JP2003234682A - Radio receiver - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To realize timing detection and frequency synchronization with a small amount of arithmetic operations in a radio receiver using an adaptive array antenna. <P>SOLUTION: A timing detection part 2 calculates correlation u(p) by using a correlation vector v<SB>1</SB>to be obtained as a correlation arithmetic operation between a known signal included in a desired signal and received signals from a plurality of antennas 1 and a correlation matrix Φ to be obtained from the received signals and detects timing of the desired signal based on the correlation. The correlation matrix Φ may be calculated by using a received signal at the time when no desired signal exists. In addition, frequency offset is detected by using a plurality of known signal lines included in the desired signal, the correlation vector of the received signal and the correlation matrix. Thus, initial synchronization capture with high performance is realized. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a receiver using a plurality of antennas in a wireless communication system.

[0002]

2. Description of the Related Art In wireless communication, many interference signals arrive at a receiving station at the same time as a desired signal. In this situation,
An adaptive array antenna is well known as a method for extracting a desired signal with high quality. The adaptive array antenna is a method of efficiently combining desired signals while reducing the interference signal level by receiving signals from a plurality of antennas and appropriately adjusting an inter-antenna combining ratio (weight).

An adaptive array antenna will be described with reference to FIGS. 6 to 9. As shown in FIG. 6, at the receiving station, M
The signals x ₁ (p), x ₂ (p), ..., x at the individual antennas 30
Receive _M (p). Here, x ₁ (p), x ₂ (p), ..., X _M (p) are complex signals representing the p-th sample value at the antennas 1, 2, ..., M, respectively. Also, the received signal vector x (p) corresponding to the p-th sample value is x (p) = [x
₁ (p), ..., x _M (p)] ^T. Here, T represents transposition. The antenna 30 is provided with K different signals k indicated by 31 to 33.
Consider the environment where (= 1, ..., K) arrives and signals are received using M antennas. Here, k = 1 represents a desired signal and k = 2, ..., K represents an interference signal. The value of the modulation component s _k (t) of the signal k changes in the unit of sample time T _s , and E [|
s _k (t) | ² ] = 1 (E [·] is the ensemble average) is satisfied.
That is, it is assumed that the sampling time interval and the symbol time interval are equal. The following deals with the environment in which a desired signal newly arrives in the presence of an interference signal. That is, s ₁ (t) is defined by t> t ₀ , and s _k (t) (k = 2, ..., K) is all t
Is defined for. FIG. 7 shows the relationship between the time when the desired signal and the interference signal exist. As shown in the figure, the interference signals (signal 2 to signal K) exist before the arrival of the desired signal (signal 1).

At the receiving station, all signals are sent to the multiple antennas 3
After simultaneous reception using 0, sampling is performed at a time interval T _s in the baseband. The pth received sample at antenna m in the baseband signal is x _m (p),
If the complex propagation coefficient of the signal k at the antenna m is a _km and the propagation vector of the signal k is a _k = [a _k1 , ..., a _kM ] ^T , the received signal vector x (p) = [x ₁ ( p), ..., x _M (p)] ^T is expressed by the following equation.

[Equation 2] Where n _m (p) is the noise component at antenna m, and n (p)
Represents a noise vector. Transpose conjugate is H, noise power is P
_{If n} (> 0) and the unit matrix is I, then E [n (p) n (p) ^H ] =
Satisfies P _n I. In the formula (1), s _k (p) = 0 for convenience.
(P <p ₀ ), _sk (p) ≠ 0 (p ≧ p ₀ ). The propagation vector a _k is a vector that changes due to fading and is originally time-varying. However, here, the transmission / reception station moves slowly, and the fading change is sufficiently slower than the weight control time of the adaptive array, and a _k is treated as a fixed value.

