JP2008178044A - Ofdm receiver - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an OFDM receiver for compensating deterioration in reception quality, regardless of the fixed reception ambience and moving receive ambience, when two or more incoming waves which come from each and every direction are received. <P>SOLUTION: The OFDM receiver 1 includes an array antenna unit 10, a rectangular demodulator 20; a direction forming unit 30, an AFC unit 40; a signal selecting switching unit 50; and a diversity composite unit 60. Even if the incoming direction of OFDM signal received by the array antenna portion 10 is not known, the direction forming unit 30 forms an N-system directional beam in an adaptive way, in a M-MSN method or M-DCMP method so that each incoming wave is subjected to separation and reception. In the AFC unit 40, the impact of the Doppler effect by the moving receive is compensated. In the signal selecting switching unit 50, an L-system signal with high signal level is chosen. The diversity composite unit 60 selects one interpolating process, according to one of symbol periods, until the switching turn of the L-system signal which has been transmitted, among two or more interpolation processes for performing symbol filter process, so that maximal ratio combining is performed for each sub-carrier. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a reception technology in the field of digital broadcasting and digital transmission that employs OFDM (Orthogonal Frequency Division Multiplexing), and in particular, receives radio waves such as digital broadcasting and wireless LAN (Local Area Network). In particular, the present invention relates to a diversity combining technique for reducing the influence of multipath fading that occurs at the time, and an adaptive array antenna technique for removing the influence of interference waves.

  Non-Patent Document 1 and Patent Document 1 describe conventional reception techniques in digital broadcasting and digital transmission employing the OFDM method. In Non-Patent Document 1, arriving waves from various angles cause frequency shifts (hereinafter referred to as “Doppler shift”) due to mobile reception, and these are added together (hereinafter referred to as “multiple Doppler shift environment”). A technique for compensating for the generated inter-carrier interference of an OFDM signal using a beam space adaptive array antenna (hereinafter referred to as BSAAA) is described. The method of compensating using this BSAAA is to arrange a directional beam formed with a fixed weight coefficient in a multi-beam type, and cover all 360 degrees omnidirectionally, thereby receiving received waves from a plurality of directions of arrival. Separately received by each directional beam, each Doppler shift is compensated by AFC (Auto Frequency Control), and maximum ratio combining for each subcarrier is performed on all signals after AFC.

  Patent Document 1 discloses a technique for compensating for inter-carrier interference of OFDM signals generated by a multiple Doppler shift environment by a multi-beam type adaptive directivity control method that separates and combines different incoming waves with different directional beams. Are listed. This method does not require a reference signal, and directly determines the eigenvector beam (directional beam) corresponding to each incoming wave in the order of the first eigenvector and the second eigenvector by directly obtaining the eigenvector of the correlation matrix as a weighting coefficient. is there. In this method, it is possible to calculate at high speed by combining the process of power multiplication and reduction.

IEICE Trans Commun Vol. E87-B, no. 1, Page. 20-28, Jan. 2004, “Beam-Space Adaptive Array Antenna for Suppressing the Doppler Spread in OFDM Mobile Reception” JP 2006-54603 A

  In the method of Non-Patent Document 1 described above, a directional beam is formed using a fixed weight coefficient. Therefore, a desired wave arrives in the main beam direction of each directional beam, and an unnecessary wave is generated from a direction that can secure a dynamic range. When it arrives, the desired wave and the unnecessary wave can be separated and received, and inter-carrier interference caused by Doppler shift can be compensated. However, when an unnecessary wave comes from a direction in which the dynamic range cannot be secured, an unnecessary wave component is also received. For this reason, this method has a problem in that it is impossible to separately receive a desired wave and an unnecessary wave, and it may not be possible to compensate for inter-carrier interference caused by Doppler shift.

  Further, the above-described method of Patent Document 1 aims at reducing the amount of calculation required for converging eigenvectors, and an eigenvector obtained in a certain calculation interval is an initial eigenvector to be obtained in the next calculation interval. It is realized by using it as a vector (updated sequentially). Therefore, there is no problem when the fluctuation speed of the signal level between calculation sections is slow as in a reception environment that moves at a low speed. However, there is a problem that eigenvectors cannot be converged and reception quality deteriorates when the fluctuation speed of the signal level between calculation sections is fast as in a reception environment that moves at high speed.

  Therefore, the present invention has been made to solve the above-mentioned problems, and its purpose is to receive quality when receiving a plurality of incoming waves arriving from all directions regardless of a fixed reception environment and a mobile reception environment. An object of the present invention is to provide an OFDM receiver capable of compensating for deterioration.

  In order to solve the above-mentioned problem, the invention of claim 1 is a digital transmission receiver using an OFDM modulation method, wherein an array antenna unit configured with K multiple antennas as array elements, and the array antenna unit A quadrature demodulator that orthogonally demodulates OFDM signals for K elements received via the IQ signal, and a plurality of N directional beams based on the IQ signal, corresponding to the N directional beams A directivity forming unit that generates an array combined signal from an array combining weighting factor and the IQ signal distributed to the N systems, and generates an N system signal obtained by separating the incoming waves to the array antenna unit, and the generated For each of the N system signals, the AFC unit that compensates the frequency shift and each signal level of the compensated N system signals are measured, and the higher L system signals having the higher signal levels are measured. A signal selection switching unit for selecting and switching, and a diversity combining unit for combining the L system signals for each subcarrier with a maximum ratio, and the directivity forming unit is set in advance for the orthogonally demodulated IQ signal. A sample buffer unit that accumulates data corresponding to the number of samples, a distribution unit that distributes IQ signals read from the sample buffer unit, and a correlation matrix that includes correlation values of signals for K elements in the distributed IQ signals , Based on a vector having angle-of-arrival information corresponding to the N systems, an array synthesis weight coefficient calculation unit that calculates N system array synthesis weight coefficients, and the distributed N system IQ signals, A delay adjusting unit for delaying the processing time in an array combining weight coefficient calculation unit; the delayed N system IQ signals; and the N system array combining weight coefficients; Characterized by comprising a array combined signal generator for generating a Luo array combined signal.

