KR20060123263A - Noise variance estimation method and apparatus, user equipment - Google Patents

Abstract

A method of noise variance estimation performed by user equipment is proposed, the method comprising receiving a signal vector and a noise vector comprising a training sequence transmitted over at least one propagation path from a base station, wherein Estimating the channel impulse response of the propagation path to construct a channel impulse response matrix and noise of the signal vector according to the channel impulse response matrix and signal vector when the channel impulse response remains largely unchanged for a particular period of the training sequence. Calculating the variance.

Description

METHOD AND APPARATUS OF NOISE VARIANCE ESTIMATION FOR USE IN WIRELESS COMMUNICATION SYSTEMS

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention generally relates to methods and apparatus for noise variance estimation used in wireless communication systems, and more particularly to methods and apparatus for noise variance estimation using a training sequence.

CDMA (Code Division Multiple Access) is a new wireless communication technology developed after FDMA (Frequency Division Multiple Access) and TDMA (Time Division Multiple Access). In CDMA wireless communication, different user equipments (UEs) are assigned different vertical spreading codes, and spread signals by different UEs having different spreading codes can be transmitted on the same frequency band.

The CDMA downlink transmission model is shown in FIG. 1 and proposed in A.Klein, "Data Detection Algorithms Specially Designed For The Downlink of CDMA Mobile Radio Systems", VTC, 1997. Signal vector d ⁽¹⁾ , ..., d ^(k) , ..., d ^(K) , where d ^(k) (k = 1 ... K) consists of N complex components In order to transmit to UE1, ..., UEk, ... UEK, respectively, the base station 200 first transmits a spreading code ( c _d ⁽¹⁾ , ... ^{_{, c d (k), ...}} , c d (k)) and vector signal ^{(d (1),} using a ^{..., d (k), ...} , and diffusing the d ^(k)) that This spreading signal vector is then combined with the signal vector s _d and transmitted to each corresponding UE 220 over the same channel 210. The signal vector s _d reaches the UE _k (k = 1 ... K) through the multi propagation path and the CIR (channel impulse response) of each propagation channel is h _{d (i)} ^(k) (i = 1, 2, ...), the signal vector ( e _d ^(k) ) received by the UE _k can be represented by the following equation (1). In other words,

(1)

(One)

Where H _d ^(k) is a CIR matrix composed of CIR h _{d (i)} ^(K) (i = 1,2, ... ⁾ of each propagation channel, and C _d is a spreading code ( c _d ⁽¹⁾ , is a spreading code matrix composed of ..., c _d ^(k) , ..., c _d ^(K) ), and the method of constructing (( H _d ^(k) and C _d ) D ) ( d ^{(1) T} , ..., d ^{(k) T} , ..., d ^{(K) T} ) ^T , where [.] ^T represents the matrix binomial, s _d Denotes a signal vector obtained after d is spread and combined, s _d = C _d d , and n _d ^(k) is a noise vector.

Equation (1) indicates that the received signal vector e _d ^(k) includes the desired signal vector d ^(k) of UE _k and the signal vector and noise vector transmitted by the base station to another UE.

In order to allow UE _k to obtain the desired signal vector d ^(k) with the minimum error from the received signal vector e _d ^(k) , a number of signal reception methods have been provided, which are described by Kimmo Kettunen. "Zero Forcing an Mininum Mean-Square-Error Equalization for Multiuser Detection in Code-Division Multiple-access channel", IEEE Transaction on by "Iterative Multiuser Receiver / Decoders with Enhanced variance Estimation", VTC, 1999, and A.Klein Vehicular Technology, vol. 45, pp. 276-287, May 1996. Both of these signal reception methods rely heavily on channel information, ie, noise variance, to obtain the desired signal vector from the received signal vector, so the noise variance needs to be accurately calculated to obtain the desired signal vector with the least error.

