KR101059878B1 - Timing offset compensation method of interference cancellation receiver - Google Patents

Abstract

The present invention relates to a timing offset compensation method of an interference canceling receiver having a feedforward filter (FF) and a feedback filter (FB) to remove interference from a received signal, the first timing offset of the training signal of the received signal and a first timing offset. Estimating 1 SNR; Training coefficients of FF and FB using a predetermined training signal; Filtering the training signal portion of the received signal by applying the trained coefficients to estimate a second timing offset and a second SNR of the training signal from which the interference signal has been removed; Comparing the first and second timing offsets with each other to determine an error of a timing offset estimate value; Correcting data of the received signal with a second timing offset value if there is an error in the timing offset estimation.

Interference cancellation receiver, timing offset, training signal, interference cancellation,

Description

Timing Offset Compensation Method for Interference Cancellation Receivers {METHOD FOR COMPENSATING TIMMING OFFSET OF INTERFERENCE CANCELLATION RECEIVER}

1 illustrates the principle of blind interference cancellation (BIC),

2 is a view showing the structure of a conventional BIC receiver;

3 is a view showing a structure of a normal burst in a conventional GSM signal,

4 is a diagram illustrating a configuration of a SAIC receiver according to the present invention;

5 is a flowchart illustrating a timing offset compensation method of a SAIC receiver according to an embodiment of the present invention.

6 illustrates an example in which the cross correlator estimates a timing offset in FIG. 4;

7 is a diagram illustrating a method of operating an adaptive algorithm used for training FF and FB coefficients in a SAIC.

******* Description of the symbols for the main parts of the drawing ********

10: phase rerotator 11: cross correlator

12 channel estimator 13 SNR estimator

14: FF (feedforward filter) 15: FB (feedback filter)

16: judge 17: matched filter

18: MLSE equalizer

The present invention relates to a mobile communication system, and more particularly, to a timing offset compensation method for an interference cancellation receiver (SIC).

In general, a large number of user signals exist in a mobile communication environment. Among them, signals other than a desired signal act as interference to degrade reception performance. For example, in the case of a time division multiplexing (TDMA) or frequency division multiplexing (FDMA) system, since the time or frequency are used separately from other users, other user signals are basically separated from the desired signals. However, when the base station is densely installed in order to increase the capacity of the system, the frequency reuse factor is increased, and several user signals exist in the same frequency or time slot, thereby causing interference.

Conventionally, technologies such as joint detection (JD) and smart antenna are used for interference cancellation. In particular, the JD is a method of increasing the detection performance by simultaneously determining a signal of another user in addition to a desired signal. The smart antenna technology is a method of attenuating other user signals spatially by changing the radiation pattern of the radio wave using multiple antennas. This method is used in a terminal sensitive to size and power consumption because multiple antennas must be mounted in the receiver. Not suitable for

Therefore, for the antipodal type signals such as BPSK, a single antenna interference cancellation (SAIC) method for removing interference with one antenna without a large amount of computation has been proposed. Standardization work is in progress.

An example of a conventional interference cancellation method is described in Korean Application No. 10-2003-7008696, which is called blind interference cancellation (BIC).

In general, without considering multipath, an antipodal signal such as BPSK usually exists on a linear axis in a constellation, and the phase component of the signal depends on the characteristics of the radio channel such as fading.

Thus, as shown in Fig. 1, a receiver without BIC determines only by compensating for the phase component of the desired signal through the channel estimator. On the other hand, a receiver applying a BIC (hereinafter referred to as a BIC receiver) estimates a phase of an interference signal component, rotates the interference signal in a specific direction, and determines a desired signal with an axis perpendicular thereto to remove the interference.

2 is a diagram showing the structure of a conventional BIC receiver described in Korean Application No. 10-2003-7008696.

Referring to FIG. 2, f _i [k] is a feedforward filter (FF) and represents a FIR filter having a plurality of coefficients for removing even an interference signal having multiple paths. b [k] is a feedback filter (FB) and receives a training sequence known in the training signal section. The FF and BF together form a Decision Feedback Equalizer (DFE), and a decision device is included in the loop of the FF and FB.

P _i {} added to the output terminal of the FF is a projection operator, and means projection to an axis perpendicular to the axis containing the interference signal. The projection operator converts a complex input into a real output.

