels CSI information is to be expressed, Thr _l denotes a power threshold at the l-th level, and V _{l, k} denotes the l-th level and the k-th subcarrier. W _k represents weights in the k-th subcarrier. Thr _l is obtained using P _avg , and Thr _l should be selected not to be larger than P _avg .

Subsequently, in step 615 of FIG. 14, a soft decision signal reflecting channel characteristics for each subcarrier is generated by giving a weight by the CSI information to a signal received using the weights of the CSI in the same manner as in Equation 24 below. do.

In Equation (24), S _{n, k} represents a soft decision signal in a sample carried on a k-th subcarrier, and is composed of n data bits. In Equation 24, n denotes the number of data bits carried on the subcarrier. In operation 615, the soft decision signal generated by Equation 24 is applied to the deinterleaver 423, and the decoder 425 inputs the same signal and decodes it.

FIG. 15A is a diagram illustrating the results in a multipath channel environment, where Ts represents OFDM symbol time. 15B is a diagram illustrating BER characteristics according to the second embodiment of the present invention. In the BER graph of FIG. 15B, when L is 4, the weight is less than 1, 1/2 if greater than 1/2 of P _avg , and less than 0.5, less than 1/4 and greater than 1/8 when greater than 1/4. This is the result in the multipath channel environment of FIG. 15A when the case is smaller than 0.25 and 1/8. Therefore, as shown in FIG. 15B, it can be seen that the BER characteristic is improved when the CSI method according to the embodiment of the present invention is used, compared to the conventional soft decision method.

As described above, by compensating I / Q imbalance in a feedforward manner in an OFDM receiver, wireless packet service has an advantage of effectively using I / Q imbalance of a received signal having a burst characteristic in a mobile communication system. The I / Q imbalance can be compensated in the frequency domain after performing the fast Fourier transform in the device that compensates for the I / Q imbalance by the feedforward method, and the I / Q imbalance can be compensated in the time domain before the fast Fourier transform. There is an advantage to that. In addition, by reflecting the CSI information in the soft decision signal which is the input signal of the decoder in the OFDM receiver, the decoder using the CSI can be implemented without changing the decoder, and the CSI information can reflect the channel characteristics for each subcarrier. Thus, there is an advantage that can improve the decoding performance.

Claims (18)