  According to a second aspect of the present invention, the array synthesis weight coefficient calculation unit of the directivity forming unit includes correlation values of signals corresponding to K elements in the distributed IQ signal by the MSN technique (Maximal SNR Method). Based on the correlation matrix and the steering vector having angle-of-arrival information corresponding to the N systems, N system array synthesis weight coefficients are calculated.

  According to a third aspect of the present invention, the array synthesis weight coefficient calculation unit of the directivity forming unit uses a DCMP technique (Directly Constrained Minimization of Power Method) to correlate signals for K elements in the distributed IQ signal. A weight coefficient for array synthesis of N systems is calculated based on a correlation matrix composed of values and a constraint vector having arrival angle information corresponding to the N systems.

  According to a fourth aspect of the invention, the signal selection switching unit measures each signal level of the N system signals compensated by the AFC unit, and outputs a system number and a level value, and the output Based on the determined system number and the level value, based on the determined system number of the L system, a level determination unit that determines the system number of the higher L system of the higher level value of the N system numbers, And a selection switching unit for selectively switching an L-system signal among the N-system signals and outputting the L-system signal after being delayed by a preset time.

  According to a fifth aspect of the present invention, the diversity combining unit inputs a transmission path characteristic calculated from a frequency domain signal in the L-system signal and performs symbol filter processing, and the signal selection switching unit. The system number of the L system is input from the level determination unit, and a signal for specifying the symbol filter processing is output based on the system number of the L system and the delay time in the selection switching unit of the signal selection switching unit A symbol filter designating unit, and combining the L system signals for each subcarrier with a maximum ratio, and the symbol filter unit based on a signal for specifying symbol filter processing from the symbol filter designating unit, Depending on the selection switching of the signal selection switching unit, signals of different systems are sent to the diversity combining unit 3 for the target symbol for transmission path estimation. The first interpolation process for estimating the transmission line characteristic by linear interpolation using the transmission line characteristic at each SP position separated by 4 symbols in the time direction when arriving at a delay of at least a symbol. Is delayed by 2 symbols with respect to the target symbol of the transmission path estimation, the transmission path characteristics obtained by linear interpolation using the transmission path characteristics of each SP position separated by 4 symbols in the time direction, and the time direction A second interpolation process is performed to estimate the transmission path characteristics from the transmission path characteristics at each SP position separated by 1 symbol and the transmission path characteristics at the SP position. The transmission path characteristics are determined by the transmission path characteristics at each SP position that is 1 symbol apart in the time direction and the transmission path characteristics that are 0 by deleting the transmission path characteristics at the original SP position. Guess The third performs the interpolation process of the signal of the different strains when only target symbol of channel estimation, and performing a fourth interpolation process is not performed the symbol filtering.

  The invention according to claim 6 does not include the signal selection switching unit of the OFDM receiver according to claim 1, and includes the array antenna unit, orthogonal demodulation unit, directivity forming unit, AFC unit, and diversity combining. And N = L.

  According to a seventh aspect of the present invention, in the OFDM receiving apparatus according to any one of the first to fifth aspects, the arrangement positions of the AFC unit and the signal selection switching unit are interchanged, and the AFC unit is placed after the signal selection switching unit. It comprises a part.

  As described above, according to the present invention, when receiving a plurality of incoming waves arriving from all directions regardless of a fixed reception environment or a mobile reception environment, an OFDM receiver capable of compensating reception quality degradation. Can be realized.

The best mode for carrying out the present invention will be described below in detail with reference to the drawings.
[Example 1]
First, Example 1 will be described. FIG. 1 is a system configuration diagram illustrating an OFDM receiving apparatus according to a first embodiment of the present invention. The OFDM receiving apparatus 1 includes an array antenna unit 10 configured with K antennas as array elements, a received OFDM signal for K elements, and an IQ signal (I: In-Phase (in-phase component)), Q: A quadrature demodulator 20 for obtaining a quadrature-phase (orthogonal component)), and adaptively forming N directional beams, N weights for array synthesis and N signals from the IQ signal from the quadrature demodulator 20 The directivity forming unit 30 that calculates the array composite signal and the directivity forming unit 30 separates and receives each incoming wave with N types of adaptive directional beams (hereinafter referred to as separate reception), and then performs Doppler shift compensation. Measure each signal level of the AFC unit 40 that generates and outputs the N system signals and the N system composite signal from the AFC unit 40, and select the upper L system with the higher measured value. A signal selection switching unit 50 which is output, and an output signal of the L system from the signal selection switching unit 50, and a diversity combining unit 60 for maximum ratio combining for each subcarrier. Details of each part will be described below.

[Array antenna section]
The array antenna unit 10 is an antenna unit configured with K antennas as each array element, and can be an array antenna unit having various shapes depending on the arrangement. FIG. 2 is a diagram showing an arrangement example of array elements of the array antenna unit 10 shown in FIG. In FIG. 2, (a) shows a linear array in which each array element is arranged on a line. (B) shows a planar array in which each array element is arranged on a plane. (C) shows a circular array in which the array elements are arranged in a circular shape. The arrangement of the array antenna unit 10 shown in FIGS. 2A to 2C is an example, and may be configured by other arrangements. Further, the array elements may be equally spaced or unequal.

(Quadrature demodulator)
The quadrature demodulator 20 receives received OFDM signals for K elements from the array antenna unit 10, generates an IQ signal by orthogonally demodulating these received OFDM signals, and outputs an IQ signal for K elements. Here, the IQ signal is an equivalent baseband signal. Hereinafter, the IQ signal will be described as an equivalent baseband signal. Note that orthogonal demodulation processing may be performed with an IF (Intermediate Frequency) signal and then converted into an equivalent baseband.

  FIG. 3 is a diagram showing a configuration of the orthogonal demodulation unit 20 shown in FIG. The quadrature demodulator 20 includes a KPF BPF (Band Pass Filter) unit 21, a local signal generator 22, a π / 2 phase shifter 26, two K distributors 23-1, 23-2, (K It is configured to include (elements × 2) multiplication units 24-1 and 24-2 and (K elements × 2) LPF units 25-1 and 25-2.

