KR100768512B1 - Method detecting multidimensional symbol and soft decision thereof - Google Patents

Abstract

The present invention relates to a multi-dimensional symbol detection method and a soft decision method using the same.

According to the present invention, the multi-dimensional symbol detection method updates the first multi-dimensional symbol in a direction opposite to the change rate obtained by partial division of the distance between the transmit symbol from the transmittable symbol in the multi-dimensional complex space into the real part and the imaginary part of the complex symbol. The second likelihood of the maximum likelihood for the received symbol is detected by gradually increasing the likelihood of the detection result.

In addition, the soft decision method applies the multi-dimensional symbol detection method to represent the rate of change of the distance function between the transmission symbol and the reception symbol on the hyperplane for each bit by expressing each symbol constituting the first multidimensional symbol that can be transmitted. The bit rate soft decision is performed by controlling the update of a specific bit in the process of updating the first multi-dimensional symbol.

Multidimensional Symbol, Symbol Detection, Soft Decision, Multiple I / O, MIMO

Description

Multi-Dimensional Symbol Detection Method and Soft Decision Method Applying Them {METHOD DETECTING MULTIDIMENSIONAL SYMBOL AND SOFT DECISION THEREOF}

1 is a configuration diagram of a multiple input / output system according to an embodiment of the present invention.

2 illustrates a schematic multi-dimensional symbol detection method according to an embodiment of the present invention.

3 is a diagram illustrating a detailed multi-dimensional symbol detection method according to an embodiment of the present invention.

4 is a view showing a soft decision method applying the multi-dimensional symbol detection method according to an embodiment of the present invention.

The present invention relates to a multi-dimensional symbol detection method and a soft decision method using the same.

Multi-Input Multi-Out (MIMO) system is a system for installing a plurality of antennas at the transmitting and receiving end, and transmits different transmission signals from each transmitting antenna. That is, by transmitting a multi-dimensional symbol using a plurality of antennas, it is possible to increase the speed of data that can be transmitted in a limited frequency band.

However, the conventional multi-input / output system has a problem in that the complexity of the receiver which is optimal in terms of maximum likelihood due to crosstalk between a plurality of antennas increases geometrically and exponentially in proportion to the number of antennas. .

In order to alleviate this problem, a simple method of ignoring the influence of noise and removing only crosstalk has been proposed. However, this method causes severe performance degradation compared to the maximum likelihood receiver due to noise enhancement in the calculation process.

In order to alleviate the noise enhancement, a technique for removing the effects of intersymbol interference has been proposed, but this also has a disadvantage in that the operation is complicated.

In addition, a method of slightly reducing the amount of computation is proposed by excluding a symbol vector far enough from a reception symbol vector in a multi-dimensional complex space as a comparison target. However, this is also complicated to identify symbol vectors within a given radius due to crosstalk between antennas.

In addition, there is still a problem that the number of symbol vectors to be compared increases geometrically and exponentially according to the number of antennas.

Accordingly, an object of the present invention is to provide a multi-dimensional symbol detection method for gradually increasing the likelihood for a symbol received in a multi-dimensional complex space.

Another object of the present invention is to provide a soft decision method for each bit constituting a symbol using a multidimensional symbol detection method.

According to a feature of the present invention in order to achieve the above problems,

Multi-dimensional symbol detection method,
A method of detecting a multi-dimensional symbol by a receiving end of a multi-in multi-out (MIMO) system, comprising: calculating a rate of change of a distance function for each symbol of a first transmittable multidimensional symbol negotiated between a transmitting end and the receiving end; Change rate calculation step; An update step of selecting a predetermined symbol having a negative rate of change from each of the symbols to update the first multi-dimensional symbol so that a distance from a received symbol is closer than the current one; And detecting the second multi-dimensional symbol in which the distance from the received symbol falls within a specific threshold by repeating the change rate calculating step and the updating step by applying the updated first multi-dimensional symbol.

