JP2001045084A - Detection system and detection circuit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the demodulation characteristic over that of a simple pilot symbol detection system by estimating up to a fluctuation component of a transfer function. SOLUTION: A pilot symbol generating circuit 10 and a multiplier 11 calculate a transfer function (h) of a received signal for a time period when the received signal is a pilot symbol. Application of this calculation to an optional number of or all of pilot symbols within a time period of the pilot symbols obtains a plurality of transfer functions h1-hn. An approximate curve estimate circuit 13 estimates a parameter of an approximate curve on the basis of a plurality of the transfer functions obtained as above to acquire an estimate value of the transfer function, a fluctuation speed of the transfer function and a fluctuation acceleration of the transfer function. Thus, the estimation of the transfer functions can include the effect of the fluctuation speed and the fluctuation acceleration of the transfer functions.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a detection system and a detection circuit, and more particularly to digital data transmission such as wireless data communication. In particular, the pilot symbol detection system conforms to the IMT-2000 standard for next-generation mobile phones.
Is intended to be employed in the present invention,
It can be used as a demodulation technology for this next-generation mobile phone.

[0002]

2. Description of the Related Art The invention described in Japanese Patent Application Laid-Open No. Hei 10-233713 (synchronous detection method) relates to a pilot interpolation detection method, and FIG. In addition, the invention described in the publication is to improve the demodulation characteristics by feeding back the amount of phase compensation. According to the present invention, the demodulation characteristics can be improved by a simpler method than the invention described in the above publication.

[0003]

FIG. 11 is a diagram showing a signal configuration in the pilot symbol detection system. As a form of a transmission signal used in the pilot symbol detection system, as shown in FIG. And intermittently insert pilot symbols for transmission. FIG. 11 shows a case where n pilot symbols are assigned to m information symbols.

[0004] Since the pilot symbol is a known signal on the receiving side, the transfer function of the communication channel can be obtained by comparing the received pilot symbol with the pilot symbol originally sent by the transmitting side. For example, the QPSK modulation the modulation scheme, the received signal I, consider the case of performing quadrature detection of Q, pilot signal sent on the transmission side and the complex vector Z _p, a complex vector Z pilot signals received side this ′ _P , the transfer function h is
It can be calculated by equation (1). Note that in equation (1),
Z ^* _p represents the complex conjugate of Z _p.

[0005]

(Equation 1)

FIG. 12 is a diagram showing an example of a conventional demodulation method using the pilot symbol detection method.
Is a pilot symbol generation circuit for generating pilot symbols, 11 is a complex conjugate multiplier, 60 is an averaging circuit, and 12 is a multiplier. In FIG. 12, a transfer function h is calculated by a pilot symbol generation circuit 10 and a multiplier 11 in a time section in which a received signal is a pilot symbol. Normally, as shown in FIG. 11, a plurality of pilot symbols P _{1 to} P _n are prepared, and a simple average or a weighted average is calculated by the averaging circuit 60. Next, when the received signal is in the information symbol section,
The transfer function h obtained immediately before is multiplied by the received signal by the multiplier 12 and demodulated.

A pilot interpolation detection system has been proposed as having better demodulation characteristics. This method is shown in FIG.
In uses a transfer function obtained by the pilot symbols P ₁ to P _n (which is referred to as h ₁₎ and (a h ₂₎ transfer function obtained by the pilot symbol P _{n + 1} ~P _2n, 2 This is a method of capturing between the _two transfer functions h ₁ and h ₂ . That is, the transfer function h used to demodulate the i-th information D _i to information symbol D ₁ to D _m in FIG. 11, a method of calculating the equation (2).

[0008]

(Equation 2)

However, in such a conventional demodulation method, the former demodulation method has a poor demodulation characteristic, and the latter pilot interpolation method has a complicated circuit and the following transfer function h ₂ is required. Until the calculation, the information symbols must be stored, and there is a problem that a delay of one time slot occurs.

According to the first aspect of the present invention, as described above, the demodulation characteristic is improved over the simple pilot symbol detection system by estimating the transfer function variation method by estimating the transfer function variation. The purpose is to let them. It is another object of the present invention to realize a demodulation system with a simpler configuration than the pilot interpolation detection system and further reduce a time delay.

According to a second aspect of the present invention, the use of the least squares method in the method of estimating the transfer function detects statistical fluctuations, thereby suppressing fluctuations in the estimated value of the transfer function. An object is to increase resistance to noise.