  The BPF unit 21 receives the received OFDM signals for K elements from the array antenna unit 10 and passes only signals in a preset frequency range. The local signal generation unit 22 generates a preset local signal, and the K distribution unit 23-1 distributes the local signal to signals of the same level and the same phase by the number of K elements. Further, the K distributing unit 23-2 distributes the local signal, the phase of which is shifted by π / 2 by the π / 2 phase shifter 26, to the signals of the same level and the same phase by the number of K elements. Multipliers 24-1 and 24-2 multiply (orthogonally demodulate) the signals after passing through BPF 21 and the local signals distributed by K distributors 23-1 and 23-2, respectively, for K elements. I signal and Q signal are output respectively. The LPF unit 25 inputs IQ signals for K elements, and allows only signals below a preset cutoff frequency to pass through. The IQ signals for the K elements generated in this way are sent to the directivity forming unit 30.

(Direction forming part)
The directivity forming unit 30 inputs an IQ signal for K elements from the quadrature demodulating unit 20, adaptively forms N directional beams, and generates an array combining weight coefficient for generating the directional beams; N system array synthesis signals (IQ signals) are generated from the input IQ signals and output.

  FIG. 4 is a diagram showing a configuration of the directivity forming unit 30 shown in FIG. The directivity forming unit 30 includes an M sample buffer unit 31, an (N + 1) distribution unit 32, an array synthesis weight coefficient calculation unit 33, an N system delay adjustment unit 34, and an N system array synthesis signal generation unit 35. Configured.

  The M sample buffer unit 31 receives the IQ signals for the K elements from the quadrature demodulation unit 20, accumulates and writes the IQ signals for M samples in accordance with an input / output timing signal from an input / output timing generation unit (not shown), read out. The M sample buffer unit 31 is configured to read data in the order of writing when data is written and read as in a FIFO (First In First Out) configuration. The sample unit for output is specified. The (N + 1) distribution unit 32 receives the IQ signal in units of M samples from the M sample buffer unit 31, distributes the signals having the same signal level and synchronized in time to the (N + 1) system and outputs them. To do. Here, the case where the FIFO is used as an example of the memory that is the M sample buffer unit 31 is shown, but the present invention is not limited to this, and a similar process may be used instead.

The array synthesis weight coefficient calculation unit 33 inputs one system IQ signal out of the (N + 1) system IQ signals distributed by the (N + 1) distribution unit 32 and inputs a step angle from a step angle generation unit (not shown). Then, the optimum weight W _opt (n) of the weight coefficient for array synthesis is calculated by the method described later. The optimal weight W _opt (n) is expressed by a vector, and W _opt (n) = [w ₁ (n), w ₂ (n),..., W _k (n)] (hereinafter the same) .) Here, n is a directional beam number (the maximum is N), and is increased or decreased according to the step angle A as shown in equation (1) described later. Note that the number N of IQ signal lines distributed by the (N + 1) distribution unit 32 is set in advance by a preset step angle A.

The delay adjustment unit 34 inputs each of the remaining N system IQ signals of the (N + 1) system IQ signals distributed by the (N + 1) distribution unit 32, and sets the processing time of the array synthesis weight coefficient calculation unit 33. The delay is adjusted by the time taken into consideration, and the delay adjusted IQ signal is output. The array synthesis signal generation unit 35 calculates the optimum weight W _opt (n) of the array synthesis weight coefficient calculated by the array synthesis weight coefficient calculation unit 33 and the IQ signal [x ₁ , x ₂ ,. .., X _k ], as shown in FIG. 5, each of the element components of the K elements is added by the adder 37 after the complex multiplication by the multiplier 36, and the N-system array composite signal y (n) is obtained. Generate. The N systems of combined output signals (IQ signals) generated in this way are sent to the AFC unit 40.

  Hereinafter, with respect to the algorithm for forming the adaptive directivity in the weight coefficient calculation unit 33 for array synthesis, a case where an M-MSN (Multiple-Maximal SNR Method) adaptive array method is used, and a case of M-DCMP (Multiple-Directional Constrained Minimized Preference). ) Each case where the adaptive array method is used will be described.

[M-MSN adaptive array method]
The MSN (maximum SNR method) adaptive array method is a method of adaptively controlling directivity under the condition that the arrival direction (V: steering vector) of a desired wave is known. Since the mobile reception environment of terrestrial digital broadcasting can be considered as a multiple Doppler shift environment for receiving a plurality of desired waves, the optimum weights of N directional beams each having a plurality of directions of arrival as reference signals are obtained. An effective method is to separate the incoming wave components and synthesize diversity at a later stage. For details of the MSN adaptive array method, refer to the document “Kikuma,“ Adaptive Signal Processing by Array Antenna ”, Science and Technology Publishing”.

The M-MSN adaptive array system described here is an extension of the MSN adaptive array system. In order to estimate the arrival directions of a plurality of desired waves, the M-MSN adaptive array system can be obtained in step A as shown in the equation (1). An optimum weight is obtained using steering vectors for N systems.

The optimum weight W _opt (n) of the M-MSN adaptive array can be obtained by equation (2). R _xx is a correlation matrix. As described above, the optimum weights W _opt (n) and V (n) are represented by vectors.

  FIG. 6 is a flowchart showing the procedure for calculating the optimum weight obtained by the array synthesis weight coefficient calculator 33 shown in FIG. Hereinafter, description will be made with reference to the flowchart of FIG.

ここで、Ｍはサンプル数、Ｋはアレー素子数、Ｈは複素共役転置、Ｔは転置を示す。 (Step 501)
First, the array synthesis weight coefficient calculator 33 calculates R _xx (correlation matrix). The correlation matrix R _xx is a matrix representation of the correlation value of the received signal of each array element, takes the inner product of the input vector X (t) and the complex conjugate transpose X (t) ^H of the input vector, and takes M samples It can be obtained by performing the temporal addition average of.

Here, M is the number of samples, K is the number of array elements, H is a complex conjugate transposition, and T is transposition.