In FIG. 7, the signal 1 (desired signal) is p ≧ p.
_It exists at ₀ , and frames of N _frame symbols are continuously transmitted. There are known N symbols at the beginning of the frame, and data symbols are transmitted in the other parts. In the figure, p ₀ ≤p≤p ₀ + N-1 and p ₀ + N _frame ≤p≤p ₀ + N _frame
The + N-1 portion (hatched portion in the figure) is a known symbol, and symbol series r ₁ = [r ₁ (0), ..., r ₁ (N-1)], r ₂ =
[r ₂ (0), ..., r ₂ (N-1)]. Normally, in array signal processing, signal synthesis between antennas is performed based on the weight w = [w ₁ , ..., w _M ] ^T, and the synthesized output z (p) of signal k
= W ^H x (p) is generated. At this time, by using a suitable weight w, it is possible to receive the desired signal with strong power while reducing the interference signal, and it is possible to generate a high-quality desired signal even in an environment where many interference signals exist. Becomes

FIG. 8 is a diagram for explaining the outline of the receiving station radio receiving apparatus. Referring to this figure, an MMSE (Minimum Mean)
Square Error) The weight calculation method of the adaptive array based on the combining method will be described. Here, 1 is a plurality of antennas (here, four antennas are shown),
Reference numeral 25 denotes a weight calculator that calculates the weight w using the signals received by the respective antennas 1. Reference numeral 3 denotes a plurality of weight signals that multiply the received signal from the respective antennas 1 by the weight w supplied from the weight calculator 25. Weight multiplication section,
Reference numeral 4 denotes a signal synthesizing unit that adds the outputs from the weight multiplying units 3 and synthesizes them.

In the MMSE synthesizing method, the weight w is determined by the following equation.

[Equation 3] Here, Φ is a correlation matrix, v ₁ is a correlation vector for signal 1, and H is a transposed conjugate. By using this weight, it becomes possible to receive a strong desired signal while removing an interference signal.

In the MMSE combining method described above, the weight calculation is performed in the receiving station while using the known signal r ₁ , but at that time, it is premised that the arrival time timing p ₀ of the known signal r ₁ included in the received signal can be recognized. Weight calculation is performed. This is because it is assumed that the timing synchronization of the desired signal has been sufficiently established, but actually establishing the timing synchronization itself is an important issue. Especially, in the initial synchronization stage, it is required to establish timing synchronization of a desired signal in the presence of many interference signals,
The correlation detection method generally used in MA communication cannot obtain sufficient synchronization performance due to the influence of interference signals. If an error occurs in the timing synchronization, the weight calculation of the equation (2) is also hindered, and therefore, there is a demand for a technique for obtaining highly accurate synchronization in the usage environment of the adaptive array.

To address such problems, several timing synchronization methods suitable for using the adaptive array have been proposed so far. Among them, reference [1] AV Keer
thiand JJ Shynk, "Separation of cochannel signa
ls in TDMA mobile radio, "IEEE Trans. Signal Proces
sing, vol. 46, no. 10, pp. 2684--2697, Oct. 1998.
Has proposed a method of performing timing synchronization while performing array signal processing, and it is possible to obtain timing synchronization of a desired signal while removing interference signals.

The timing synchronization method of the document [1] will be described with reference to FIG. Here, it is assumed that the receiving side has previously recognized the known symbol sequence r ₁ and uses the information to perform timing synchronization. Reference [1]
In the above method, weight calculation and beam forming are performed using known symbols for all possible timings. That is, the timing of the known signal in the received signal is unknown, but the weight calculation unit 26 performs weight calculation for all the timings. In addition, the signal synthesis unit 4
The correlator 5 performs a correlation calculation with the known symbol r _{1 on} the output of the array signal processing obtained in ( ₁₎ . Weight w (p) = when starting weight calculation from received sample p
[w ₁ , ..., W _M ] ^T is determined by the following equation.

[Equation 4] Here, Φ (p) is a correlation matrix starting from sample p, v ₁
(p) represents a correlation vector for the signal 1 starting from the sample p.

When array signal processing is performed using the weight w (p) described above, array output z (p) = w (p) ^H X (p) is obtained.
In the correlator 5, this array output and the known symbol sequence r ₁
The correlation coefficient f (p) with and is calculated according to the following equation.