As another method of realizing the present invention, a method of performing a linear approximation several times from a plurality of obtained transfer functions and taking an average of the straight lines to detect the transfer function and its variation. is there. A third object of the present invention is to make it possible to estimate a transfer function by a simple method by using an average of approximate straight lines.

The present invention can also be realized by detecting only the variation of the transfer function and calculating the average. It is another object of the present invention to estimate a transfer function by a simple method by calculating an average of a variation.

As a method for further simplifying the invention of claim 4, there is a method of taking the difference between transfer functions and taking the average. It is another object of the present invention to simplify the circuit by detecting the variation of the transfer function by the average of the difference between the transfer functions.

[0015]

According to a first aspect of the present invention, in a pilot symbol detection system, a transfer function and a variation speed of the transfer function are estimated from a plurality of pilot symbols.
It is characterized in that an information symbol following a pilot symbol is demodulated.

According to a second aspect of the present invention, in the first aspect of the present invention, the transfer function and the variation speed of the transfer function are estimated by using a least squares method, and the pilot symbol is estimated from the transfer function and the variation speed of the transfer function. Detecting an information symbol subsequent to.

According to a third aspect of the present invention, in the first aspect of the present invention, a plurality of transfer functions and a variation speed of the transfer function are estimated from a plurality of pilot symbols, and the plurality of estimated transfer functions and a variation speed of the transfer function are statistically estimated. , A transfer function in a pilot symbol section and a variation speed of the transfer function are determined, and an information symbol following the pilot symbol section is detected.

According to a fourth aspect of the present invention, in accordance with the first aspect of the present invention, the transfer function is statistically processed from a plurality of pilot symbols, and a variation of the transfer function is detected. , A transfer function is estimated from the results of these two statistical processes, and an information symbol following the pilot symbol is detected.

According to a fifth aspect of the present invention, in the first aspect, statistical processing of a transfer function is performed from a plurality of pilot symbols, further, a difference component of the transfer function is detected, and statistical processing of the difference component is performed. , A transfer function is estimated from the results of these two statistical processes, and an information symbol following the pilot symbol is detected.

According to a sixth aspect of the present invention, there is provided a detection circuit used in a receiver of a communication system in which a predetermined length of a pilot symbol and a predetermined length of an information symbol are transmitted alternately. A first multiplier for multiplying the received signal by the signal generated by the pilot symbol generator, and a second multiplier for multiplying the received signal by the inverse characteristic of the transfer characteristic estimated by the transfer characteristic estimator. , And a transfer characteristic is estimated by a transfer characteristic estimating unit from a signal obtained by multiplying a pilot symbol included in the received signal and a pilot symbol generated by the pilot symbol generating unit by a first multiplier. , By multiplying the information symbol appearing subsequent to the pilot symbol by the inverse characteristic estimated by the transfer characteristic estimating unit in the second multiplier, It is obtained by, characterized in that detection.

[0021]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (Invention of Claim 1) FIG. 1 is a block diagram of a main part for explaining an embodiment of the invention of claim 1, in which reference numeral 10 denotes a pilot symbol generation for generating a pilot symbol. Circuit, 11 a multiplier for performing multiplication by complex conjugate, 12 a normal multiplier, 13 an approximate curve estimating circuit for estimating an approximate curve, and 14 a transfer for estimating a transfer function at a certain time from the estimated approximate curve. It is a function estimating circuit.

Next, the operation of the embodiment shown in FIG. 1 will be described. In FIG. 1, a transfer function h is calculated by a pilot symbol generation circuit 10 and a multiplier 11 in a time section in which a received signal is a pilot symbol. For example,
If the received pilot signal P is represented by a complex signal P = P _I + jP _Q and the pilot symbol generated by the pilot symbol generation circuit 10 is represented by S = S _I + jS _Q , the transfer function h is represented by a complex number divided by a conjugate complex number. from being expressed by the product of, by calculating h from equation _{(1) = (P I S} I + P Q S Q) + j (P Q S I -P I S Q) equation (3), the transfer function Can be calculated. By applying this for any number or all of the pilot symbols in the time interval of the pilot symbols, it is possible to obtain a plurality of transfer functions h ₁ to h _n.