(Step 502)
Next, the array synthesis weight coefficient calculator 33 calculates an inverse matrix R _xx ⁻¹ of the correlation matrix R _xx . The inverse matrix R _xx ⁻¹ can be obtained by using a calculation algorithm such as Gaussian elimination or LU decomposition. For details of the inverse matrix calculation method, see Togawa, “Science and Technology Calculation Handbook”, Science Corporation.

(Step 503)
Next, the array synthesis weight coefficient calculation unit 33 calculates V (n) (steering vector). The steering vector V (n) determines the main lobe direction of the directional beam, and is expressed by equation (4).

Note that υ _K (n) in the case of a linear array in which K elements are arranged at equal intervals is expressed by equation (5).

(Step 504)
Next, the array synthesis weight coefficient calculation unit 33 calculates the optimum weight W _opt (n).

(Step 505)
Since the optimum weight W _opt (n) has N values, the array synthesis weight coefficient calculation unit 33 calculates N steps 503 and 504 described above. Note that the N types of calculations are not performed in time series, but are performed simultaneously.

[M-DCMP adaptive array system]
DCMP (Directionally Constrained Output Power Minimization Method) Adaptive array method specifies the constraint direction (C: constraint matrix) of the desired wave (s), leaves the components protected by the constraint conditions intact, This is a technique for minimizing the output power so as to suppress the component. The optimal weight W _opt (n) of the DCMP adaptive array is shown in equation (7). The optimum weight W _opt (n) can be obtained by weighting the desired wave component restrained by the restraint matrix C with the restraint response vector H. For details of the DCMP adaptive array method, refer to the document “Kikuma,“ Adaptive Signal Processing by Array Antenna ”, Science and Technology Publishing”.

  Since the mobile reception environment of terrestrial digital broadcasting can be considered as a multiple Doppler shift environment for receiving a plurality of desired waves, the optimum weights of N directional beams each having a plurality of directions of arrival as reference signals are obtained. An effective method is to separate the incoming wave components and synthesize diversity at a later stage. The M-DCMP adaptive array system described here is an extension of the DCMP adaptive array system. Like the M-MSN adaptive array system, a single constraint condition is used to estimate the arrival directions of a plurality of desired waves. The optimum weight is obtained using the constraint vector C (n) (= steering vector) for N systems obtained in the A degree step when the constraint response vector is H = 1.

以下、最適ウェイトの計算手順を、図６のフローチャート図に従って説明する。 Therefore, the optimum weight W _opt (n) of the M-DCMP adaptive array is expressed by equation (10).

Hereinafter, the procedure for calculating the optimum weight will be described with reference to the flowchart of FIG.

(Steps 501 and 502)
The array synthesis weight coefficient calculator 33 calculates R _xx (correlation matrix) and calculates an inverse matrix R _xx ⁻¹ of the correlation matrix R _xx . This calculation process is as described above.

(Step 503)
Next, the array synthesis weight coefficient calculator 33 calculates a constraint vector C (n). The constraint vector C (n) determines the main lobe direction of the directional beam and is expressed by the equation (11).

Note that c _K (n) in the case of a linear array in which K elements are arranged at equal intervals is expressed by equation (12).

(Step 504)
Next, the array synthesis weight coefficient calculation unit 33 calculates the optimum weight W _opt (n).

(Step 505)
Since the optimum weight W _opt (n) has N values, the array synthesis weight coefficient calculation unit 33 calculates N steps 503 and 504 described above. Note that the N types of calculations are not performed in time series, but are performed simultaneously.

[AFC Department]
Returning to FIG. 1, the AFC unit 40 separates and receives the incoming waves by the directivity forming unit 30 using N directivity beams, and then receives the N composite array signals (IQ signals) from the directivity forming unit 30. Input, estimate and compensate for Doppler shift frequency, and output IQ signal after Doppler compensation.

  FIG. 7 is a diagram showing a configuration of the AFC unit 40 shown in FIG. The AFC unit 40 includes effective symbol length delay units 41-1 and 41-2, multiplication units 42-1 and 42-2, moving average units 43-1 and 43-2, a Doppler frequency calculation unit 44, and a frequency correction unit. 45 is provided for each of N systems.

The effective symbol length delay unit 41-1 receives the I signal I (t) from the directivity forming unit 30, and generates an I signal I (t + T _u ) delayed by an effective symbol period _Tu . The effective symbol length delay unit 41-2 receives the Q signal Q (t) from the directivity forming unit 30, and generates a Q signal Q (t + T _u ) delayed by an effective symbol period _Tu .

The multiplier 42-1 multiplies (correlates) the I signal I (t) from the directivity forming unit 30 and the I signal I (t + T _u ) delayed by the effective symbol length delay unit 41-1. . Further, the multiplying unit 42-2 multiplies the I signal I (t) from the directivity forming unit 30 by the Q signal Q (t + T _u ) delayed by the effective symbol length delay unit 41-2 (correlation is performed). Take).

The moving average unit 43-1 generates Sii (t) obtained by taking a moving average of samples (NT _g ) in the guard interval period for the signal multiplied by the multiplying unit 42-1. Further, the moving average unit 43-2 generates Siq (t) obtained by taking a moving average of samples (NT _g ) in the guard interval period for the signal multiplied by the multiplying unit 42-2. These signals Sii (t) and Siq (t) are sent to the Doppler frequency calculation unit 44.

Sii (t) and Siq (t) are calculated by equations (14) and (15), respectively.

The Doppler frequency calculation unit 44 performs addition averaging for I symbols for each OFDM symbol from Sii (t) and Siq (t) calculated by the moving average units 43-1 and 43-2, and to calculate the Doppler frequency _{f d.}

The frequency correction unit 45 inputs the IQ signal from the directivity forming unit 30 and the Doppler frequency f _d from the Doppler frequency calculation unit 44, respectively, and converts the Doppler frequency f _d into a phase rotation amount θ _fd (17). The IQ signal after Doppler compensation is obtained by the equations (18) and (18).

  Such processing is performed for N systems, and the obtained IQ signals of N systems after Doppler compensation are sent to the signal selection switching unit 50. Although the above-described Doppler compensation method by the AFC unit 40 has been described by taking the feedforward type as an example, there are many other Doppler compensation methods, and the present invention is not limited to the feedforward type.