[Equation 5] Here, ‖ ・ ‖ represents the norm of the vector. The correlation coefficient f (p) is calculated for various p, and the timing is detected based on the magnitude.

[0012]

In the method of the above-mentioned document [1], since the weight calculation and the correlation coefficient are calculated for each received sample, there is a problem that the calculation amount is large. Therefore, a technique that can establish synchronization with a smaller amount of calculation is an issue. Further, although the method of reference [1] deals only with timing synchronization,
In the initial synchronization, not only timing synchronization but frequency synchronization is important. However, there have been no examples of dealing with frequency synchronization when using an adaptive array, and a frequency synchronization method suitable for using an adaptive array has become a problem. Furthermore, in an environment where the interference signal exists before the arrival of the desired signal, it is considered that the synchronization accuracy can be improved by effectively using the received signal at the time when the desired signal does not exist. In the conventional synchronization method, it is also an issue to construct a highly accurate synchronization method that does not use the received signal before the arrival of the desired signal and effectively uses the section in which the desired signal does not exist.

[0013] Therefore, an object of the present invention is to provide a radio receiving apparatus capable of establishing initial synchronization by a method suitable for using the adaptive array. It is another object of the present invention to provide a wireless reception device using an adaptive array that can establish timing synchronization with a smaller amount of calculation. Further, it is another object of the present invention to provide a wireless reception device capable of performing frequency synchronization suitable when using the adaptive array.

[0014]

In order to achieve the above object, a radio receiving apparatus of the present invention is a radio receiving apparatus for receiving a signal using a plurality of antennas, and a known signal and a received signal included in a desired signal. The arrival timing of the desired signal is detected by using an operation that combines both the correlation vector v obtained as the correlation operation and the correlation matrix Φ obtained from the received signal. Further, the operation of combining the correlation vector v and the correlation matrix Φ is

[Equation 6] (H represents transposed conjugate). further,
As the correlation matrix, a correlation matrix calculated using the received signal at the time when the desired signal does not exist may be used. Furthermore, another radio receiving apparatus of the present invention is a radio receiving apparatus that receives a signal using a plurality of antennas, and a plurality of correlations obtained as a correlation calculation between a plurality of known signal sequences included in a desired signal and a received signal. The carrier frequency or frequency offset of the desired signal is determined using a vector and a correlation matrix obtained from the received signal. Here, as the correlation matrix, a correlation matrix calculated using a received signal at a time when a desired signal does not exist may be used. Furthermore, still another radio receiving apparatus of the present invention is a radio receiving apparatus that receives a signal using a plurality of antennas, and a plurality of a plurality of known signal sequences included in a desired signal and a plurality of obtained as a correlation calculation of the received signal. The timing of the desired signal is detected and the frequency offset is determined using the correlation vector and the correlation matrix obtained from the received signal, and the initial synchronization is completed.

[0015]

BEST MODE FOR CARRYING OUT THE INVENTION [Embodiment 1] FIG. 1 is the most basic diagram for explaining the outline of a radio receiving apparatus of the present invention.
Here, 1 is a plurality of antennas, and 2 is a timing detection unit 2 that detects the arrival timing of a desired signal based on the received signals of each antenna 1. In addition, the usage environment of the present invention is
Assume the same environment as the conventional method described above. That is, as shown in the temporal transmission signal diagram of FIG.
~ Signal K exists, signal 1 is the desired signal, and signals 2 to K are interference signals.

Next, the method of detecting the arrival timing of the desired signal executed in the timing detecting section 2 will be described. As a basic principle, first, a process of performing a correlation calculation between a received signal array output z (p) = w (p) ^H X (p) and a known symbol r ₁ will be discussed. Here, the weight w (p) = [w ₁ , ..., W _M ] ^T calculated with the sample p as the starting point is determined by the following equation as in the equation (3) of the conventional method described above.

