KR20040046499A - A method of improved beam forming for smart antenna in base station - Google Patents

Abstract

PURPOSE: A method for forming an improved beam of a base station smart antenna is provided to make a value of an error signal be a minimal, and to update the minimal value with an optimal coefficient vector value, thereby improving reception performance of a blind series. CONSTITUTION: A system initializes a value of a coefficient vector set in a base station beam former(S100), and inputs a signal applied in CDMA method(S110). The system multiplies the initialized coefficient value by the inputted signal, and outputs the multiplied signal(S120). The system divides the outputted signal with a power ratio of a number of a smart antenna and a control channel, and extracts a pure control signal(S130). The system offsets the pure control signal from a normalized value, and calculates an error value(S140). The system conjugates the error value, adds the coefficient vector value to a multiplied value of a reception signal input value and a convergent constant, and obtains an updated coefficient vector value(S150). The system divides the updated coefficient vector value, and calculates a normal coefficient vector(S160).

Description

Improved beamforming method of base station smart antenna {A METHOD OF IMPROVED BEAM FORMING FOR SMART ANTENNA IN BASE STATION}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for forming a smart antenna receive beam (BEAM) in an extended code division multiple access (WCDMA) mobile communication base station system, and more particularly, to improve the performance of forming a blind series receive beam that does not require a reference signal. It is about a method.

Code division multiple access mobile communication system is a method to occupy a limited number of radio frequency resources to use for communication, and a mobile terminal (UE) registered and subscribed to the system is a service area (SERVICE AREA). Communicate with your opponent instantly anytime, anywhere while on the go.

The service area is a range in which a wireless signal transmitted from a base station (BS) antenna (ANT: ANTENNA) is transmitted to a minimum effective level (LEVEL) that a mobile terminal can receive and process. Also referred to as a cell (CELL) area, the mobile communication system is to receive a mobile communication service anywhere in the country by covering the whole country (COVER) a plurality of the continuous cells, as described above, a plurality of cell areas are like a honeycomb It is also called a cellular system because it is continuous.

Since the wireless signal output from the portable terminal is transmitted at a very weak level, the base station BS receiving the signal should increase the gain GAIN of the received signal, and among the various methods of increasing the gain as described above, There is a method of increasing the wireless reception gain through, and one of the plurality of antennas are arranged at a predetermined interval of the reference multiple multipled by the wavelength (λ), the plurality of antennas of the directional or beam to receive the signal The SMART antenna can be used to arbitrarily control the direction along with the width WIDTH.

A program or a task that arbitrarily adjusts the direction and width of a reception beam without using a smart antenna as described above and using a reference signal is called a beamformer or a CMA algorithm, called a CONSTANT MODULUS ALGORITHM algorithm. In the case of receiving and analyzing a signal at a receiving side, a method of not sharing information about a position at which a signal starts or a reference signal with a transmitting side is called a blind system, and the blind system as described above. In the CMA algorithm, the factor that forms the reception beam in the desired direction and width is called the coefficient vector, and the error or error of the coefficient vector as described above is minimized in terms of MMSE (MINIMUM MEAN SQUARED ERROR). Development of technology is necessary.

Hereinafter, a beamforming method of a base station smart antenna according to the prior art will be described with reference to the accompanying drawings.

Attached to describe the prior art, FIG. 1 is a functional explanatory diagram of a general mobile communication base station smart antenna, and FIG. 2 is a flowchart of a smart antenna beam forming method according to the prior art.

Referring to FIG. 1, a general mobile communication base station smart antenna is registered and subscribed to a mobile switching center, and performs first and second mobile terminals (UE: USER EQUIPMENT) 10 to communicate wirelessly with a counterpart anytime and anywhere while moving. And 15, which are wirelessly connected to the mobile terminals 10 and 15 to transmit and receive a signal, thereby selectively receiving a wireless signal of the first mobile terminal 10 or the second mobile terminal 15. A base station (BS) 30 for controlling the smart antenna 20 and the smart antenna 20 and simultaneously processing radio signals transmitted and received with the portable terminals 10 and 15 in a code division multiple access scheme; It is composed of a mobile switching center (MSC) 40 connected to the base station 30 to monitor and control the entire mobile communication system and to process the requested call by SWITCHING.