delete

According to another feature of the invention,

The soft decision method is

A soft decision method of a multiple input / output receiver, comprising: a first soft decision step of detecting a distance by detecting a second multidimensional symbol having a maximum likelihood for a received symbol among first transmittable multidimensional symbols negotiated between a transmitter and the receiver; A first updating step of toggling a value of a first bit of the second multidimensional symbol; Selecting a bit having a negative rate of change of a distance function among bits of the second multidimensional symbol except the first bit to toggle the bit so that the distance from the received symbol is closer than the current to update the second multidimensional symbol A second update step; And a second soft decision step of determining a reliability of the first bit by obtaining a distance of the updated second multidimensional symbol and a difference from the distance obtained in the first soft decision step.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. As those skilled in the art would realize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

In addition, when a part is said to "include" a certain component, this means that it may further include other components, except to exclude other components unless otherwise stated.

Now, a multidimensional symbol detection method and a soft decision method employing the same according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a configuration diagram of a multiple input / output system according to an embodiment of the present invention.

개의 송신 안테나(120)를 사용하는 송신단(100)과

개의 수신 안테나(210)를 사용하는 수신단(200)으로 구성되어 있다.That is, as shown in FIG. 1, the multiple input / output system

Transmitting stage 100 using two transmitting antennas 120 and

It consists of a receiving end 200 using two receiving antennas (210).

Each symbol (symbol 1, symbol 2... Symbol N) constituting the multi-dimensional symbol of the transmitter 100 is transmitted through a modulator (modulator 1, modulator 2 ... modulator N, 110) as a transmission symbol vector as a transmission antenna 120. It is transmitted to the receiving end 200 through).

When the transmitting symbol vector is transmitted to the receiving terminal via the transmitting antenna 120 and received by the receiving antenna 210 as the receiving symbol vector, the receiving symbol vector is a demodulator (demodulator 1, demodulator 2 ...). Through 220, noise is mixed into the transmission symbol vector and extracted as a distorted multidimensional reception symbol (receive symbol 1, received symbol 2 ... received symbol N). The extracted multi-dimensional received symbol is restored to the multi-dimensional transmitted symbol transmitted from the transmitter 100 via the multi-dimensional symbol detector 240.

At this time, the channel estimator 230 estimates channel characteristics between the transmitter 100 and the receiver 200 using pilot symbols and obtains a channel model and transmits the channel model to the multi-dimensional symbol detector 240.

The multi-dimensional symbol detector 240 is a means for detecting a transmission symbol vector having a maximum likelihood for the received symbol vector. The multi-dimensional symbol detector 240 minimizes the distance function f (X) formed by the relation between the transmission symbol vector, the received symbol vector, and the channel model. Detect vectors That is, the distance between the transmittable symbol in the multidimensional complex space and the received symbol is represented by a multivariate quadratic polynomial, and the multidimensional symbol closest to the received symbol is detected by an optimization technique using partial differential coefficients for each component of the multidimensional symbol. .

In addition, each symbol is represented as a first-order function of bits, and the distance between the transmit and receive symbols on the hyperplane for each bit is calculated using the same method as that of the multi-dimensional symbol. Decisions can be made.

Such multi-dimensional symbol detection and soft decision using the same will be described in more detail with reference to the accompanying drawings.

First, FIG. 2 is a diagram illustrating a schematic multidimensional symbol detection process according to an embodiment of the present invention.

That is, as shown in FIG. 2, the multi-dimensional symbol detection process includes an input step (S110), a first multi-dimensional symbol setting step (S120), a change rate calculation step (S130), an update step (S140), and an iterative update determination step ( S150) and the second multi-dimensional symbol detection step S160, which will be described for each step as follows.

The input step S110 is a step of receiving the received symbol vector Y and the estimated channel model H from the demodulator 220 and the channel estimator 230, respectively.

The first multidimensional symbol determination step (S120) is a step of initializing an arbitrary first multidimensional symbol that can be transmitted. Here, the first multidimensional symbol that can be transmitted is a multidimensional symbol that is agreed between the transmitting end 100 and the receiving end 200 and belongs to a symbol set known to the receiving end. In addition, the first multi-dimensional symbol that can be initialized in step S120 may include a multi-dimensional symbol output from another detector. That is, it may be a transmission multi-dimensional symbol estimated from the received signal Y and the channel model H using a linear detection method. For example, a transmission multi-dimensional symbol estimated by the product of the inverse of H and Y may be used. Thus, if a multi-dimensional symbol having a relatively small distance has already been detected, it can be utilized.