[Equation 7]

Hereinafter, for simplification, w (p), X (p), z (p)
(= W (p) ^H X (p)), v ₁ (p) are respectively w, X, z, v ₁
The correlation between z and r ₁ is given by the following equation.

[Equation 8] This means that the above equation 4 is converted into ‖z‖ = (v ^H Φ ^-1 v) ^1/2 , ‖r‖ =
Obtained by transforming using the relationship of N ^1/2 , u (p)
= F (p) ² . Here, a modification of the following equation holds for u (p).

[Equation 9] According to the above equation (4), the correlation vector v ₁ and the correlation matrix Φ
u (p) can be obtained by the operation using (p), and u (p)
Can be used to perform timing detection. That is, as described above, u (p) = f (p) ² , and whichever is used to perform timing detection, the detected sample is the same.

Thus, the basic principle of this embodiment is
This method is the same as the conventional method shown in FIG. 9 in that the arrival timing of the desired signal is detected based on the correlation between z and r ₁ , but the array signal processing procedure as in the conventional method is used. Since u (p) can be directly calculated using v ₁ and Φ (p) instead of performing the correlation calculation with the known symbol on the signal processing output, the device configuration becomes simple. Further, the method of the present embodiment can detect the timing with a smaller amount of calculation than the conventional method.

There are several methods for determining the timing detection depending on the usage environment. The first is a method of handling as timing detection when the value of u (p) exceeds a certain threshold β, and this detection method is often used when the arrival time of a desired signal is completely unknown. In timing synchronization, the arrival time range of the desired signal may be known to some extent. In this case, u (p) is calculated for all possible samples p, and the sample p having the maximum u (p) is selected as the synchronization timing.

In order to verify the performance, there are 7 interfering signals under the half-wave spacing circular 8 antenna shown in FIG.
Assume an environment in which three desired signals arrive. The timing detection unit 2 uses u (p) for 20 possible samples.
Is calculated, and p that maximizes u (p) is selected, and timing detection is performed using this as the arrival timing of the desired signal. FIG. 3 shows the relationship between the timing detection error rate and the desired signal power when this method is used. In the figure, a curve 10 is the error rate characteristic of the correlation detection method in which u (p) = v ₁ ^H v ₁ / N which is often used in mobile communication that does not use an adaptive array. A curve 11 is the timing detection error rate characteristic when the present embodiment is used, and it is possible to suppress the timing detection error rate to a low level as compared with the case of the correlation detection method. The conventional system shown in FIG. 9 and this embodiment have the same timing detection error rate characteristic. However, even in that case, the amount of calculation can be suppressed to a lower level by the method of the present embodiment.

[Second Embodiment] Next, a second embodiment will be described in which the timing detection characteristic can be further improved in an environment where an interference signal exists before the arrival time of the desired signal. FIG. 4 is the most basic diagram for explaining the outline of the wireless reception device according to the present embodiment. Where 2
Reference numeral 0 represents the timing detection unit in the present embodiment.

In general array signal processing, the above equation (3) is often used. However, if a method of performing a correlation matrix operation using a received signal before the arrival time of the desired signal is introduced, the accuracy of the weight is further improved. (2) Yoshitaka Hara, Chihiro Fujita, Yoshihide Kamio, “Fast Initial Weight Determination Method for Adaptive Array Antenna”, IEICE Tech., RCS2002-1, Jan.2002).
In order to apply this principle to the synchronization method, the weight calculation of Expression (3) is changed to the following expression.

[Equation 10] Here, a sample (p-N _A) from a sample (p-1) is assumed that there is no desired signal, which is a section for performing only a correlation matrix calculation. w'represents the result of weight calculation using a section where the desired signal does not exist.

The correlation between the array output and the reference signal when the weight w '= w' (p) is used is given by the following equation.