[Signal selection switching section]
The signal selection switching unit 50 receives the N system IQ signals after Doppler compensation from the AFC unit 40, measures each signal level, and sets a predetermined L system (L < For N), the upper L system having a high signal level measurement value is selected, and the IQ signal of the L system is output.

  FIG. 8 is a diagram showing a configuration of the signal selection switching unit 50 shown in FIG. The signal selection switching unit 50 includes an N-system M sample level measurement unit 51, a level determination unit 52, and a selection switching unit 53. The selection switching unit 53 includes an input port 54, an output port 55, and an L-system delay unit 56.

The M sample level measuring unit 51 receives the N system IQ signals from the AFC unit 40, measures the average signal level in the M sample period by the equation (19), and determines the system number (1, 2,..., N ) And a level value (for example, −10 dBm) are output to the level determination unit 52.

  The level determination unit 52 inputs the system number and the level value from the M system M sample level measurement unit 51, selects the signal of the upper L system having a higher signal level from the N system signal based on the level value, and selects The system number is output to the selection switching unit 53 and a symbol filter designating unit 69 (see FIG. 9) described later.

  The selection switching unit 53 inputs the system number of the L system from the level determination unit 52 and, based on the system number, out of the N system IQ signals input from the M sample level measurement unit 51 to the input port 54, the output port In order to switch to the L system IQ signal output from 55, the connection between the input port 54 and the output port is switched. Then, the delay unit 56 of the selection switching unit 53 corresponds to an L-system IQ signal from the output port 55 for the preparation time of the switching corresponding process in a symbol filter unit 63 (see FIG. 9) of the diversity combining unit 60 described later. Output with delay.

  When the diversity combining unit 60 described later uses a general transmission path estimation method in the transmission path estimation processing, it performs time interpolation processing using a total of 7 symbols including symbols that are transmission path estimation targets. In some cases, transmission path estimation is performed using the signal of the system selected by the selection switching unit 53 and the signal of the new system selected after selection switching, resulting in degradation of reception quality. Therefore, the selection switching unit 53 includes an L-system delay unit 56 and causes a delay of several symbols, so that the transmission path estimation in accordance with the switching timing in the selection switching unit 53 in the symbol filter unit 63 of the diversity combining unit 60. In order to perform the filtering process, a preparation time for the switching corresponding process is given. Specifically, the delay unit 56 provides data accumulation for at least 7 symbols when performing interpolation processing I (see FIG. 11) described later in the symbol filter unit 63, and for at least 5 symbols when performing interpolation processing II. Therefore, when performing interpolation process III, it is necessary to provide data accumulation for at least three symbols, and when performing interpolation process IV, it is necessary to provide data accumulation for at least one symbol. A delay of the symbol period shall be given. It is assumed that the target symbol for transmission path estimation is the fourth symbol data out of a total of 7 symbol data.

  Further, in order to prevent frequent selection switching in the selection switching unit 53, it is assumed that the signal once selected does not change the connection between the input port 54 and the output port 55 until it is not selected. In the above-described signal selection switching unit 50, the signal level is used as the signal selection reference. However, the signal quality (MER: Modulation Error Ratio) may be used as the selection reference.

[Diversity synthesis department]
The diversity combining unit 60 receives the L system IQ signal from the signal selection switching unit 50, performs the maximum ratio combining for each subcarrier when the L system diversity combining is performed, and outputs the IQ signal after the maximum ratio combining. To do.

  FIG. 9 is a diagram showing a configuration of diversity combining unit 60 shown in FIG. The diversity combining unit 60 includes an FFT unit 61, an SP extracting unit 62-1, an SP generating unit 62-2, a dividing unit 62-3, a symbol filter unit 63, a carrier filter unit 64, a diversity combining weight coefficient calculating unit 65, The delay adjustment unit 66 and the multiplication unit 67 are configured to include L systems, respectively, and further include a symbol filter designating unit 69 and an in-phase synthesis unit 68.

The FFT unit 61 receives an L-system IQ signal from the signal selection switching unit 50 and converts a time-domain signal into a frequency-domain signal. The SP extraction unit 62-1 receives the frequency domain signal from the FFT unit 61 and extracts an SP signal that serves as a reference signal for demodulation.
The SP position kp is defined by equation (20).

を求めることができる。 The SP generator 62-2 generates a known SP signal. The division unit 62-3 performs complex division on the SP signal extracted by the SP extraction unit 62-1 and the known SP signal generated by the SP generation unit 62-2. As a result, transmission path characteristics (amplitude and phase) at the SP position

Can be requested.

The symbol filter unit 63 performs symbol filter processing that is interpolation processing in the time direction, and the carrier filter unit 64 performs carrier filter processing that is interpolation processing in the frequency direction. Each transmission line characteristic H _l (s, c) is estimated. The diversity combining weight coefficient calculation unit 65 obtains the weight coefficient W _l (s, c) for each subcarrier for maximum ratio combining using the transmission path characteristic H _l (s, c). Multiplier 67 complex-multiplies weighting factor W _l (s, c) from diversity combining weighting factor calculator 65 and the other output signal from FFT unit 61. Here, the other output signal from the FFT unit 61 is sent from the delay adjustment unit 66 to the SP extraction unit 62-1, the division unit 62-3, the symbol filter unit 63, the carrier filter unit 64, and the diversity combining weight coefficient calculation. The processing time of the unit 65 is delayed. The in-phase synthesis unit 68 performs in-phase synthesis of the output signals of the respective systems from the multiplication unit 67 and outputs them. Here, s is a symbol number, c is a subcarrier number, and l is an input number to the diversity combining unit.

  Hereinafter, symbol filter processing, carrier filter processing, maximum ratio combining processing, and a transmission path estimation method will be described.

[Symbol filter processing]
A straight line interpolation method will be described as an example of symbol filter processing by the symbol filter unit 63.