Referring to the configuration of the general mobile communication base station smart antenna in more detail, the smart antenna 20 is controlled by a CMA algorithm provided in the base station, the smart antenna 20 in the cell area for transmitting and receiving radio signals Among a plurality of portable terminals located, when a signal of a specific portable terminal, for example, the first portable terminal 10 is to be received, a first reception beam 22 is formed, and the signal of the second portable terminal 15 is received. If desired, the second reception beam 24 is formed.

The reception beams 22 and 24 as described above are formed by the control of the CMA algorithm embedded in the base station 30, and the signals of the mobile terminals 10 and 15 are increased by accurately forming the direction and width of the reception beams. It receives the gain GAIN and transfers it to the mobile switching center 40, and the portable terminals 10 and 15 continue to move so that the continuous control is performed so that the direction of formation of the reception beam is also optimal.

As described above, the beamforming of the smart antenna 20 to be in an optimal state without a reference signal is a CMA algorithm that finds a coefficient vector in terms of MMSE (MINIMUM MEAN-SQUARED ERROR).

Referring to FIG. 2, the smart antenna beam forming method according to the related art includes a first step S10 of initializing a coefficient vector WE _{k, p} and a smart antenna 20. A second step (S20) of inputting a signal (V _{k, p, Q} (m)) of a symbol (SYMBOL) unit received through a code division multiple access method and a code division multiple access method, and the input signal and the coefficient value The third step S30 of outputting the signal (Y _{k, p, Q} (m)) multiplied by and the error value (e _{k, p} (m)) is calculated using the output signal of the step (S30). The fourth step S40 and the error values e _{k and p} (m) according to the step S40 are conjugated (e ^* _{k, p} (m)) to receive the received signal V _{k, p, Q} (m)) and a convergence constant (μ) and the multiplied coefficient vector values (W _{k, p} (m) of 2-fold) and more hameuroseo newly updated (uPDATE) coefficient values (W _{k, p} (m +1)) and a fifth step S50 for feeding back to the third step S30.

Referring to the above-described conventional method in more detail, the code division multiple access (CDMA) mobile communication base station 30 is for receiving signals from the mobile terminals 10 and 15 using the smart antenna 20. Initializes the coefficient vector WEIGHT VECTOR of the beamformer or CMA algorithm (S10), and receives the signal received through the smart antenna 20 in the state where the coefficient vector WEIGHT VECTOR is initialized as described above. (S20).

As described above, the signal received and input through the smart antenna 20 is multiplied by the normalized coefficient vector value of the beamformer or the CMA to output the vector VECTOR in the desired direction, that is, the received signal in the desired direction. It is estimated by calculating using Equation 1 as follows (S30).

Y _{k, p, Q} (m) = W ^H _{k, p} (m) V _{k, p, Q} (m)

That is, assuming the N antenna array of the smart antenna 20, k-th user, m-th despread N * 1 DPCCH symbol vector V _{k, p, Q} (m) of the p-th path, the previous step The beamforming output signal Y _{k, p, Q} (m) is made by multiplying the N * 1 coefficient vectors W _{k, p} (m) obtained from.

In order to exactly match the direction of the received signal estimated as described above, that is, to reduce the error, the error value is measured by calculating the following Equation 2 (S40).

e _{k, p} (m) = {Y _{k, p, Q} (m) / | Y _{k, p, Q} (m) | }-Y _{k, p, Q} (m)

That is, since there is no reference signal _, the value obtained by normalizing Y _{k, p, Q} (m) {Y _{k, p, Q} (m) / | Y _{k, p, Q} (m) | } Is used as a reference signal value, and the error signals e _{k and p} (m) are calculated and obtained.

The third error value after the error correction is calculated using the following equation (3), which is multiplied by the received signal with the convergence constant (μ) and added with the previous coefficient value. By returning to step S30, the above steps are repeated, and the received signal in the desired spatial direction is accurately extracted without error.

W _{k, p} (m + 1) = W _{k, p} (m) + 2 μV _{k, p, Q} (m) e ^* _{k, p} (m)

That is, the coefficient vector (WEIGHT VECTOR) is updated by using the error signal value calculated by Equation 2 and the received signal input value input to the beam former or the CMA algorithm.

However, the prior art using the above method has a problem that the magnitude of the signal (Y _{k, p, Q} (m)) output through the beamformer or the CMA is up to four times larger than that of the reference signal compared to the reference signal. Therefore, there is a problem of balancing that the value of 0 is not calculated when Equation 2 is calculated.