[Equation 11]

[0024] Compared with the formula (4), the equation (6), the section performing the calculation samples (pN _A) ~ samples (p + N
−1), and the number of calculation samples of the correlation matrix w ′ is increasing. By increasing the number of calculation samples of the correlation matrix in this way, more accurate timing detection can be expected. Further, if the expression (6) is a special equation (4) (N _A =
It is included as 0). As described above, by performing the timing detection based on the equation (6) using the reception sample in which the desired signal does not exist, more accurate synchronization detection can be performed.

In order to verify the performance, there are 7 interference signals under the circular 8 antenna as in the first embodiment.
Assume an environment in which three desired signals arrive. The timing detection unit 20 uses u for 20 possible samples.
(p) is calculated, and p that maximizes u (p) is selected and timing detection is performed. The curve 12 in FIG. 3 shows the relationship between the timing detection error rate and the desired signal power when the method of this embodiment is used. It can be seen that the use of this method can further reduce the timing detection error rate as compared with the method of the first embodiment. As described above, according to the present embodiment, a highly accurate timing synchronization method can be realized.

[Third Embodiment] Normally, although the carrier frequency of an incoming desired signal can be predicted to some extent, it is difficult to accurately grasp it in advance. In the receiver, the received signal is frequency-converted into the baseband, but when the incoming carrier frequency and the frequency of the conversion signal are different, a frequency offset occurs. The present embodiment relates to frequency synchronization that compensates for frequency offset when using an adaptive array. FIG. 5 is the most basic diagram for explaining the outline of the wireless reception device of the present embodiment. In this figure, reference numeral 21 represents a frequency offset estimation unit in the present embodiment.

Generally, in the initial synchronization of a desired signal, frequency synchronization is performed after timing synchronization is performed. The frequency synchronization after detecting the arrival time timing (sample p) of the desired signal by the timing synchronization will be described below. In order to estimate the frequency offset, it is necessary to estimate the phase rotation in different time zones for the array output based on the specific weight w. For example, it is possible to estimate the frequency offset by estimating the phase change corresponding to the frequency offset between the two sampling intervals and converting it into the change amount per time.

Firstly, the correlation output of known symbol r ₁ and the array output of the N samples (w ^'H X (p)) for the sample p as a starting point is given by the same equation as the equation (6).

[Equation 12]

Next, the array output of N samples starting from the sample (p + N _frame ) and the correlation output of the known symbol r ₂ are given by the following equation. Here, the same correlation matrix Φ ′ as the above equation is used.

[Equation 13] The two correlation outputs u ₁ (p) and u ₂ (p) represent the phase of the frequency offset of the received signal at two times separated by N _frame samples. Therefore, the amount of phase rotation in the two sections is u ₁ (p) ^*
It can be estimated by the phase of u ₂ (p).

Here, the following expression holds for u ₁ (p) ^* .

[Equation 14] That is, u ₁ (p) is a real number and is in a phase 0 state. Therefore, the amount of phase rotation of u ₁ (p) ^* u ₂ (p) is equal to the amount of phase rotation of u ₂ (p). Therefore, the frequency offset Δf can be estimated by the following equation.

[Equation 15] Frequency synchronization can be established by compensating for this frequency offset Δf and performing frequency conversion of the received signal. That is, the carrier frequency can be determined based on the frequency offset Δf.

Note that Φ ′ is Φ ′ = X ′ (p) in the equation (5).
Although it is given by X '(p) ^H , the value of N _A included in X' (p) can be set arbitrarily. When N _A = 0 (that is, when Φ in Expression (3) is used), only a received signal having a desired signal is used, and when N _A > 0, a received signal having no desired signal is also included. Frequency offset is compensated.

[Embodiment 4] In the present embodiment, initial synchronization is performed which is a combination of the timing synchronization and the frequency synchronization described above. That is, the timing detection unit 20 or 21 and the frequency offset estimation unit 2
An initial synchronization unit having the function of 1 is provided. The initial synchronization section uses the timing synchronization method of the first or second embodiment and the frequency synchronization method of the third embodiment described above while using the correlation vectors v ₁ and v ₂ and the correlation matrix Φ (p), Control is performed in the order of synchronization. Therefore, it is possible to perform timing synchronization and frequency synchronization using common parameters. Thus, by performing frequency synchronization using the same parameters after performing timing synchronization, all initial synchronization can be established at the same time.