、及びこのシンボルよりも４シンボル後のＳＰの位置の伝送路特性

が除算部６２−３により既に算出されていると仮定すると、残りのシンボルは、まずシンボル間の差分量ΔＨ_ｌ，ｋｐを（２１）式により計算し、残りの各シンボルの伝送路特性を（２２）式により推定する。

Transmission path characteristics at SP position

, And transmission path characteristics at the SP position 4 symbols after this symbol

Is already calculated by the dividing unit 62-3, first, the remaining symbols are calculated by calculating the difference amount ΔH _{l, kp} between the symbols by the equation (21), and the transmission path characteristics of the remaining symbols are expressed by ( 22) Estimate by the equation.

[Carrier filter processing]
Next, the FFT method will be described as an example of carrier filter processing by the carrier filter unit 64. In the FFT method, first, zero interpolation is performed in addition to the transmission path characteristics estimated by the symbol filter unit 63.

Further, in order to reduce the error at the end of the OFDM signal band, IFFT is performed by inserting vectors indicated by A and B at the left end and the right end of the OFDM signal band. cl is the total number of OFDM carriers.

に乗算し、フーリエ変換して伝送路特性Ｈ_ｌ（ｓ，ｃ）を得る。

After that, the band limiting filter coefficient F is set to remove aliasing.

Is multiplied by Fourier transform to obtain a transmission path characteristic H _l (s, c).

はＨ_ｌの共役複素数である。

[Maximum ratio synthesis processing]
The diversity combining weight coefficient calculation unit 65 calculates the diversity combining weight coefficient for maximum ratio combining from the transmission path characteristics H _l (s, c) obtained by performing the two-dimensional filter of the symbol filter processing and the carrier filter processing. Is calculated for each subcarrier. The weight ratio between the systems is determined by the size of the transmission path estimation value as shown in equation (29). still,

Is the conjugate complex number of H _l .

[Transmission path estimation method]
Under the influence of multipath fading, mobile reception characteristics and multipath tolerance differ depending on the interpolation processing of the symbol filter unit 63. Here, a general transmission path estimation scheme and a transmission path estimation scheme for high-speed mobile reception will be described. As shown in FIG. 10 (a), a general transmission path estimation method is a method of linear interpolation using SPs that are separated by a total of 7 symbols in time, and is therefore easily affected by time fluctuations. However, in a multipath environment, transmission path estimation is performed using SPs that are interpolated every three carriers in the frequency direction, so that it is not easily affected by multipath waves having a long delay time. On the other hand, as shown in FIG. 10B, the transmission path estimation method for high-speed mobile reception is a method that performs only the interpolation process in the frequency direction without performing the interpolation in the time direction. . However, in a multipath environment, only the SP for every 12 carriers in the frequency direction is used, so that it is easily affected by a multipath wave having a long delay time.

[Symbol filter processing for switching measures]
Next, the symbol filter processing for switching countermeasures by the symbol filter unit 63 will be described. In this process, the interpolation process is sequentially switched based on the system number from the signal selection switching unit 50 and the delay time (preset delay time) in the delay unit 56 of the signal selection switching unit 50. The symbol filter designating unit 69 receives the system number from the level determining unit 52 of the signal selection switching unit 50 and constantly monitors the system number, so that a signal of a system different from the system signal currently selected can be obtained after what symbol. You can know if it will be sent. In other words, when the symbol filter designating unit 69 determines that the input system number has changed, the symbol filter designating unit 69 starts the symbol corresponding to a preset time for delaying the signal by the delay unit 56 of the signal selection switching unit 50 from that time. The switching generation line to which a new system signal is sent is recognized. Therefore, the symbol filter designating unit 69 recognizes the timing of the switching generation line based on the timing when the input system number changes and the subsequent time count, and uses the interpolation processes I to IV shown in FIG. To do. Hereinafter, β is a non-negative integer. The priority order of the interpolation processing is I>II>III> IV.

(Interpolation process I)
This interpolation processing I is performed when the symbol filter designating unit 69 determines that a signal of a different system arrives at a delay of 3 symbols or more with respect to a target symbol for transmission path estimation with respect to a signal of the currently selected system. use. The same processing as the symbol filter processing described above is performed.

(Interpolation process II)
This interpolation process II is performed when the symbol filter designating unit 69 determines that a signal of a different system arrives at a delay of two or more symbols with respect to a target symbol for transmission path estimation with respect to a signal of the currently selected system. use. (A) a transmission line response obtained by linear interpolation using transmission line responses at each SP position separated by 4 symbols in the time direction, and (b) one symbol in the time direction with respect to the target section of the symbol filter A transmission path in the time direction is estimated using the transmission path response at each remote SP position and (c) the transmission path response at the SP position.

(Interpolation III)
This interpolation process III is performed when the symbol filter designating unit 69 determines that a signal of a different system arrives at a delay of one symbol or more with respect to a target symbol for transmission path estimation with respect to a signal of the currently selected system. use. (D) Transmission path response at each SP position separated by one symbol in the time direction with respect to the target section of the symbol filter, and (e) Transmission path response that is 0 by deleting the transmission path response at the original SP position. To estimate the transmission path.

[Interpolation IV]
This interpolation process IV is used when the symbol filter designating unit 69 has only a target symbol for transmission path estimation for a signal of a different system with respect to the signal of the currently selected system. In this process, the symbol filter process is not performed.

  As described above, when the symbol filter designating unit 69 determines that the input system number has changed, the symbol corresponding to the delay time set in advance in the delay unit 56 is that a new system signal is sent. It is determined that it is after (after 7 symbols). In this case, the symbol filter unit 63 continues the interpolation process I. When the symbol filter designating unit 69 determines that the new system signal is sent two symbols after the target symbol for transmission path estimation based on the time count, the symbol filter designating unit 69 interpolates the symbol filter unit 63. Process II is performed. Further, when it is determined that the symbol is one symbol after the target symbol for transmission path estimation, the symbol filter unit 63 is caused to perform the interpolation process III. When it is determined that there is only a target symbol for transmission path estimation, the symbol filter unit 63 is caused to perform interpolation processing IV.

  As described above, according to the OFDM receiver 1 of the first embodiment, it is possible to compensate for reception quality degradation when receiving a plurality of incoming waves arriving from all directions regardless of a fixed reception environment and a mobile reception environment. Is possible.