In order to obtain accurate noise variance, various noise estimation methods have been proposed. For example, conventional variance estimation techniques used in AWGN channels are described in "A novel variance estimator for turbo-code decoding" by M. Reed and J. Asenstorfer, Proc. Of ITC'97, pp173-178, April 1997, a rake technique for mitigating multipath interference is disclosed in US Patent US200220110199 entitled "Method for Noise Energy Estimation in TDMA Systems." There is also a predetermined noise estimation method for calculating noise variance by convolving training sequences. These noise estimation methods can meet the exact requirements of 2G wireless communication systems.

However, in 3G wireless communication systems, more accurate noise variance is required for signal reception, for example, the core technology of both multi-user detection and turbo-code has a high demand for accurate noise variance. Current noise estimation methods cannot satisfy the exact demand for noise variance of 3G wireless communication systems.

It is an object of the present invention to provide a method and apparatus for noise variance estimation for use in a wireless communication system, in which a training sequence is used to calculate the noise variance to obtain more accurate noise variance.

SUMMARY OF THE INVENTION The present invention proposes a method of noise variance estimation for use in a wireless communication system, the method comprising receiving a signal vector and a noise vector comprising a training sequence transmitted over at least one propagation path from a base station; Estimating the channel impulse response of each propagation path according to the vector to construct a channel impulse response matrix, and if the channel impulse response remains largely unchanged for a particular period of the training sequence, Calculating a noise variance of the signal vector.

1 is a diagram illustrating a conventional CDMA downlink transmission model,

2 is a flowchart illustrating a noise variance estimation method of the present invention;

3 is a block diagram showing a UE equipped with a noise variance estimating apparatus according to an embodiment of the present invention;

4 is a block diagram showing an apparatus for estimating noise variance according to an embodiment of the present invention.

Hereinafter, an embodiment of the present invention will be described in detail using TD-SCDMA.

In TD-SCDMA, the base station sends a signal vector to each UE in the corresponding timeslot. According to the timeslot format of TD-SCDMA, the signal vector transmitted by the base station to each UE in the corresponding timeslot consists of a training sequence and a spreading user signal.

Regarding the UEs assigned in the same timeslot, the base station first combines the signal vectors to be transmitted to each UE into one combined signal vector, and then transmits the combined signal vectors to each UE in that timeslot. The combined signal vector also consists of a user signal and a training sequence, wherein the user signal in the combined signal vector is obtained by combining spreading user signals in a signal vector to be transmitted to each UE, wherein the training sequence in the combined signal vector is obtained from each UE. It is obtained by combining the training sequence into a signal vector to be sent to.

The training sequence assigned to each UE in one cell is obtained by performing a different shift operation on the same basic training sequence, so the training sequence of the combined signal vector can be considered as the basic training sequence. Each UE obtains the basic training sequence used by its cell during the cell search process, so that the training sequence transmitted by the base station in the timeslot is known in advance to each UE.

The training sequence included in the signal vector transmitted by the base station in one timeslot arrives at the UE via at least one propagation path, and the signal vector received by the UE in that timeslot is the training sequence and noise vector (n). R, and the known value of the training sequence is s. According to equation (1), the signal vector r can be expressed as follows.

(2)

(2)

Here, H is a CIR matrix composed of CIRs of respective propagation paths between the UE and the base station.

)은 다음과 같이 표현될 수 있다."Low Cost Channel Estimation in the uplink receiver of CDMA mobile radio systems" by B. Steiner and PWBaier, Frequenz, vol. 47, pp. 292-298, Nov./Dec. According to the channel estimation method disclosed in 1993, the maximum probability estimation value of the training sequence included in the signal vector r (

) Can be expressed as

(3)

(3)

Here, the superscript H denotes a complex conjugated binomial.

From equation (3), according to the known value of the training sequence included in the signal vector r, the estimated value n 'of the noise vector n can be given as follows.

(4)

(4)

The covariance matrix is

(5)

(5)

)을 계산하는 후속하는 공식이 쉽게 도출된다.Where E {.} Represents the expected value operation. By performing the calculation of the matrix trace between the two sides of equation (5), the mean variance of the estimated value n 'of the noise vector n

The following formula for computing () is easily derived.

(6)

(6)

Where N is the chip duration of the training sequence, operator (trace (.)) Means the calculation of the matrix trace, and σ ² is the variance of the signal vector (r).