[0033]

As described above, according to the radio receiving apparatus of the present invention, it is possible to detect timing with a smaller amount of calculation than the conventional method. Further, the signal processing configuration becomes simpler than that of the conventional method. Further, according to the wireless reception device of the present invention, which estimates an interference signal by using a received signal at a time when a desired signal does not exist, it is possible to perform timing synchronization with higher accuracy than the conventional method. Furthermore, a frequency synchronization configuration when using an adaptive array, which has not been studied in the past, is provided, and by using the wireless reception device of the present invention, it is possible to perform frequency synchronization with high accuracy in the presence of an interference signal. . Furthermore, by estimating the interference signal using the received signal at the time when the desired signal does not exist, frequency synchronization can be performed with higher accuracy. Furthermore, according to the wireless reception device of the present invention that simultaneously performs timing synchronization and frequency synchronization using the same parameter, initial synchronization acquisition can be established with a small amount of signal processing. Moreover, since the timing synchronization and the frequency synchronization are collectively performed, the synchronization mode can be directly switched to the communication mode after the synchronization is acquired.

[Brief description of drawings]

FIG. 1 is a first embodiment of a wireless reception device of the present invention.
It is a figure which shows schematic structure of.

FIG. 2 is a diagram showing a receiver antenna arrangement for performing performance evaluation according to the first embodiment.

FIG. 3 is a diagram showing timing detection error rate characteristics when timing detection is performed using the first and second embodiments and the correlation detection method.

FIG. 4 is a second embodiment of the wireless reception device of the present invention.
It is a figure which shows schematic structure of.

FIG. 5 is a third embodiment of the wireless reception device of the present invention.
It is a figure which shows schematic structure of.

FIG. 6 is a diagram showing a spatial signal arrival model.

FIG. 7 is a diagram showing a temporal signal arrival model.

FIG. 8 is a diagram showing a signal processing configuration for weight calculation in the adaptive array.

FIG. 9 is a diagram showing a configuration for timing synchronization in a conventional method.

[Explanation of symbols]

1 antenna 2,20 Timing detector 21 frequency offset estimation unit 25,26 Weight calculator 3 Multiplier 4 Signal synthesizer 5 Correlator

Claims (6)

[Claims] 1. A radio receiving apparatus for receiving a signal using a plurality of antennas, both of a correlation vector v obtained as a correlation operation between a known signal included in a desired signal and a received signal and a correlation matrix Φ obtained from the received signal. A radio receiving apparatus, characterized in that the arrival timing of a desired signal is detected by using an operation in which the above are combined. 2. The operation that combines the correlation vector v and the correlation matrix Φ is as follows: The wireless reception device according to claim 1, wherein the wireless reception device is an operation including (H represents transposed conjugate). 3. The radio receiving apparatus according to claim 1, wherein a correlation matrix calculated using a received signal at a time when a desired signal does not exist is used as the correlation matrix. 4. A radio receiving apparatus for receiving a signal using a plurality of antennas, wherein a plurality of correlation vectors obtained as a correlation operation between a plurality of known signal sequences included in a desired signal and the received signal and a correlation obtained from the received signal. A radio receiving apparatus, wherein a carrier frequency or a frequency offset of a desired signal is determined using a matrix. 5. The radio receiving apparatus according to claim 4, wherein a correlation matrix calculated by using a received signal at a time when a desired signal does not exist is used as the correlation matrix. 6. A radio receiving apparatus for receiving a signal using a plurality of antennas, wherein a plurality of correlation vectors obtained as a correlation calculation between a plurality of known signal sequences included in a desired signal and the received signal and a correlation obtained from the received signal. A radio receiving apparatus, characterized in that a timing of a desired signal is detected and a frequency offset is determined by using a matrix to complete initial synchronization.