Compensating the imbalance between the I and Q channels generated during the frequency down conversion in a receiving apparatus of an orthogonal frequency division multiple communication system having a frequency down converter for converting a received signal into a baseband and converting signals of I and Q channels. In the device, An analog / digital converter for converting the signals of the I and Q channels into digital signals; A fast Fourier transformer for distinguishing channels by fast Fourier transforming the I and Q channel signals; An I / Q imbalance detector for inputting the fast Fourier transform signal, and outputting an I / Q imbalance detection signal by inspecting a predetermined pattern in a frame section of the input signal; An I / Q imbalance compensator for inputting the fast Fourier transform signal and compensating for the input signal with the I / Q imbalance detection signal and outputting the same; Orthogonal frequency division multiplexing device comprising a channel compensator for inputting a signal output from the I / Q imbalance compensator and compensating for a channel of the input signal. The method of claim 1, wherein the I / Q imbalance detector, A grouper for grouping according to a known pattern in the input fast Fourier transform signal; An averager for averaging the grouped signals within a corresponding group; The I / Q imbalance estimator is configured to detect the I / Q imbalance from the signals averaged in each group, and then add these signals and average them to output the I / Q imbalance detection signal. Split Multiple Receiver. The method of claim 2, wherein the I / Q imbalance compensator, A conjugator for complex-conjugating the fast Fourier transform signal; And a compensator for compensating the fast Fourier transform signal and the complex conjugated Fast Fourier transform signal with the I / Q imbalance detection signal. Compensating the imbalance between the I and Q channels generated during the frequency down conversion in a receiving apparatus of an orthogonal frequency division multiple communication system having a frequency down converter for converting a received signal into a baseband and converting signals of I and Q channels. In the device, An analog / digital converter for converting the signals of the I and Q channels into digital signals; An I / Q imbalance detector for inputting the digitally converted I and Q channel signals and detecting the I / Q imbalance detection signals of amplitude and phase by examining the input signals; An I / Q imbalance compensator for inputting the digitally converted I and Q channel signals and compensating for the input signal with the I / Q imbalance detection signal; A fast Fourier transformer for distinguishing channels by fast Fourier transforming the signal compensated for the I / Q imbalance; Orthogonal frequency division multiplexing device comprising a channel compensator for compensating for the channel of the fast Fourier transformed signal The method of claim 4, wherein the I / Q imbalance detector, Expected value of signal power of I and Q channel of transmission signal, I channel expectation and Q channel expectation when I and Qc channel signals are assumed to be independent and uncorrelated, and expectation of actual I / Q channel expectation Detectors, An amplitude imbalance detector for detecting an I / Q imbalance detection signal for amplitude from the expected values; And a phase imbalance detector for detecting an I / Q imbalance detection signal for a phase from the expected values. The method of claim 5, wherein the I / Q imbalance compensator, First multipliers for multiplying the I-channel signal and the Q-channel signal by the amplitude imbalance detection signals, respectively, to compensate for the amplitude of the I-channel and Q-channel signals, respectively; Second multipliers for first compensating the phases of the I and Q channel signals by multiplying the amplitude compensation signals of the I and Q channels by the phase imbalance detection signal, and 90 to the amplitude compensation signals of the I and Q channels, respectively. Third multipliers for second compensating phases of the I channel and the Q channel signal by multiplying the shifted phase imbalance detection signal, and the first phase compensated Q channel signal to the second phase compensated I channel signal. And a subtractor for subtracting and subtracting the first phase compensated I channel signal from the second phase compensated Q channel signal. Compensating the imbalance between the I and Q channels generated during the frequency down conversion in a receiving apparatus of an orthogonal frequency division multiple communication system having a frequency down converter for converting a received signal into a baseband and converting signals of I and Q channels. In the way, Converting the signals of the I and Q channels into digital signals; Classifying channels by fast Fourier transforming the I and Q channel signals; Inputting the fast Fourier transform signal and detecting an I / Q imbalance from a known pattern in a frame section of the input signal to generate an I / Q imbalance detection signal; Inputting the fast Fourier transform signal and compensating the input signal with the I / Q imbalance detection signal; Compensating the I / Q imbalance of the orthogonal frequency division multiplexing device, characterized in that the process of compensating the channel of the I / Q imbalance compensation signal. The method of claim 7, wherein the detecting of the I / Q imbalance comprises: Grouping according to a pattern known in advance in the input fast Fourier transform signal; Averaging the grouped signals within a corresponding group; And detecting the I / Q imbalance from the signals averaged in each group, adding these signals, and averaging them to output the I / Q imbalance detection signal. How to compensate for I / Q imbalance. The method of claim 8, wherein the compensating for the I / Q imbalance comprises: Complex-conjugating the fast Fourier transform signal; And compensating the fast Fourier transform signal and the complex conjugated fast Fourier transform signal with the I / Q imbalance detection signal. Compensating the imbalance between