The approximation curve estimating circuit 13 estimates the parameters of the approximation curve from the plurality of obtained transfer functions, thereby obtaining the estimated value of the transfer function, the change speed of the transfer function, and the change acceleration of the transfer function. By this, the fluctuation speed,
The effect of the fluctuating acceleration can be incorporated into the transfer function estimation. Of course, the variation speed and the variation acceleration may be obtained by expressing h as an approximate curve as a function of time t, and obtaining the first derivative and the second derivative.

As the approximation function, for example, a polynomial or an exponential function is used. In particular, when approximated by a quadratic expression, the transfer function can be expressed as h (t) = at ² + bt + c (4) as a function of time h (t). The value of the transfer function obtained from each pilot symbol is substituted into this, and the parameters a, b, c
Can be obtained. This is shown in FIG. FIG.
Here, for simplicity, the case where the number of pilot symbols is four is shown. Although FIG. 8 illustrates the case where h is a real number, the same calculation can be performed even when h is a complex number. In that case, for example, an approximate function may be obtained for each of the real part and the imaginary part.

In FIG. 8, assuming that the first pilot symbol is a time base 0, the transfer functions at each pilot symbol are (0, h ₁ ), (T, h ₂ ), (2T, h ₃ ),
(3T, h ₄ ). However, here:
T is the period of the pilot symbol. Therefore, the parameters a, b, and c can be determined by substituting the three points into equation (4). For example, P ₁
Using a value of up to P _3, the ternary simultaneous equations is obtained below by solving _{h 1 = c h 2 = aT} 2 + bT + c h 3 = 4aT 2 + 2bT + c equation (5) this following a , B, c can be obtained.

[0026]

(Equation 3)

In FIG. 1, a transfer function estimating circuit 14 estimates a transfer function in an information symbol time section based on the obtained approximate function. That is, by substituting the time in the information symbol into the approximate function h _t with the parameter obtained by the approximate function estimating circuit 13, the transfer function at that time can be estimated. That is, speaking in Figure 8, the transfer function to be used to demodulate the information symbols D ₁ applying the h (4T), it may be applied and so h (5T) in D _2. In the time section of the information symbol, the estimated value of the transmission relation obtained here is multiplied by the multiplier 12.
, Detection can be performed.

As described above, according to the present invention, it is possible to estimate the transfer function taking into account the fluctuation speed component and the fluctuation acceleration component of the transfer function, thereby improving the detection characteristics. .

(Invention of claim 2) Actually, the fluctuation of the transfer function is largely due to the influence of the rotation of the phase due to the remaining carrier component, and the fluctuation speed component affects the detection characteristics. It can be considered that the influence of the fluctuation acceleration due to the Doppler effect or the like on the transmission path is small. In such a case, a transfer function that is not easily affected by noise can be estimated by applying an approximation using a straight line statistically obtained instead of a curve to the approximation function. The least square method is known as such a statistical linear approximation.

FIG. 2 is a block diagram of a main part for explaining an embodiment of the second aspect of the present invention. In FIG. 2, reference numeral 20 denotes a regression line estimating circuit using a least squares method. The same numbers are assigned to components that play a role.
In FIG. 2, n points to be approximated now are (t ₁ ,
_{h 1), (t 2,} h 2), ..., ( and t _n, and h _n), if a regression line h = at + b equation (7), the parameters a, b, the following equation (8), It is given by equation (9).

[0031]

(Equation 4)

Therefore, when there are four pilot symbols, (0, h ₁ ), (T, h ₂ ), (2T, h ₃ ), (3
The parameters a and b can be obtained by substituting the coordinates of T, h ₄ ) into the above equation. The regression line estimation circuit (least squares estimation circuit) 20 is a circuit that performs this calculation. Using the regression line determined here, 4T, 5T,
May be determined by the transfer function estimating circuit 14.

As described above, according to the present invention, by using the least squares method, it is possible to suppress the fluctuation due to the noise of the pilot symbol.
Detection characteristics can be improved.

The equation for obtaining the parameters in the least squares method is complicated, and the circuit may be complicated to be realized by hardware. Such a case can be easily realized by appropriately setting the coordinate system.

FIG. 3 is a block diagram of a main part for explaining an embodiment of the second aspect of the present invention, and shows portions of the regression line estimation circuit 20 and the transfer function estimation circuit 14 of FIG. FIG.
, 31 is an averaging circuit, 32 is a counter for estimating a transfer function, 33 is a multiplication circuit, 34 is an integration circuit that calculates the sum of input signals for a certain period, 35 is an integration circuit that divides by a constant K, and 36 is A multiplier, 37 is an adder, and 38 is an information symbol counter.