이 종래의 방법으로 계산되는 경우, 그것은 매우 복잡할 것이다. 사실, 분산(

)의 계산은 채널이 일정한 것으로 간주될 수 있으면 하나의 트레이닝 시퀀스 시간 기간에 위치한 잡음 벡터(n)의 추정된 값(n')에 대해 모든 요소의 평균 제곱 값을 계산함으로써 근사화될 수 있다. 신호 벡터(r)의 잡음 분산(σ²)은 이제 다음과 같이 추론될 수 있다.

If it is calculated by the conventional method, it will be very complicated. In fact, variance (

Can be approximated by calculating the mean squared value of all elements against the estimated value n 'of the noise vector n located in one training sequence time period if the channel can be considered constant. The noise variance σ ² of the signal vector r can now be inferred as follows.

(7)

(7)

의 평균을 타임슬롯에서의 신호 벡터(r)의 잡음 분산(σ²)으로서 취급할 수 있다.For estimating further improve performance, in a time slot formula averaged by summing the noise variance (σ ²⁾ calculated from the noise variance (σ ²⁾ and the formula (7) in each of the previous time slot calculated from (7) and , Different

Can be treated as the noise variance σ ² of the signal vector r in the timeslot.

In the above description, the principle of calculating the noise variance using the training sequence has been described.

Hereinafter, the proposed noise variance estimation method will be described in detail with reference to FIG. 2.

Firstly, the UE receives a signal vector and a noise vector including a training sequence transmitted through at least one propagation path in one timeslot from the base station (step S10).

Secondly, the UE estimates the CIR of each propagation path according to the received signal vector and constructs a CIR matrix H using the estimated CIR of each propagation path (step S20).

)을 추정한다(단계 S30).Thirdly, the UE uses the equation (3) according to the signal vector and the CIR matrix to estimate the maximum probability estimated value of the training sequence included in the signal vector.

) Is estimated (step S30).

Fourthly, the UE uses the equation (4) according to the MLE (maximum probability estimation) value of the training sequence included in the signal vector and the known value of the training sequence to estimate the estimated value of the noise vector included in the signal vector ( n ') is calculated (step S40). Here, the known value of the training sequence included in the signal vector is obtained by the UE during cell search.

Fifth, the UE calculates the noise variance σ ² of the signal vector using equation (7) according to the estimated value n 'of the signal vector and the noise vector included in the CIR matrix H ( Step S50). Here, first, n 'squared p _n ² can be calculated according to the equation p _n ² = (n') ^H (n '), and then the matrix trace (cf) can be calculated ((H ^H H ) Can be calculated, cf = trace ((H ^H H) ^-1 ), finally, the noise variance (σ ² ) is calculated according to the equation σ ² = p _n ² / cf, ie equation (7) Can be.

의 평균을 타임슬롯에서의 신호 벡터(r)의 잡음 분산(σ²)으로서 취급한다.Finally, UE is averaged by summing the formula (7), the noise variance (σ ²⁾ and the noise variance (σ ²⁾ calculated from the formula (7) in each of the previous time slot calculated from the in-time slot, a different

The average of is treated as the noise variance σ ² of the signal vector r in the timeslot.

Hereinafter, the proposed noise variance estimating apparatus will be described in detail with reference to FIGS. 3 and 4.

3 is a block diagram illustrating a UE equipped with a proposed noise variance estimating apparatus. As shown in FIG. 3, in the cell search procedure before the UE communicates with the base station, the cell search means 40 obtains the basic training sequence used by the cell in which the UE is located. When the UE communicates with the base station, the antenna of the UE first transmits the signal vector Rx received in the timeslot to the multiplier 10, and the multiplier 10 receives the RF carrier generated by the VCO 20. The signal vector Rx is converted into a baseband signal vector by multiplying the signal vector Rx. The ADC 30 then converts the baseband signal vector r output from the multiplier 10 into a digital baseband signal vector r. Then, the cell searching means 40 synchronizes the digital baseband signal vector r output from the ADC 30, and the channel estimating means 50 performs the CIR of each propagation channel for the synchronized digital baseband signal vector. Is calculated using a conventional channel estimation method, and a CIR matrix is constructed using the calculated CIR of each propagation path. Next, the noise variance estimating means 60 is obtained by the CIR matrix calculated by the channel estimating means 50, the digital baseband signal vector r output by the ADC 30, and the cell searching means 40. The noise variance of the digital baseband signal vector r is calculated using the basic training sequence. Finally, the data detecting means 70 uses a digital basis according to the noise variance calculated by the noise variance estimating means 60, using a conventional data detection method, for example a multi-user detection method, turbo-code decoding or the like. A desired user signal is obtained from the band signal vector r.