The change rate calculation step S130 calculates a change rate of a distance function for each symbol constituting the first multidimensional symbol set in the step S120.

In the update step S140, a predetermined symbol having a negative change rate is selected from the first multi-dimensional symbols to update in a direction opposite to the change rate calculated in the change rate calculation step S130. As a result, the distance between the first multidimensional symbol, that is, the multidimensional transmission symbol and the received symbol is gradually reduced. Here, a specific update process will be described with reference to FIG. 3.

The iterative update determination step (S150) is a step of determining whether the above steps, that is, the first multidimensional symbol setting step (S120), the change rate calculation step (S130), and the updating step (S140) are repeatedly performed. That is, as a result of determining whether to repeat the update, if the repeated update is to be performed, the steps S130 and S140 are repeated by applying the updated symbol to the first multi-dimensional symbol. In addition, if it is determined that the repeated update is not to be performed as a result of the repeated update, the next step (S160) is performed.

The second multi-dimensional symbol detection step S160 outputs a second multi-dimensional symbol composed of the updated symbols when the repeated update is not to be performed in the iterative update determination step S150. That is, the second multidimensional symbol means the first multidimensional symbol having the maximum likelihood detected through the iterative update.

3 is a diagram illustrating a detailed multi-dimensional symbol detection process according to an exemplary embodiment of the present invention, and a description of the steps overlapping with FIG. 2 will be omitted.

That is, as shown in FIG. 3, the channel model H and the reception symbol Y are first input (S210).

Next, it initializes to a first multidimensional symbol that can be transmitted (S220). In this case, the first multidimensional symbol vector is represented by Equation 1 below.

는 제1 다차원 심볼 벡터이고,

는 제1 다차원 심볼 벡터를 구성하는 각각 인덱스 (k=1, 2, n_T)에 해당하는 심볼 벡터이다.here,

Is the first multidimensional symbol vector,

Is a symbol vector corresponding to an index (k = 1, 2, n _T ) constituting the first multidimensional symbol vector.

를 구성하는 각 심볼 벡터(

)에 대한 변화율을 계산한다(S230).Next, the first multidimensional symbol vector

Each symbol vector (

Calculate the rate of change for (S230).

In this case, the rate of change is obtained by partial-dividing the distance function into the real part and the imaginary part of the complex symbol. Here, the distance function f (X) represents a distance between the multi-dimensional symbol received in the multi-dimensional complex space and the transmittable symbol and is expressed by Equation 2 below.

차원 수신 심볼 벡터이고 H는 안테나 간 크로스토크를 표현하는 채널 모델로서

행렬이다. 이러한 수신 심볼 벡터(Y)에 대한 우도를 높이는 문제는 거리 함수f(X)를 최소화하는 문제로 귀착한다. Where Y is

Is a dimensional received symbol vector and H is a channel model representing crosstalk between antennas

It is a matrix. The problem of increasing the likelihood for the received symbol vector Y results in a problem of minimizing the distance function f (X).

Next, the rate of change of the distance function is obtained by partial derivative of the distance function into the real part and the imaginary part of the complex symbol. Equations 2 and 3 are used.

는 k번째 심볼 벡터(

)의 실수부의 변화율이고

는 k번째 심볼 벡터(

)의 허수부의 변화율이고, k는 1, 2 ㆍㆍㆍ

를 포함하고

는 송신 심볼 벡터의 차원이다. 또한,

및

는 각각 거리 함수를 다차원 심볼의 근방에서 평면으로 근사할 때의 기울기 중에서 실/허수부에 비례하는 정도 및 허수부/실수부에 비례하는 정도이며,