When the invention of claim 2 is realized by hardware, it can be easily realized by placing the origin of the coordinate axis at the center of the pilot symbol time section.
This is shown in FIG. FIG. 9 shows a case where there are four pilot symbols, similarly to FIG. The origin of this case will be taken in the middle of the pilot symbols P ₂ and P _3. Then, the time axis coordinates of the four pilot symbols are -3T / 2, -T / 2, T / 2, 3T, respectively.
/ 2, and the sum is 0. This is determined by the placement of the origin and does not depend on the number of pilot symbols. Therefore,

[0037]

(Equation 5)

Equations (8) and (9), which are the equations for the parameters a and b of the regression line, are as follows:

[0039]

(Equation 6)

## EQU4 ## Therefore, in FIG.
To realize 1), the transfer function counter 32 uses the transfer function counter 32 to represent the part of t _i , -3/2, -1/2, 1/2, 3
/ 2 is generated and multiplied by a multiplier 33 for each pilot symbol. In order to obtain the sum of the multiplied signals, the sum is obtained in a time section of the pilot symbol by the integrating circuit 34. Constant K, the denominator of formula (11), since a constant determined only by t _i, can be previously determined. This is the constant K,

[0041]

(Equation 7)

Is as follows. Dividing the integration result by K is a circuit 35 that divides by the constant K in FIG. Equation (12) is the average of a simple transfer function. Therefore, the averaging circuit 31
By calculating the average, the value of b can be obtained. As described above, the values of the parameters a and b are determined by the least square method.

In the time section of the information symbol, the time of each information symbol is calculated by the information symbol counter 38, and from this value and a and b obtained by the above method, the equation (7) is multiplied by a multiplier 36. The transfer function can be estimated by calculating with the adder 37. As described above, according to the present invention, it is possible to realize a detection device that estimates a transfer function by the least square method with a simple configuration.

(Invention of Claim 3) FIG. 4 is a schematic diagram of a main part for explaining an embodiment of the invention of claim 3, in which FIG.
Is a straight line approximation circuit, 21 is an averaging circuit for averaging the straight lines, and the same numbers are assigned to the same components as those in FIG. In FIG. 4, in order to perform straight line approximation, if the coordinates of two points are known, the slope and intercept of the straight line can be determined. Therefore, among the transfer functions h _i obtained by each pilot symbol, using any two values, performing linear approximation. This is shown in FIG. FIG.
0 shows an example in which, of the four transfer functions, a first straight line is obtained from h ₁ and h ₃ , and a second straight line is obtained from h ₂ and h ₄ . The other two points are, of course, permitted. The straight line approximation circuit 20 in FIG. 4 is a circuit for obtaining a straight line from two points. The average of the obtained straight line is averaged by the averaging circuit 21.
Taken by The average is obtained, for example, by calculating the average of the slopes of all the straight lines and the average of the intercepts of all the straight lines. First-order approximation is performed using the slope and intercept obtained by the above operation, and the transfer function in the information symbol time section can be estimated.

Obtaining the straight line equation directly requires a division circuit, which complicates the circuit. Therefore, the circuit can be simplified by combining the processing for obtaining the straight line and the processing for obtaining the average. FIG. 5 is a main part configuration diagram for explaining the embodiment of the third aspect of the present invention, and shows the parts of the linear approximation circuit 20, the averaging circuit 21, and the transfer function estimating circuit 14 of FIG. Each role is enclosed in a dotted box. Also, the same numbers are assigned to the components that play the same role as in FIG. 5, 51 and 52 are coefficient generating circuits, 53 and 54 are multipliers,
55 and 56 are integrators, and 57 and 58 are multipliers of certain constants.