[Example 2]
Next, Example 2 will be described. FIG. 12 is a system configuration diagram showing an OFDM receiver according to Embodiment 2 of the present invention. The OFDM receiving apparatus 2 includes an array antenna unit 10, an orthogonal demodulation unit 20, a directivity forming unit 30, an AFC unit 40, and a diversity combining unit 60. Comparing the OFDM receiver 1 according to the first embodiment shown in FIG. 1 with the OFDM receiver 2, the OFDM receiver 2 does not include the signal selection switching unit 50 in the configuration of the OFDM receiver 1. Is different.

  That is, the OFDM receiver 2 includes an array antenna unit 10 configured with K antennas as array elements, an orthogonal demodulator 20 that demodulates received OFDM signals for K elements to obtain an IQ signal, and N systems of directivity. A directivity forming unit 30 for adaptively forming a directional beam and calculating an N system array composite signal from the N system array synthesis weight coefficients and the IQ signal from the orthogonal demodulator 20; After each incoming wave is separated and received by the N adaptive directional beams, the AFC unit 40 that generates and outputs the N system signals compensated for Doppler shift, and the N system from the AFC unit 40 (in this case, N = L ) And a diversity combining unit 60 that combines the maximum ratio for each subcarrier. Since the details of each part have been described above, a description thereof will be omitted.

  According to the OFDM receiving apparatus 2 of the second embodiment, since the signal selection switching unit 50 is not provided as compared with the OFDM receiving apparatus 1 of the first embodiment, the apparatus configuration is simplified as a whole in addition to the effects of the first embodiment. Become. This configuration is suitable when the number of directional beams (N) of the directivity forming unit 30 is as small as about 1 to 4, for example.

Example 3
Next, Example 3 will be described. FIG. 13 is a system configuration diagram showing an OFDM receiver according to Embodiment 3 of the present invention. The OFDM receiver 3 includes an array antenna unit 10, an orthogonal demodulation unit 20, a directivity forming unit 30, a signal selection switching unit 50, an AFC unit 70, and a diversity combining unit 60. Comparing the OFDM receiver 1 according to the first embodiment shown in FIG. 1 with the OFDM receiver 3, the OFDM receiver 3 is configured such that the AFC unit 70 is provided in the subsequent stage of the signal selection switching unit 50, and Since the N system signals are input to the AFC unit 40 of the OFDM receiver 1, the system is configured by N systems of processing units, whereas the AFC unit 70 of the OFDM receiver 3 has an L system signal. Is input, and thus is different in that it is configured by processing units for L systems.

  That is, the OFDM receiver 3 includes an array antenna unit 10 configured with K antennas as array elements, an orthogonal demodulator 20 that demodulates received OFDM signals for K elements and obtains IQ signals, and N systems of directivity. A directivity forming unit 30 that adaptively forms a directional beam and calculates an N system array composite signal from the N system array synthesis weight coefficients and the IQ signal from the orthogonal demodulator 20; Each signal level of the N array composite signals is measured, and a signal selection switching unit 50 for selectively switching and outputting the upper L systems of the measured values, and a directivity forming unit 30 with each of the N adaptive directional beams. An AFC unit 70 that receives and separates incoming waves, and selects and switches the L system by the signal selection switching unit 50, and then generates and outputs an L system signal that has been compensated for Doppler shift, and the L system from the AFC unit 70 Configured to output signals, and a diversity combining unit 60 for maximum ratio combining for each subcarrier. Since the details of each part have been described above, a description thereof will be omitted.

  According to the OFDM receiver 3 of the third embodiment, the AFC unit 70 that inputs and processes a signal having a smaller number of systems than the OFDM receiver 1 of the first embodiment is provided. In addition, the overall apparatus configuration is simplified.

1 is a system configuration diagram illustrating an OFDM receiver according to Embodiment 1 of the present invention. It is a figure which shows the example of arrangement | positioning of an array antenna part. It is a figure which shows the structure of an orthogonal demodulation part. It is a figure which shows the structure of a directivity formation part. It is a figure which shows the structure of an array synthetic signal production | generation part. It is a flowchart figure which shows the calculation procedure of the weight coefficient calculation part for array composition. It is a figure which shows the structure of an AFC part. It is a figure which shows the structure of a signal selection switching part. It is a figure which shows the structure of a diversity synthetic | combination part. It is a figure explaining the general transmission-path estimation system and the transmission-path estimation system for high-speed mobile reception. It is a figure explaining the symbol filter process for a switching countermeasure. It is a system block diagram which shows the OFDM receiver of Example 2 by this invention. It is a system block diagram which shows the OFDM receiver of Example 3 by this invention.

Explanation of symbols

1, 2, 3 OFDM receiver 10 Array antenna unit 20 Orthogonal demodulation unit 21 BPF unit 22 Local signal generation unit 23-1, 23-2 K distribution unit 24-1, 24-2 Multiplication unit 25-1, 25-2 LPF unit 26 π / 2 phase shifter 30 Directivity formation unit 31 M sample buffer unit 32 (N + 1) distribution unit 33 Array synthesis weight coefficient calculation unit 34 Delay adjustment unit 35 Array synthesis signal generation unit 36 Multiplication unit 37 Addition unit 40 , 70 AFC unit 41-1 and 41-2 Effective symbol length delay unit 42-1 and 42-2 Multiplying unit 43-1 and 43-2 Moving average unit 44 Doppler frequency calculation unit 45 Frequency correction unit 50 Signal selection switching unit 51 M sample level measurement unit 52 level determination unit 53 selection switching unit 54 input port 55 output port 56 delay unit 60 diversity combining unit 61 F T unit 62-1 SP extraction unit 62-2 SP generation unit 62-3 Division unit 63 Symbol filter unit 64 Carrier filter unit 65 Diversity combining weight coefficient calculation unit 66 Delay adjustment unit 67 Multiplication unit 68 In-phase synthesis unit 69 Symbol filter designation Part

Claims (7)