과

는 다차원 심볼의 실수부/허수부 값에 영향을 받지 않는 상수이다. 이때,

,

,

및

는 상기 제1 다차원 심볼 설정단계(S120)에서 미리 계산해두어, 반복계산을 피할 수 있다.here,

Is the kth symbol vector (

Is the rate of change of the real part of

Is the kth symbol vector (

) Is the rate of change of the imaginary part of k, k is 1, 2

Including

Is the dimension of the transmit symbol vector. Also,

And

Are proportional to the real / imaginary part and proportional to the imaginary part / real part of the inclinations when the distance function is approximated to the plane in the vicinity of the multidimensional symbol.

and

Is a constant that is not affected by the real / imaginary part value of a multidimensional symbol. At this time,

,

,

And

Calculate in advance in the first multi-dimensional symbol setting step (S120), it is possible to avoid iterative calculation.

Next, a predetermined symbol having a negative change rate is selected and updated in the opposite direction to the change rate within a transmittable range (S240).

At this time, first, a symbol in which the absolute values of the real part and the imaginary part of the change rate calculated in the step S230 has the maximum value is selected (S241). Here, the symbol is selected using Equation 5 below.

In other words, r is an index at which the absolute value of the rate of change of the real part becomes maximum. I is an index in which the absolute value of the rate of change of the imaginary part becomes maximum.

Next, the absolute value of the real part of the symbol selected in the step S241 and the magnitude of the absolute value of the imaginary part are compared (S242). At this time, the comparison expression is as shown in Equation 6 below.

은 상기 단계(S241)에서 선택한 인덱스 r에 해당하는 심볼 벡터의 실수부의 변화율이고,

는 상기 단계(S241)에서 선택한 인덱스 i 에 해당하는 심볼 벡터의 허수부의 변화율이다.here,

Is the rate of change of the real part of the symbol vector corresponding to the index r selected in step S241,

Is the rate of change of the imaginary part of the symbol vector corresponding to the index i selected in step S241.

)이 상기 허수부의 절대값(

)보다 큰 경우 상기 실수부의 변화율(

)에 반대 방향으로 상기 심볼의 실수부(

)를 갱신한다.As a result of the comparison in step S242, the absolute value of the real part (

Is the absolute value of the imaginary part (

If greater than the rate of change of the real part (

The real part of the symbol in the opposite direction to

).

)이 상기 실수부의 절대값(

)보다 큰 경우 상기 허수부의 변화율(

)에 반대 방향으로 상기 심볼의 허수부(

)를 갱신한다.Or, as a result of the comparison in step S242, the absolute value of the imaginary part (

Is the absolute value of the real part (

If greater than the rate of change of the imaginary part (

The imaginary part of the symbol in the opposite direction to

).

)또는 허수부(

)의 갱신은 각각 다음과 같은 과정을 통해 이루어진다.At this time, the real part of the symbol (

) Or imaginary part (

) Is updated through the following process.

)를 갱신하는 경우 상기 실수부의 변화율(

)의 부호를 확인한다(S243). 즉

이 양(+)인지 음(-)인지 여부를 확인한다.First, the real part of the symbol (

), The rate of change of the real part (

Check sign (S243). In other words

Check whether this is positive or negative.

)의 부호가 양(+)인 경우 상기 실수부(

)를 감소한다(S244).As a result of checking in the step S243, the rate of change of the real part (

Is a positive sign, the real part (

Decrease (S244).

)의 부호가 음(-)인 경우 상기 실수부(

)를 증가한다(S245).As a result of the confirmation, the rate of change of the real part (

) Sign is negative, the real part (

Increase (S245).

)를 갱신하는 경우 상기 허수부의 변화율(

)의 부호를 확인한다(S246). 즉

가 양(+)인지 음(-)인지 여부를 확인한다.Meanwhile, the imaginary part of the symbol (

), The rate of change of the imaginary part (

Check sign (S246). In other words

Checks whether is positive or negative.

)의 부호가 양(+)인 경우 상기 허수부(

)를 감소한다(S247).As a result of checking in the step S246, the rate of change of the imaginary part (

The sign of the imaginary part (

) (S247).

)의 부호가 음(-)인 경우 상기 허수부(

)를 증가한다(S248).As a result of checking in the step S246, the rate of change of the imaginary part (

If the sign of the negative sign (-) the imaginary part (

Increase (S248).