), 및 기본 트 레이닝 시퀀스(s)(즉, 상기 디지털 기저대역 신호 벡터(r)에 포함된 트레이닝 시퀀스의 공지된 값)에 따라 상기 디지털 기저대역 신호 벡터(r)에 포함된 잡음 벡터의 추정된 값(n')을 수식(4)을 사용하여 계산하는 잡음 추정 수단(602)과, 잡음 추정 수단(602)에 의해 계산된 상기 디지털 기저대역 신호 벡터(r)에 포함된 잡음 벡터의 추정된 값(n')에 따라 상기 잡음 벡터의 추정된 값(n')의 제곱 p_n ²을 수식 p_n ²=(n')^H(n')를 사용하여 계산하는 잡음 제곱 계산 수단(603)과, 매트릭스 ((H^HH)^-1)의 트레이스(cf)를 계산하는, 즉 cf=trace((H^HH)^-1)를 계산하는 등화 교정 수단(604)과, 잡음 제곱 계산 수단(603)에 의해 계산된 상기 잡음 벡터의 추정된 값(n')의 제곱 p_n ²과 등화 교정 수단(604)에 의해 계산된 트레이스(cf)에 따라 잡음 분산(σ²)을 수식 σ²=p_n ²/cf를 사용하여 계산하는 잡음 제곱 교정 수단(605)을 포함한다.4 is a block diagram showing the noise variance estimating means 60. As shown in FIG. Referring to FIG. 4, the noise variance estimating means 60 is based on the CIR matrix H calculated by the channel estimating means 50 and the digital baseband signal vector r output by the ADC 30. Equalizing means 601 for estimating the MLE value of the training sequence included in the band signal vector r using equation (3), and the digital baseband calculated by the equalization means (601). MLE value of the training sequence contained in the signal vector (r)

) And the noise vector contained in the digital baseband signal vector r according to a basic training sequence s (i.e., a known value of the training sequence included in the digital baseband signal vector r). Noise estimating means 602 for calculating the estimated value n 'using Equation 4, and noise vector included in the digital baseband signal vector r calculated by the noise estimating means 602. Noise square calculation means for calculating a square p _n ² of the estimated value n 'of the noise vector according to the estimated value n' using the formula p _n ² = (n ') ^H (n') 603, equalization correction means 604 that calculates the trace cf of the matrix ((H ^H H) ⁻¹ ), that is, cf = trace ((H ^H H) ⁻¹ ), and noise square calculation The noise variance σ ² is calculated according to the square p _n ² of the estimated value n 'of the noise vector calculated by the means 603 and the trace cf calculated by the equalization correction means 604. ² = p _n ² / cf It comprises a calibration means for calculating squared noise with 605.

Advantage of the present invention

As described above, in the proposed noise variance estimation method and apparatus used in the wireless communication system, a training sequence is used to calculate the noise variance, and thus the calculated noise variance can satisfy the demand of higher accuracy.