That is, since the NORM value, which is the value obtained by adding each square value of the coefficient vector after the update by Equation 3 and taking the square root, does not become 1, the error signal is 0 even though the beamformer forms an optimal beam. There is a problem in that it has a constant bias value without convergence and the above value is reflected in the coefficient update equation of Equation 3, so that the optimal reception beam coefficient vector is not efficiently updated.

The present invention provides a method for improving the reception performance of the blind system by updating the optimal coefficient vector value by minimizing the value of the error signal in the beamformer or the CMA algorithm for forming the reception beam of the smart antenna. For that purpose.

In order to achieve the above object, the present invention provides a base station smart antenna beam forming method; An input process of initializing a coefficient vector and inputting a despread code division multiple access method received signal; A detection step of multiplying the coefficient vector value of the process by the input signal and dividing the multiplied value by the power ratio of the control signal and the number of smart antennas to output a normal signal of the control signal; An error output process of calculating an error value by subtracting a normal signal output value of the process from a normalized value; The updated coefficient vector value is obtained by using the error value, the received signal input value, and the coefficient vector value obtained by the above process, divided by the norm value, and the normal coefficient vector value is obtained. .

1 is a functional explanatory diagram of a general mobile communication base station smart antenna;

2 is a flowchart of a method for forming a smart antenna beam of the prior art;

3 is a flowchart of an improved beamforming method of a base station smart antenna of the present invention;

4 is an output curve diagram improved by the beamforming method of the present invention.

** Explanation of symbols on the main parts of the drawing **

10,15: mobile terminal 20: smart antenna

22: first receiving beam 24: second receiving beam

30 base station 40 mobile switching center

Hereinafter, an improved beamforming method of a base station smart antenna according to the present invention will be described with reference to the accompanying drawings.

3 is a flowchart illustrating an improved beamforming method of a base station smart antenna according to the present invention, and FIG. 4 is an output curve diagram improved by the beamforming method of the present invention.

Referring to FIG. 3, an improved beamforming method of a base station smart antenna according to the present invention includes initializing a coefficient vector and inputting a code division multiple access (CDMA) received signal. Initializing a value of a coefficient vector set in the base station beamformer (S100); An input process comprising a process of inputting a signal received and applied in a code division multiple access method (S110);

Multiplying the coefficient vector value of the input process (S100, S110) and the input received signal, and outputs a normal signal consisting of a pure control signal by dividing the multiplied by the power ratio (βc) and the smart antenna number (N) of the control signal By doing so, multiplying the signal received by the initialized coefficient value and the code division multiple access method and the input signal (Y _{k, p, Q} (m)) (S120); The pure control signal ( _{Y'k, p, Q} (m)) is divided by dividing the signal (Y _{k, p, Q} (m)) output by the process by the number (N) of the smart antenna and the power ratio (β _c ) of the control channel. A detection process consisting of a normal signal output process (S130) for extracting

An error output process S140 for calculating an error value e _{k, p} (m) by subtracting the normal signal output value _{Y'k, p, Q} (m) from the normalized value;

Error value (e _{k, p} (m)), received signal input value (V _{k, p, Q} (m)) and coefficient vector value (W _{k, p} (m)) by the error output process (S140) The coefficient vector value is updated, and the normal coefficient vector value (W _{k, p} (m + 1) ') is obtained by dividing by the NORM value and feeding back to the average output process. The coefficient vector (W) updated by conjugating the error value (e _{k, p} (m)) by S140 and multiplying the received signal input value with the convergence constant value (μ) by adding the coefficient vector value. obtaining a value of _{k, p} (m + 1)) (S150); The step S160 of dividing the updated coefficient vector value W _{k, p} (m + 1) by the norm value and calculating a normal coefficient vector W _{k, p} (m + 1) '(S160). It consists of a process.

Hereinafter, an improved beamforming method of a base station smart antenna according to the present invention having the above configuration will be described in detail with reference to FIGS. 3 and 4.

The present invention uses despreaded control channel (DPCCH) symbol data for coefficient update of a base station smart antenna that transmits and receives an extended code division multiple access (WCDMA) signal. Modifications to fit the structure improve the performance of the receiver.

According to the present invention, when the conventional CMA algorithm is improved and converged to approach the coefficient vector of the optimal beam, the signal size becomes the same as the noise, meaning that the signal size is four times larger than the conventional one. The beamformer output signal improves the signal-to-noise ratio (SIR or SNR: SIGNAL TO NOISE RATIO) by 6 dB relative to the input signal.