)가 특정 임계치(

)를 초과하는지를 판단함으로써 반복 갱신 여부를 판단한다(S249).Next, the distance function of the multi-dimensional symbol applying the updated symbol through the step (S241) (

) Has a certain threshold (

It is determined whether or not to repeat the update by determining whether or not (S249).

)가 특정 임계치(

)를 초과하는 경우 상기 단계(S230, S240)를 반복한다.As a result of the determination in step S249, the distance function of the multi-dimensional symbol (

) Has a certain threshold (

If is exceeded) repeat the steps (S230, S240).

)가 특정 임계치(

) 미만인 경우 해당하는 다차원 심볼 벡터(

)출력한다(S250). 즉 이때의 다차원 심볼 벡터(

)가 최대 우도의 제2 다차원 심볼 벡터이다.However, as a result of the determination in step S249, the distance function of the multi-dimensional symbol (

) Has a certain threshold (

Less than), the corresponding multidimensional symbol vector (

The output (S250). That is, the multidimensional symbol vector (

) Is the second multidimensional symbol vector of maximum likelihood.

4 is a diagram illustrating a bit-by-bit soft decision process applying a multi-dimensional symbol process according to an embodiment of the present invention.

That is, as shown in FIG. 4, the soft decision process first detects the second multi-dimensional symbol closest to the received symbol by using the method shown in FIG. 3 and the like and stores distances from the received symbol (S310 and S320). That is, the first soft decision is performed through the steps S310 and S320.

Then, the first bit of the second multidimensional symbol is toggled by selecting a specific first bit (S330). At this time, the toggle changes to '1' if the value of the first bit is '0'.

Then, the rate of change of the distance function for each bit is calculated for each symbol constituting the second multidimensional symbol (S340). That is, the rate of change of the distance function between the symbol on the hyperplane and the received symbol is calculated by expressing each symbol constituting the first multidimensional symbol as a linear function of bits. In this case, the rate of change of the distance function for each bit is expressed by linear combination of binary variables of the real part and the imaginary part of the complex symbol.

The second multi-dimensional symbol is updated by selecting and toggling a bit having a negative rate of change among the bits of the second multi-dimensional symbol except for the first bit selected in the step S330 (S350). At this time, more specifically, the update first selects the bit with the largest absolute value of the real part and the imaginary part of the change rate. The real part and the imaginary part of the selected bit are increased or decreased in the opposite direction to the sign of the rate of change. In other words, if the sign is positive, it decreases. If it is negative, it decreases.

Here, the steps S340 and S350 are performed by applying the multi-dimensional symbol method described with reference to FIGS. 2 to 3 in units of bits instead of symbols, and a detailed description thereof will be omitted.

Then, if it is determined whether to repeat the update (S360) to repeat the update, the steps (S340, S350) are repeated.

In addition, if the update is not to be repeated, a second sequence for determining the distance between the updated second multi-dimensional symbol and the received symbol and determining the reliability of the first bit based on the ratio of the distance stored in the step S320. The determination is performed (S370).

Accordingly, by selecting a bit to be updated (S350), the specific bit, that is, the first bit, is excluded from the update candidate, thereby controlling the update of the specific bit and making a soft decision on each bit. In addition, since the soft decision for each bit is performed independently, parallel processing is possible.

The embodiments of the present invention described above are not implemented only through the apparatus and the method, but may be implemented through a program for realizing a function corresponding to the configuration of the embodiment of the present invention or a recording medium on which the program is recorded. Implementation may be easily implemented by those skilled in the art from the description of the above-described embodiments.

The scope of the present invention is not limited to the above-described embodiments, but various modifications and improvements of those skilled in the art using the basic concept of the present invention as defined in the following claims are also within the scope of the present invention.

By the above-described configuration, since the symbol of the minimum distance is gradually obtained, the complexity and performance of the operation can be controlled by the calculation time. It also provides an effect that can be used as a post-processing tool for other detectors.

In addition, the soft decision on each bit constituting the symbol is independently performed to provide the effect of parallel processing.