Those skilled in the art will appreciate that the methods and apparatus of noise variance estimation used in the wireless communication system disclosed in the present invention may be modified significantly without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims (13)

In the method of noise variance estimation performed by user equipment, (a) receiving a signal vector and a noise vector comprising a training sequence transmitted over at least one propagation path from a base station; (b) estimating a channel impulse response of each propagation path according to the signal vector to construct a channel impulse response matrix; (c) calculating a noise variance of the signal vector according to the channel impulse response matrix and the signal vector when the channel impulse response remains largely unchanged for a particular period of the training sequence. Noise variance estimation method comprising. The method of claim 1, Wherein the specific period is a time period of the training sequence. The method of claim 2, Step (c) is, estimating an MLE (maximum probability estimate) value of the training sequence included in the signal vector according to the channel impulse response matrix and the signal vector; (c2) calculating an estimated value of the noise vector included in the signal vector according to the MLE value of the training sequence and a known value of the training sequence; (c3) calculating the noise variance of the signal vector according to the estimated value of the noise vector and the channel impulse response matrix. Noise variance estimation method comprising. The method of claim 3, wherein Step (c3) is a formula

Calculate the noise variance of the signal vector according to σ ² is the noise variance of the signal vector, n 'is an estimated value of the noise vector included in the signal vector, and H is the channel impulse response matrix. Superscript H represents the complex conjugated binomial, trace {.} Represents the calculation of the matrix trace. The method according to claim 3 or 4, And summing and averaging the noise variance of the signal vector and the noise variance computed in a previous time slot, and treating the average noise variance as the noise variance of the signal vector. Receiving means for receiving a signal vector and a noise vector comprising a training sequence transmitted over at least one propagation path from a base station; Channel estimation means for estimating a channel impulse response of each propagation path according to the signal vector to form a channel impulse response matrix; Calculating means for calculating a noise variance of the signal vector according to the channel impulse response matrix and the signal vector when the channel impulse response remains largely unchanged for a particular period of the training sequence. Noise variance estimation device comprising. The method of claim 6, And the specific period is a time period of the training sequence. The method of claim 7, wherein The calculation means, Equalizing means for estimating an MLE value of a training sequence included in the signal vector according to the channel impulse response matrix and the signal vector; Noise estimation means for calculating an estimated value of the noise vector included in the signal vector according to the MLE value of the training sequence and a known value of the training sequence; Noise square calculation means for calculating a square of an estimated value of the noise vector according to the estimated value of the signal vector; Noise square correction means for calculating the noise variance of the signal vector according to the square of the estimated value of the noise vector and the channel impulse response matrix. Noise variance estimation device comprising. The method of claim 8, The noise square correction means is a formula

Calculate the noise variance of the signal vector according to σ ² is the noise variance of the signal vector, n 'is an estimated value of the noise vector included in the signal vector and n' ^H n 'is the noise A square of the estimated value of the vector, H is the channel impulse response matrix, superscript H represents the complex conjugated binomial, and trace {.} Represents the calculation of the matrix trace. Receiving means for receiving a signal vector and a noise vector comprising a training sequence transmitted over at least one propagation path from a base station; Channel estimation means for estimating a channel impulse response of each propagation path according to the signal vector to form a channel impulse response matrix; Noise variance estimation means for calculating a noise variance of the signal vector according to the channel impulse response matrix and the signal vector when the channel impulse response remains largely unchanged for a particular period of the training sequence; Data detecting means for detecting the received signal vector to obtain a desired signal according to the calculated noise variance of the signal vector; Including user equipment. The method of claim 10, Wherein said particular period is a time period of said training sequence. The method of claim 11, The noise variance estimating means, Equalization means for estimating an MLE value of a training sequence included in the signal vector according to the channel impulse response matrix and the signal vector; Noise estimation means for calculating an estimated value of the noise vector included in the signal vector according to the MLE value of the training sequence and a known value of the training sequence; Noise square calculation means for calculating a square of an estimated value of the noise vector according to the estimated value of the signal vector; Noise square correction means for calculating the noise variance of the signal vector according to the square of the estimated value of the noise vector and the channel impulse response matrix. Including user equipment. The method of claim 12, The noise square correction means is a formula

Calculate the noise variance of the signal vector according to σ ² is the noise variance of the signal vector, n 'is an estimated value of the noise vector included in the signal vector and n' ^H n 'is the noise A square of the estimated value of the vector, H is the channel impulse response matrix, superscript H represents the complex conjugated binomial, and trace {.} Represents the calculation of the matrix trace.

Images