Next, the operation will be described. Assuming that two points for obtaining a straight line are a pilot symbol time and a transfer function represented by (t _i , h _i ) and (t _k , h _k ), the slope a of the straight line is

[0047]

(Equation 8)

Is represented by Here, if the time width t _i -t _k is set to be the same for all the straight lines, the denominator of the above equation a becomes the same value for all the straight lines. For example, in FIG. 10, the straight line determined by the pilot symbol p ₁ and p _3, and a straight line determined by the pilot symbol p ₂ and p ₄ are both _{t 3 -t 1 = t 4 -t} 2 = 2T. Therefore,
Since the slopes a of all the straight lines have a constant denominator, the average of the slopes a is the average of the numerator. The sum of the molecules is, for example, as shown in FIG.
For _{0, (h 3 -h 1)} + (h 4 -h 2) = - h 1 -h 2 + h 3 + h 4 Expression and so (15), multiplied by 1 or -1 in the transfer function _{h i} The sum of the values. Further, (t _{i, h i)} when using more than one point of calculating a plurality of straight lines, will appear a value multiplied by the number using the (t _{i, h i)} to h _i. In any case, the sum of the molecule multiplied by the appropriate constant in a transfer function h _i, can be obtained by taking the sum. Therefore, in FIG. 5, the coefficient generation circuit 51, by calculating in the multiplier 55 and the sum to generate a coefficient to multiply the transfer function h _i, the slope of the straight line a, i.e., detects the variation in the transfer function be able to. Similarly, the intercepts of the straight line are (t _i , h _i ), (t _k ,
h _k ),

[0049]

(Equation 9)

Is given by If the value in this case is also t _k -t _i is constant, the sum of the intercept b is multiplied by a predetermined constant transfer function h _i, can be obtained by taking the sum. For example, in the case of FIG. 10, the intercept of a straight line determined by p ₁ and p ₃ is given by Expression 17, and p ₂ and p ₄
The intercept of the straight line defined by Therefore,
The sum of these two intercepts is given by Equation 19, where h ₁ , h ₂ ,
h ₃ and h ₄ are multiplied by −1, 0, −3/2, and 、, respectively, and the sum is obtained.

[0051]

(Equation 10)

Therefore, in FIG.
By multiplying the transfer function by a constant and taking the sum thereof, the intercept b of the straight line, that is, the average of the transfer function can be detected. Multiplying the variation of the transfer function and the average of the transfer function obtained as described above by an appropriate constant by the constant multipliers 57 and 58 to determine a linear equation, and estimating the transfer function in the information symbol from this equation. Can be.

(Invention of Claim 4) The variation of the transfer function can also be obtained by calculating the difference between the transfer functions.
FIG. 6 is a block diagram showing a main part of a fourth embodiment of the present invention. In FIG.
Numeral 61 denotes a circuit for calculating the variation of the transfer function, and 62 denotes a circuit for calculating the average. In FIG.
As the circuit 61 for detecting the fluctuation, the straight-line approximation circuit described in the third aspect of the present invention can be used as it is. Thus, the variation is detected, and the average is obtained by the averaging circuit 62. On the other hand, a circuit 60 for calculating the average of the transfer function
Is a circuit for obtaining a simple average and a weighted average of a transfer function used in a conventional pilot symbol detection circuit. Therefore, the transfer function can be estimated using the variation of the transfer function as the slope a of the linear approximation and the average of the transfer function as the intercept b of the linear approximation. The invention of claim 4 is particularly equivalent to the invention of claim 3 in which the origin of time is set at the last pilot symbol in a time slot.

(Invention of Claim 5) The variation of the transfer function can be obtained, most simply, by taking a difference signal from the transfer function one symbol before. FIG. 7 is a block diagram of a main part for explaining an embodiment of the invention of claim 5;
0 is a circuit that takes the difference from the transfer function of one symbol before,
In addition, the same numbers are assigned to components that perform the same functions as in FIG. In FIG. 7, the difference signal of the transfer function corresponds to the calculation of the slope a when the pilot symbols that are sometimes adjacent to each other are used in Expression (14), which is the expression for the variation of the transfer function. Therefore, the difference in the transfer function can be detected by the difference circuit 70, and the transfer function can be estimated using this value.

[0055]

According to the first aspect of the present invention, a detection method that is simpler than the pilot interpolation detection method and has better demodulation characteristics than the simple pilot symbol detection method is realized by estimating the variation of the transfer function. And a time delay at the time of detection can be reduced.

According to the second aspect of the present invention, by using the least squares method for estimating the transfer function, the statistical fluctuation is detected, and thereby the fluctuation of the estimated value of the transfer function is suppressed. Can be made more resistant to noise. Further, by appropriately setting the coordinate system to which the least square method is applied, it is possible to provide a circuit for estimating a transfer function with a simple configuration.

According to the third aspect of the present invention, it is possible to provide a detection method for estimating a transfer function by a simple method by using the average of the approximate straight lines. In addition, a simple and good detection device can be provided by using a simple multiplier and an integrator.