Claims (11)

In the method of detecting a multi-dimensional symbol at the receiving end of a multi-in multi-out (MIMO) system, A change rate calculation step of calculating a rate of change of a distance function for each symbol of the first transmittable multidimensional symbol negotiated between a transmitter and the receiver; An update step of selecting a predetermined symbol having a negative rate of change from each of the symbols to update the first multi-dimensional symbol so that a distance from a received symbol is closer than the current one; And A detection step of detecting a second multi-dimensional symbol in which a distance from the received symbol is within a specific threshold by repeating the change rate calculating step and the updating step by applying an updated first multi-dimensional symbol Multi-dimensional symbol detection method comprising a. The method of claim 1, The update step, A symbol selecting step of selecting a symbol whose absolute value of the real part and the imaginary part of the change rate has a maximum value; A comparison step of comparing an absolute value of the real part of the selected symbol with an absolute value of the imaginary part; A real part updating step of updating the real part in a direction opposite to a change rate of the real part when the absolute value of the real part has a large value as a result of the comparison; And The imaginary part updating step of updating the imaginary part in a direction opposite to the change rate of the imaginary part when the absolute value of the imaginary part has a large value as a result of the comparison. Multi-dimensional symbol detection method comprising a. The method of claim 2, The real part update step, Checking the sign of the rate of change of the real part; Reducing the real part when the sign of the rate of change of the real part is positive (+); And Increasing the real part when the sign of the rate of change of the real part is negative Multi-dimensional symbol detection method comprising a. The method of claim 2, The imaginary part updating step, Checking the sign of the rate of change of the imaginary part; Reducing the imaginary part when the sign of the rate of change of the imaginary part is positive (+); And Increasing the imaginary part when the sign of the rate of change of the imaginary part is negative (-) Multi-dimensional symbol detection method comprising a. The method according to any one of claims 1 to 4, Before the change rate calculation step, And a first multidimensional symbol determining step of determining the first multidimensional symbol that can be transmitted. The method of claim 5, The first multi-dimensional symbol determination step And determining the first multidimensional symbol from a set of symbols that can be transmitted by the transmitting end stored in the receiving end in advance between the receiving end and the transmitting end. The method of claim 5, The first multi-dimensional symbol determination step And determining another transmittable multi-dimensional symbol determined as the multi-dimensional symbol having the maximum likelihood as the first multi-dimensional symbol. The method of claim 1, The detecting step, Determining whether the distance function of the updated first multidimensional symbol exceeds the specific threshold; Repeating the change rate calculating step and the updating step if the distance function of the updated first multidimensional symbol exceeds the specific threshold; And Detecting the updated first multidimensional symbol as the second multidimensional symbol when the distance function of the updated first multidimensional symbol is less than the specific threshold. Multi-dimensional symbol detection method comprising a. In the method of soft decision by the multiple input and output receiver, A first soft decision step of detecting a distance by detecting a second multidimensional symbol having a maximum likelihood for the received symbol among the first transmittable multidimensional symbols negotiated between the transmitter and the receiver; A first updating step of toggling a value of a first bit of the second multidimensional symbol; Selecting a bit having a negative rate of change of a distance function among bits of the second multi-dimensional symbol except the first bit, toggling the bit to update the second multi-dimensional symbol so that the distance from the received symbol is closer than the current one; 2 update step; And A second soft decision step of determining a reliability of the first bit by obtaining a distance of the updated second multidimensional symbol and a difference from the distance obtained in the first soft decision step Soft decision method comprising a. The method of claim 9, The second update step, A change rate calculating step of calculating a change rate of the bit-by-bit distance function for the second multidimensional symbol; A bit selecting step of selecting a bit in which the absolute value of the real part and the imaginary part of the change rate is maximum; And An update step of increasing or decreasing the real part and the imaginary part of the bit in a direction opposite to the sign of the rate of change Soft decision method comprising a. The method of claim 9 or 10, Between the second update step and the second soft decision step, The repetition determination step of the second update step for determining whether the second multi-dimensional symbol is repeatedly updated and repeating the second update step or outputting the updated second multi-dimensional symbol according to the determination result. Soft decision method further comprising.

Images