the I and Q channels generated during the frequency down conversion in a receiving apparatus of an orthogonal frequency division multiple communication system having a frequency down converter for converting a received signal into a baseband and converting signals of I and Q channels. In the way, Converting the signals of the I and Q channels into digital signals; Inputting the digitally converted I and Q channel signals, detecting an I / Q imbalance of amplitude and phase from the input signal, and generating an I / Q imbalance detection signal; Inputting the digitally converted I and Q channel signals and compensating the input signal with the I / Q imbalance detection signal; A method for compensating I / Q imbalance in an orthogonal frequency division multiplexing device, characterized in that the I / Q imbalance compensated signal is fast Fourier transformed to distinguish channels. The method of claim 10, wherein the detecting of the I / Q imbalance comprises: The process of calculating the expectation of the signal power of the I and Q channels of the transmission signal, the expectation of the I channel and the Q channel when the signals of the I and Qc channels are independent and uncorrelated, and the expectation of the actual I / Q channel. and, Detecting an I / Q imbalance detection signal for amplitude from the expected values; And detecting an I / Q imbalance detection signal with respect to a phase from the expected values. The method of claim 11, wherein the I / Q imbalance compensator, Compensating the amplitudes of the I and Q channel signals by multiplying the amplitude imbalance detection signals corresponding to the I and Q channel signals, respectively; Performing a first compensation on the phases of the I and Q channel signals by multiplying the amplitude compensation signals of the I and Q channels by the phase imbalance detection signal; Performing a second compensation on the phases of the I and Q channel signals by multiplying the amplitude compensation signals of the I and Q channels by the phase imbalance detection signal shifted by 90 degrees; Orthogonally subtracting the first compensated Q channel signal from the second phase compensated I channel signal and subtracting the first compensated I channel signal from the second phase compensated Q signal. A method for compensating I / Q imbalance in a frequency division multiplexing device. Claims [1] A receiver of an orthogonal frequency division multiple communication system comprising a frequency down converter for converting a received signal into a baseband and converting signals of I and Q channels. An analog / digital converter for converting the signals of the I and Q channels into digital signals; A fast Fourier transformer for dividing channels by fast Fourier transforming the digital signal; An estimator for extracting a channel compensation signal and channel state information from the fast Fourier transformed signal; A channel compensator for compensating the fast Fourier transformed signal by the channel compensation signal; An inverse signal converter for inverting the channel compensated signal; A soft decision signal generator for generating a soft decision signal reflecting the channel characteristics of each subcarrier based on the channel state information of the inverse signal converted signal; A deinterleaver for deinterleaving the soft decision signal; And a decoder for decoding the deinterleaved signal. The method of claim 13, wherein the soft decision signal generator, And transmitting the code bits of the inverse signal-converted signal as they are and weighting the confidence bits based on the channel state information to generate a soft decision signal. The method of claim 14, wherein the soft decision signal generator, The average power of the signal is obtained by adding the reverse power of the estimated channel coefficients in symbol units, The channel state information is compared with each of at least two preset reference levels, and the weight is calculated in the corresponding subcarrier. If the reverse power of each subcarrier is greater than the average power of the signal, the soft-determined data is transmitted as it is. If greater than the receiver of an orthogonal frequency division multiplexing system, multiplying the weight of the corresponding subcarrier to change the confidence bit of consecutively scheduled data. In the decoding method of an orthogonal frequency division multiple communication system comprising a frequency down converter for converting the received signal to the baseband to convert the signal of the I channel and the Q channel, Converting the signals of the I and Q channels into digital signals; Classifying channels by fast Fourier transforming the digital signal; Extracting a channel compensation signal and channel state information from the fast Fourier transformed signal; Compensating the fast Fourier transformed signal with the channel compensation signal; Inverse signal conversion of the channel compensated signal; Generating a soft decision signal reflecting the channel characteristics of each subcarrier based on the channel state information of the inverse signal converted signal; Deinterleaving the soft-determined signal; And decoding the deinterleaved signal. The method of claim 16, wherein the generating of the soft decision signal comprises: And transmitting the code bits of the inverse signal-converted signal as they are and weighting the confidence bits by the channel state information to generate a soft decision signal. The method of claim 16, wherein the soft decision signal generation process, Obtaining the average power of the signal by adding the reverse power of the channel estimated coefficients in symbol units, Comparing the channel state information with each of at least two preset reference levels and calculating a weight in a corresponding subcarrier; If the reverse power of each subcarrier is greater than the average power of the signal, the soft-determined data is transmitted as it is, and if it is greater than 1, multiplied by the weight of the corresponding subcarrier to change the confidence bit of the soft-scheduled data. A decoding method of an orthogonal frequency division multiple communication system.