In a digital transmission receiver using an OFDM modulation scheme,
An array antenna unit configured with a plurality of K antennas as each array element;
An orthogonal demodulation unit that orthogonally demodulates OFDM signals for K elements received via the array antenna unit to IQ signals;
A plurality of N directional beams are formed based on the IQ signal, and an array combined signal is generated from the array combining weight coefficients corresponding to the N directional beams and the IQ signal distributed to the N systems. A directivity forming unit that generates N systems of signals separated from incoming waves to the array antenna unit;
An AFC unit that compensates a frequency shift for the generated N systems of signals;
A signal selection switching unit that measures each signal level of the compensated N system signals and selectively switches the upper L system signals having a higher signal level;
A diversity combining unit that combines the signals of the L system with a maximum ratio for each subcarrier;
The directivity forming part is
A sample buffer unit that accumulates data for a preset number of samples for the orthogonally demodulated IQ signal;
A distribution unit for distributing the IQ signal read from the sample buffer unit;
An array for calculating weight coefficients for array synthesis of N systems based on a correlation matrix composed of correlation values of signals for K elements in the distributed IQ signal and a vector having arrival angle information corresponding to the N systems. A weighting factor calculation unit for synthesis;
A delay adjustment unit that delays the distributed IQ signals of the N systems by a processing time in the array combination weight coefficient calculation unit;
An OFDM receiving apparatus comprising: an array combined signal generating unit configured to generate an array combined signal from the delayed N IQ signals and the N array combining weight coefficients. The OFDM receiver according to claim 1, wherein
The array synthesis weight coefficient calculation unit of the directivity forming unit is:
Based on an MSN technique (Maximal SNR Method), N systems based on a correlation matrix composed of correlation values of signals for K elements in the distributed IQ signal and steering vectors having arrival angle information corresponding to the N systems An OFDM receiver characterized by calculating a weighting factor for array synthesis. The OFDM receiver according to claim 1, wherein
The array synthesis weight coefficient calculation unit of the directivity forming unit is:
Based on a DCMP technique (Directly Constrained Minimization of Power Method) based on a correlation matrix composed of correlation values of signals for K elements in the distributed IQ signal and a constraint vector having arrival angle information corresponding to the N systems. An OFDM receiver characterized by calculating weight coefficients for N array combining. In the OFDM receiver according to any one of claims 1 to 3,
The signal selection switching unit is
A level measuring unit that measures each signal level of the N system signals compensated by the AFC unit and outputs a system number and a level value;
Based on the output system number and the level value, a level determination unit that determines the system number of the upper L system having a high level value among the system numbers of the N systems,
A selection switching unit for selectively switching an L system signal among the N system signals based on the determined system number of the L system, and outputting the L system signal after being delayed by a preset time; An OFDM receiving apparatus comprising: The OFDM receiver according to claim 4,
The diversity combining unit
A symbol filter unit for inputting a transmission path characteristic calculated from a frequency domain signal in the L system signal and performing symbol filter processing;
In order to specify the symbol filter processing based on the system number of the L system from the level determination unit of the signal selection switching unit and based on the system number of the L system and the delay time in the selection switching unit of the signal selection switching unit A symbol filter designating section that outputs the signal of
The symbol filter unit includes:
Based on the signal for specifying the symbol filter processing from the symbol filter designating unit,
When signals of different systems arrive at the diversity combining unit with a delay of 3 symbols or more with respect to the transmission path estimation target symbol due to the selection switching of the signal selection switching unit, transmission at each SP position separated by 4 symbols in the time direction. Performing a first interpolation process for estimating the transmission path characteristics by linear interpolation using the path characteristics;
A transmission path obtained by linear interpolation using transmission path characteristics at each SP position 4 symbols away in the time direction when the signals of the different systems arrive with a delay of 2 symbols with respect to the transmission path estimation target symbol. A second interpolation process is performed to estimate the transmission line characteristic from the characteristic, the transmission line characteristic of each SP position separated by one symbol in the time direction, and the transmission line characteristic of the SP position;
When the signal of the different system arrives one symbol later than the target symbol for transmission path estimation, the transmission path characteristics at each SP position separated by one symbol in the time direction and the transmission path characteristics at the original SP position are deleted. And performing a third interpolation process for estimating the transmission line characteristic from the transmission line characteristic set to 0,
An OFDM receiving apparatus characterized by performing a fourth interpolation process that does not perform the symbol filter process when the signal of the different system is only a target symbol for channel estimation. In a digital transmission receiver using an OFDM modulation scheme,
An array antenna unit configured with a plurality of K antennas as each array element;
An orthogonal demodulation unit that orthogonally demodulates OFDM signals for K elements received via the array antenna unit to IQ signals;
A plurality of N directional beams are formed based on the IQ signal, and an array combined signal is generated from an array combining weight coefficient by the N directional beams and the IQ signal distributed to the N systems, A directivity forming unit that generates N systems of signals separated from incoming waves to the array antenna unit;
An AFC unit that compensates a frequency shift for the generated N systems of signals;
A diversity combining unit that combines the compensated N-system signals with a maximum ratio for each subcarrier;
The directivity forming part is
A sample buffer unit that accumulates data for a preset number of samples for the orthogonally demodulated IQ signal;
A distribution unit for distributing the IQ signal read from the sample buffer unit;
An array for calculating weight coefficients for array synthesis of N systems based on a correlation matrix composed of correlation values of signals for K elements in the distributed IQ signal and a vector having arrival angle information corresponding to the N systems. A weighting factor calculation unit for synthesis;
A delay adjustment unit that delays the distributed IQ signals of the N systems by a processing time in the array combination weight coefficient calculation unit;
An OFDM receiving apparatus comprising: an array combined signal generating unit configured to generate an array combined signal from the delayed N IQ signals and the N array combining weight coefficients. In the OFDM receiver according to any one of claims 1 to 5,
The signal selection switching unit measures each signal level of the N system signals generated by the directivity forming unit, and selectively switches the upper L system signal having a higher signal level,
The AFC unit compensates the frequency shift for the L system signal that has been selectively switched,
2. The OFDM receiver according to claim 1, wherein the diversity combiner combines the compensated L-system signals with a maximum ratio for each subcarrier.

Images