According to the fourth aspect of the present invention, the transfer function can be estimated by a simple method by obtaining the average of the variation.

According to the fifth aspect of the present invention, the circuit can be simplified by detecting the variation of the transfer function based on the average of the difference between the transfer functions.

[Brief description of the drawings]

FIG. 1 is a main part configuration diagram for explaining an embodiment of the invention of claim 1;

FIG. 2 is a main part configuration diagram for explaining an embodiment of the invention of claim 2;

FIG. 3 is a main part configuration diagram for explaining another embodiment of the invention of claim 2;

FIG. 4 is a main part configuration diagram for explaining an embodiment of the invention of claim 3;

FIG. 5 is a main part configuration diagram for explaining another embodiment of the invention of claim 3;

FIG. 6 is a main part configuration diagram for explaining an embodiment of the invention of claim 4;

FIG. 7 is a main part configuration diagram for explaining an embodiment of the invention of claim 5;

FIG. 8 is a diagram for explaining an approximate curve according to the invention of claim 1;

FIG. 9 is a diagram for explaining an approximate straight line by the least square method.

FIG. 10 is a diagram illustrating an approximate straight line obtained by averaging a plurality of straight lines.

FIG. 11 is a diagram showing a signal configuration in a pilot symbol detection system.

FIG. 12 is a diagram illustrating an example of a conventional method of demodulation using a pilot symbol detection method.

[Explanation of symbols]

Reference Signs List 10: pilot symbol generation circuit, 11: complex conjugate multiplier, 12: normal multiplier, 13: approximate curve estimation circuit, 14: transfer function estimation circuit, 20: linear approximation circuit (regression line estimation circuit), 21 ... Averaging circuit, 30 ... Regression line estimation circuit, 31 ... Averaging circuit, 32 ... Transfer function estimation counter, 33 ... Multiplication circuit, 34 ... Integration circuit, 35 ... Integration circuit, 36 ... Multiplier, 37 ... Adder, 38 ... Information symbol counter, 51,52 ... Coefficient generation circuit, 53,54
.., Multipliers, 55, 56, integrators, 57, 58, constant multipliers, 60, averaging circuit, 61, fluctuation detecting circuit, 62, averaging circuit, 70, difference circuit.

Claims (6)

[Claims] In a pilot symbol detection system,
A detection method characterized by estimating a transfer function and a transfer function fluctuation speed from a plurality of pilot symbols, and demodulating an information symbol following the pilot symbol. 2. The method according to claim 1, wherein the transfer function and the transfer function change speed are estimated using a least squares method, and an information symbol subsequent to the pilot symbol is detected from the transfer function and the transfer function change speed. Detection method characterized by the following. 3. The method according to claim 1, further comprising: estimating a plurality of transfer functions and a change speed of the transfer function from the plurality of pilot symbols; and statistically processing the plurality of estimated transfer functions and the change speed of the transfer function. A transfer function and a fluctuation speed of the transfer function, and detecting an information symbol following the pilot symbol section. 4. The method according to claim 1, further comprising: performing a statistical process on the transfer function from the plurality of pilot symbols; detecting a variation in the transfer function; performing a statistical process on the variation; A detection method characterized by estimating a transfer function from a result of a spatial processing and detecting an information symbol following a pilot symbol. 5. The method according to claim 1, wherein a transfer function is statistically processed from a plurality of pilot symbols, a difference component of the transfer function is detected, and the difference component is statistically processed. A detection method characterized by estimating a transfer function from a result of a spatial processing and detecting an information symbol following a pilot symbol. 6. A detection circuit used in a receiver of a communication system in which a pilot symbol of a predetermined length and an information symbol of a predetermined length are transmitted alternately, a pilot symbol generator, a transfer characteristic estimator, and a received signal. And a first multiplier for multiplying the signal generated by the pilot symbol generator, and a second multiplier for multiplying the received signal by the inverse characteristic of the transfer characteristic estimated by the transfer characteristic estimator. A transfer characteristic is estimated by a transfer characteristic estimating unit from a signal obtained by multiplying a pilot symbol included in a received signal by a pilot symbol generated by a pilot symbol generating unit in a first multiplier, and the pilot symbol The detection of the information signal is performed by multiplying the succeeding information symbol by the inverse characteristic estimated by the transfer characteristic estimating unit by the second multiplier. The detection circuit that was marked.