The coefficients of the FF and FB are calculated using the training interval of the received signal (GSM signal), calculating the inverse of the Wiener-Hopf equation, or applying adaptive algorithms such as RLS (Recursive least-square) and LMS (Least means-square). Obtain by applying However, as shown in FIG. 3, the length of one normal burst training signal in the GSM signal is relatively short, which is 26. Since the calculation of the inverse matrix is practically impractical, the convergence speed is high even in the adaptive algorithm. It is appropriate to use a relatively fast RLS algorithm.

When the coefficients of the FF and FB are obtained, the BIC receiver passes the data with the coefficients fixed to obtain a determination value from the determiner.

However, when a time tracker that estimates a timing offset of a received signal does not accurately estimate an accurate timing offset value, the data value output from the BIC receiver is detected by pushing the timing offset. As a result, even if the interference included in the received signal is removed through the BIC receiver, the received signal cannot be decoded properly.

Accordingly, an object of the present invention is to provide a method for effectively removing interference signals from a received signal by preventing performance degradation caused by timing offset.

In order to achieve the above object, a timing offset compensation method of an interference cancellation receiver according to the present invention includes estimating a first timing offset and a first SNR for a training signal of an initial received signal; Training coefficients of a feedforward filter (FF) and a feedback filter (FB) using a predetermined training signal; Filtering the training signal portion of the received signal by applying the trained coefficients to estimate a second timing offset and a second SNR of the training signal from which the interference signal has been removed; Comparing the first and second timing offsets with each other to determine an error of a timing offset estimate value; And correcting the data of the received signal with the second timing offset value if there is an error in the timing offset estimation.

Preferably, the removal of the interference signal is performed only when the ratio of the first and second SNRs is larger than the threshold value, and the first and second SNRs are estimated from the training signals of the received signal before and after the interference cancellation.

Preferably, the received signal is a normal burst of the GSM signal.

Preferably, when the first and second timing offsets are different from each other, it is determined that there is an error in the timing offset estimate.

The present invention provides a method for preventing performance degradation caused by a timing offset while removing interference from a received signal in a GSM mobile communication terminal. Hereinafter, preferred embodiments of the present invention will be described in detail.

4 is a block diagram of a SAIC receiver according to the present invention.

As shown in FIG. 4, the SAIC receiver according to the present invention is a kind of BIC receiver and is composed of two paths. The first path is a path that does not use Single Antenna Interference Cancellation (SAIC) and is used when it is determined that no interference signal exists. This path uses matched filters and the Maximum-Likelihood Sequence Estimation (MLSE) equalizer, which are known to be optimal in the environment of white additive Gaussian noise.

In addition, the second path includes a FF (feedforward filter) and a FB (feedback filter) as a path used when an interference signal exists and the use of SAIC shows good performance.

The first and second paths are connected to an output of a phase derotator 10. The phase rerotator 10 is a block for transforming a GMSK signal into an antipodal signal and removing differential encoding. Blocks commonly used in.

The cross-correlator 11 takes a cross-correlation between the received signal and a known training sequence, estimates the impulse response, and evaluates the grouped taps. The coefficients are summed to determine the time index with the maximum value as the timing offset.

In addition, the channel estimator 12 obtains a channel estimate using the estimated impulse response, and the signal-to-noise ratio estimator 13 uses the SAIC, that is, the first or the second. Determine whether to use a route. In particular, in FIG. 4, two autocorrelators, two channel estimators, and two SNR estimators exist (including a dashed block), but two blocks are not required to exist, but one block is exemplarily shown to be operated over time. will be.

The FF is in principle a FIR filter having a complex input and a complex coefficient and a real output. Therefore, since the FF outputs only the real part of the complex output value, the complex filter provided can be separated into two independent filters that process the real part and the imaginary part ((a + bj) (c + dj) = (ac -bd) (bc + ad) j). FB is a FIR filter with real input, real coefficients and real output. The operation of the FB is the same as in the existing invention. That is, the input of the FB uses a known training signal in the training section, and uses the data value after the determination in the data section.

The timing offset compensation method according to the present invention in the SAIC receiver configured as described above is as follows.

6 is a flowchart illustrating a timing offset compensation method of a SAIC receiver according to an embodiment of the present invention.

First, the phase rerotator 10 of the SAIC receiver receives a received signal (GSM burst) to perform phase derotation (S10). The cross correlator 11 cross-correlates the phase re-rotated received signal with a known training signal to estimate an impulse response, and as shown in FIG. 6, a grouped tap ( The coefficients of the taps are added together to determine a time index having the maximum value as the timing offset τ _f . In FIG. 6, the timing offset τ _f ) is determined as 5 since the time index at which the 5-tap coefficients are maximum is 5.