In the conventional CMA algorithm, when an error signal is obtained, a signal that normalizes an input signal of the beamformer to have a magnitude of 1 is used as a reference signal, and a square of an error signal that is a difference between the reference signal and the beamformer input signal is used. The coefficient vector is updated to minimize the error (MEAN SQUARE ERROR) (Equations 1 to 3), but when the output signal size of the CMA beamformer is multiplied by the optimal coefficient compared to the reference signal, the signal becomes four times larger. In addition, the control signal is input by multiplying (SCALE) by βc compared to the data signal during the modulation (MODULATION) of the output stage.

Assuming that βc does not change significantly, the beamformer output signal is normalized using βc and N of smart antennas known before one frame, and then an error signal is obtained. The point of the present invention is to increase the efficiency of coefficient update by minimizing the MEAN SQUARE ERROR value of the error by converging to the noise signal and normalizing it to the NORM of the coefficient after updating the coefficient vector. Experimental results showed excellent performance.

In order to understand the output signal normalization process of the beamformer by the CMA algorithm, a theoretical first look at what values the despread DPCCH symbol signal contains.

The data channel (DPDCH) signal D and the control channel (DPCCH) signal C at the transmitting end of the base station (BS) 30 or the mobile terminal (UE) 10, 15 are β _d and for respective power (POWER) balancing. By multiplying β _c to form a QPSK (QUADRATURE PHASE SHIFT KEYING) signal, multiplying by OVSF (ORTHOGONAL VARIABLE SPREADING FACTOR) code, SPREADING, and then scrambling with COMPLEX SCRAMBLING CODE WCDMA) signal.

The transmitted signal is transmitted through MULTI-PATH by fading in the air, and the noise is input with THERMAL NOISE added to the receiver.

Assuming the K-th user (USER), the P-th Rayray (FALE) fading path (FADING PATH), the received signal is expressed by the following equation.

X _{k, p} (m) = (β _{d, k} C _{ch, d, k} D _{k, p} (m) + jβ _{c, k} C _{ch, c, k} C _{k, p} (m)) C _{scr, k} h (t-τ _{k, p} ) a (θ _{k, p} ) + n _{k, p} (m)

In the above _description, C _{ch, d, k} and C _{ch, c, k} denote the k'th USER CHANNELIZATION CODE of the data channel DATA CHANNEL and CONTROL CHANNEL, respectively, and C _{scr, k} denotes the COMPLEX SCRAMBLE CODE, and h (t-τ _{k, p} ) = α _{k, p} e ^{jφk, p} σ (t-τ _{k, p} ) denotes the propagation delay τ _k, pointing to the ray ray fading channel having a _{p, a (θ k, p} ) = [1 exp (j2π (d / λ) sin (θ k, p)) λ exp (j2π (N-1) (d / λ) sin (θ _{k, p} ))] ^T denotes an array response vector (ARRAY RESPONSE VECTOR) at the number N of antennas, and n denotes ADAPTIVE WHITE GAUSSIAN NOISE (AWGN).

The input signal of the beamformer CMA algorithm is simply a signal which is despread with a user's channelization code and a scramble code and is expressed as follows.

V _{k, p, Q} (m) = jβ _{c, k} C _{k, p} (m) h (t-τ _{k, p} ) a (θ) + n '

In general, the output signal of the beamformer CMA algorithm is expressed by the following equation.

Y _{k, p, Q} (m) = V _{k, p, Q} (m) W ^H _{k, p} (m)

The ideal output signal of the beamformer CMA algorithm is expressed by the following equation.

| Y _{k, p, Q} (m) | = N β _{c, k} | h (t-τ _{k, p} ) |

The absolute value is a case where noise is ignored.

The present invention is to minimize the error signal in the process of making a reference signal using the output signal of the beamformer CMA algorithm as described above, and to normalize the beamformer output signal to a known value in the conventional CMA algorithm.

That is, in Equation 7, since N and β _{c, k} values are known in advance, the error signals are obtained in the following order after normalizing by N and β _{c, k} .

In addition, after the beamforming coefficient update equation is performed, the value is normalized to the NORM value of the coefficient vector, thereby increasing the signal-to-noise ratio (SNR) without changing the magnitude of the output signal of the beamformer.