In addition, a channel estimator 12 obtains a channel estimate using the estimated impulse response, and a signal-to-noise ratio (SNR) estimator 13 calculates the channel estimate and the phase reconstruction. SNR _f using training signal of rotated received signal Estimate (S11).

Once the SNR _f is estimated, the SAIC receiver trains (finds) the coefficients of the FF 14 and the FB 15 using the training signals of the received GSM bursts. In this case, the FF and FB coefficients are set such that the error signal e [k] is minimized. Where u (n) represents the delay lines of FF and FB, w (n) represents the vector of FF and FB, and L _f and L _b represent the length of FF and FB, respectively. . Adaptive algorithms used for the training of the FF and FB coefficients include LMS or RLS, and hierarchical RLS (HRLS) or fast RLS (Fast RLS) can be used to further reduce the amount of computation (S12). ).

When the FF and FB coefficients are trained by performing the above process (obtained), the cross correlator 11 filters (DFE) only the training signal portion of the output signal of the FF (a signal received by applying the trained coefficients). The SNR _b and the timing offset τ _b ) of the training signal from which the interference signal is removed are estimated (S13).

When the SNR value SNR _b and the timing offset τ _b ) are estimated, the determiner 16 compares the ratio of SNR _f and SNR _b with a threshold to determine the presence or absence of an interference signal. If the ratio is greater than the threshold value, the switch SW1 is controlled to operate the first path and DFE is performed on the data.

Further, the determiner 16 compares the timing offset (τ _f )) value before the interference cancellation with the timing offset (τ _b ) value after the interference cancellation to determine whether the timing offset estimate is in error. If an error exists due to different offset values, the data is finally corrected for the timing offset τ _b ) after interference cancellation.

On the other hand, if the ratio is smaller than the threshold value in step S14, the controller controls the switch SW1 so that the received signal is input to the second path not using the SAIC, so that the received signal is matched with the matching filter 17 and Output through the MLSE equalizer 18.

Therefore, the reception performance is improved by transmitting the output of the first path or the output of the second path in which the interference is eliminated and the error of the timing offset estimation is corrected and transmitted to the muxing chain (interleaver, channel decoder, etc.).

As described above, the present invention can effectively solve the problems that may occur in the existing invention by determining whether the timing offset estimate is an error and taking a correction process when an error exists.

Although the present invention has been described with reference to the embodiments shown in the drawings, this is merely exemplary, and it will be understood by those skilled in the art that various modifications and equivalent other embodiments are possible. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

Claims (8)

In the interference cancellation receiver having a feedforward filter (FF) and a feedback filter (FB), to remove interference from the received signal, Estimating a first timing offset and a first SNR for the training signal of the received signal; Training coefficients of FF and FB using a predetermined training signal; Filtering the training signal portion of the received signal by applying the trained coefficients to estimate a second timing offset and a second SNR of the training signal from which the interference signal has been removed; Comparing the first and second timing offsets to determine an error of a timing offset estimate value; And correcting data of a received signal with a second timing offset value if there is an error in the timing offset estimation. The method of claim 1, wherein the interference signal is removed. The timing offset compensation method, characterized in that only when the ratio of the first, second SNR is greater than the threshold value. The method of claim 1, wherein the first and second SNRs are A timing offset compensation method characterized in that it is estimated from a training signal of a received signal before and after interference cancellation, respectively. The method of claim 1, wherein the received signal is Timing offset compensation method characterized in that the normal burst (normal burst) of the GSM signal. The method of claim 1, wherein the first and second timing offset The timing offset compensation method, characterized in that it is determined that there is an error in the timing offset estimation value. The method of claim 1, wherein the interference cancellation receiver  A timing offset compensation method characterized in that the blind interference cancellation receiver or a single antenna interference cancellation receiver. The method of claim 1, wherein the coefficients of FF and FB are Using a training interval of the received signal, the timing offset compensation method characterized in that obtained by applying Recursive least-square (LLS) and Least means-square (LMS). 8. The method of claim 7, wherein the coefficients of FF and FB are And a hierarchical RLS or a fast RLS, using the training interval of the received signal, to obtain the timing offset compensation method.

Images