Referring to FIG. 3, the CMA algorithm of the base station BS initializes the coefficient vector WEIGHT VECTOR value (S100), and receives the spread code division multiple access (WCDMA) scheme signal V _{k. , p, Q} (m)) and multiply the coefficient vector value initialized as described above with the received signal and output (Y _{k, p, Q} (m)) (S120).

Equation for outputting a signal in the process (S120) is as follows.

Y _{k, p, Q} (m) = W ^H _{k, p} (m) V _{k, p, Q} (m)

The signal output as above is divided by the base station smart antenna 20 number N and the power ratio value β of the control channel to obtain a normal output value (S130), and the equation by the process (S130) is as follows.

Y ' _{k, p, Q} (m) = Y _{k, p, Q} (m) / (N β)

Subtracting from the normalized value using the normal output value obtained in the step (S130) calculates the improved error value, the above equation is as follows.

e _{k, p} (m) = {Y ' _{k, p, Q} (m) / | Y' _{k, p, Q} (m) | }-Y ' _{k, p, Q} (m)

As obtained in the step S140, the improved error value e _{k, p} (m) is conjugated (*), multiplied by the input signal and the convergence constant (μ), and added to the coefficient vector value, thereby updating ( The UP-DATE) coefficient vector value is calculated (S150), and the equation for obtaining the updated coefficient vector value is as follows.

W _{k, p} (m + 1) = W _{k, p} (m) + 2 μV _{k, p, Q} (m) e ^* _{k, p} (m)

Using the updated coefficient vector values W _{k, p} (m + 1) obtained by the above process (S150), the result of taking the root value after adding each of the squares and adding them, || W _{k, p} After dividing by the NORM value of (m + 1) ||), the normal coefficient vector value W _{k, p} (m + 1) 'is obtained (S160), and then fed back to the detection process (S120). The process of repeating the steps to be further improved, and to obtain the constant coefficient vector value (W _{k, p} (m + 1) ') as described above.

W _{k, p} (m + 1) '= W _{k, p} (m + 1) / W _{k, p} (m + 1) ||

The output curve of the CMA algorithm applying the method according to the present invention having the above configuration as an experimental example is shown in FIG. 4, which will be described in detail with reference to BER (BIT ERROR RATE) 1e-3 (10 ^-3 ). As a result, it can be seen that the performance of about 2dB is improved, and that the array response vector (ARRAY RESPONSE VECTOR) value can be known in advance, and there is an improvement effect of about 0.5 dB at a low signal-to-noise ratio (SNR). Higher signal-to-noise ratio (SNR) results in nearly similar results.

The present invention having the configuration described above has an industrial use effect of improving the reception performance of the receiver by complementing the CMA algorithm using the base station smart antenna.

In addition, it can be easily applied to the signal of the blind (BLIND) series that does not require a reference signal, and there is an industrial use effect of improving the beamforming function of the smart antenna while using the conventional CMA algorithm as it is.

In addition, since the reception function is improved, the communication quality of the subscriber is improved and there is a convenient effect of using the communication signal without error transmission.

Claims (4)

In the base station smart antenna beamforming method, An input process of initializing a coefficient vector and inputting a despread code division multiple access method received signal; A detection process of multiplying a coefficient vector value of the process by an input signal and dividing the multiplied value by a power ratio of a control signal and a number of smart antennas to output a normal signal of the control signal; An error output process of calculating an error value by subtracting a normal signal output value of the process from a normalized value; The updated coefficient vector value is obtained by using the error value, the received signal input value, and the coefficient vector value, and the normal coefficient vector value is obtained by dividing by the norm value. Improved beamforming method of base station smart antenna. The method of claim 1, wherein the input process, Initializing a value of a coefficient vector set in the base station beamformer; Improved beamforming method of a base station smart antenna, characterized in that the step of inputting a signal received and applied in a code division multiple access method. The method of claim 1, wherein the detecting process comprises: Multiplying and outputting the initialized coefficient value and the despread input signal by the code division multiple access method; Improved beamforming method of the base station smart antenna, characterized in that the normal signal output process for extracting the pure control signal by dividing the signal output in the process by the number of the smart antenna and the power ratio of the control channel. The method of claim 1, wherein the updating process, Obtaining an updated coefficient vector value by conjugating the error value obtained by the error output process and adding the coefficient vector value to a value obtained by multiplying a received signal input value and a convergence constant value; And calculating a normal coefficient vector by dividing the updated coefficient vector value by a norm value.