JP2000188568A - Receiver - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow a receiver adopting a time division transmission system to conduct adaptive array diversity reception with simple signal processing at a high convergence speed of synthesis weight. SOLUTION: A reception circuit 2 converts a signal received by each antenna (branch) 1 into a base band signal and a complex multiplier section 100 and a complex adder section 15 synthesizes the base band signals. A weight calculation section 30 calculates the weight of the complex multiplier section 100. In a weight update algorithm of the weight calculation section 30, the step size in the LMS is variable. That is, the initial value is set to a large value and the set value is decreased with the lapse of time in each time slot and the value is initialized every time a time slot is changed. Thus, the convergence speed of updating the weight is increased at an arithmetic quantity nearly equal to that for a conventional LMS.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a receiving apparatus used for receiving a digitally modulated radio frequency signal and receiving the signal by a plurality of antennas.

[0002]

2. Description of the Related Art In recent years, in the field of mobile communication, improved secrecy and ISDN (Integrated Service) have been developed.
(es Digital Network) From the viewpoint of affinity with a network or a computer, effective use of frequency resources, and the like, digitization of wireless communication is progressing. In order to use frequency resources effectively, it is desirable to reuse radio waves of the same frequency (channel) at a repetition distance as short as possible. However, shortening the repetition distance of frequency increases interference (co-channel interference) from neighboring mobile stations or base stations using the same channel, and thus causes a problem of deterioration in transmission quality.

[0003] By the way, fading occurs in mobile communication, so that transmission quality (error rate in digital communication) deteriorates remarkably. For this reason, transmission quality degradation due to fading is compensated for by spatial diversity reception, which is usually received by two or more antennas and a receiving circuit (branch).

[0004] As a branch combining method, that is, a method of combining signals output from a plurality of receiving circuits into one, a post-detection selection in which an output of a branch having the highest received signal strength (hereinafter referred to as RSSI) is set as a received output. Synthesis is most common. As a combining method for further improving the reception characteristics, a maximum ratio combining method after detection is known. In general, when the maximum ratio combining is performed, a baseband signal obtained by a demodulation circuit for each branch is obtained by weighting and adding the same weight for each of two components of quadrature and in-phase.

[0005] It is known that the diversity reception improves transmission quality degradation not only for fading but also for co-channel interference. However, as a method for realizing more effective co-channel interference characteristics, "adaptive diversity" is used. An adaptive array diversity receiver called "least squares combining diversity", "LMS adaptive array" or the like has been proposed. Specifically, Japanese Patent Laid-Open No. 7-1
54129 and JP-A-9-820400 disclose the configuration of the receiver. By performing such adaptive array diversity reception, co-channel interference is reduced and frequency use efficiency can be increased.

[0006] In this adaptive array diversity receiver, a receiving circuit for converting a received signal obtained by two or more antennas into a complex baseband signal is provided for each branch. Then, the weight calculation unit calculates, for each complex baseband signal, a complex weight for weighting in the amplitude direction or the phase direction so that the signals are strengthened when synthesized. Complex weight Wk obtained for each branch
Is multiplied by a complex baseband signal Xk (k = 1, 2,..., K) obtained from the receiving circuit for each branch,
Further, by adding the complex baseband signals weighted for each branch, a combined baseband signal with fading canceled can be obtained. The demodulated data is extracted by comparing the synthesized baseband signal with an appropriate threshold value.

More specifically, the baseband signals (X1, X2,...) Received by the individual antennas and the receiving circuits (branches) and obtained for each branch have phase shifts, and are different from each other. Weight (W1, W2, ...
..), the phases of the respective baseband signals can be aligned. By adding the respective baseband signals after adjusting the phases in this way, it is possible to increase the level of the desired signal while suppressing the noise level. Here, the weights (W1, W2,...) Are used to reduce the error between the reference signal and the combined signal or to make the absolute value of the combined complex signal constant, .. And the error E are adaptively and sequentially updated.

As a weight updating algorithm, a least mean square (Le) similar to a linear equalizer using a tapped delay line is used.
An ast Mean Square (hereinafter simply referred to as LMS) or a recursive Least Mean Square (hereinafter simply referred to as RLS) algorithm is used. Specifically, for example, the document “Modulation / demodulation technology for digital radio” (Yoichi Saito, edited by the Institute of Electronics, Information and Communication Engineers) or the document “Simon
Haykin, Adaptive Filter Theory, 3rd edition, Prent
ice Hall, Upper Saddle River, NJ, 1996. Among them, LM whose weight is updated by (Equation 1)
The S algorithm is often used because of its simplest operation.

[0009]

(Equation 1)

Here, n is the time in units of the number of symbols from the beginning of the training sequence (training signal), and μ is the step size. For example, International Patent Publication W097 /
20400 discloses a diversity receiver in which the above LMS is applied to spread spectrum communication. If μ is increased, the convergence becomes faster, but the characteristics after convergence deteriorate,
It is also known to diverge. Conversely, if μ is reduced, the characteristics after convergence are improved, but convergence is slowed. For this reason, μ is usually made small at the expense of the convergence speed to some extent.

[0011]

However, in the above conventional receiving apparatus, LM is used as a weight updating algorithm.
When S is used, there is a problem that the convergence speed of the weight is slow. If the RLS algorithm is used, the convergence is fast but the calculation becomes extremely complicated. Therefore, a high-speed arithmetic circuit and a large-scale complicated arithmetic circuit are required. There was a problem.

[0012] An object of the present invention is to solve the above-mentioned problems and to provide an LM
It is an object of the present invention to provide an adaptive array diversity receiving apparatus that uses a simple algorithm having almost the same calculation amount as S and that converges weights at high speed.

[0013]

SUMMARY OF THE INVENTION The present invention relates to a receiving apparatus in a time division transmission system for transmitting data by dividing data into blocks called time slots in a time division manner. Is a function that becomes smaller as time elapses in each time slot, and is initialized each time the time slot changes. As a result, an adaptive array diversity receiving apparatus that uses a simple algorithm whose calculation amount is almost the same as that of the LMS and in which the weights converge at high speed can be realized.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The invention according to claim 1 of the present invention is provided separately for each antenna, detecting means for detecting a signal received by the antenna and outputting a baseband signal; Combining means for multiplying the band signal by a complex weight and adding and combining the signals; determining means for determining a transmission symbol from the combined baseband signal added and combined by the combining means; and a known symbol to be sent as a reference signal A reference signal generating means for generating, an error detecting means for outputting an error signal between the synthesized baseband signal and the reference signal, a weight corresponding to each antenna is calculated from the error signal and the baseband signal, and the weight is sequentially calculated. And a weight calculating means for updating the weight in the weight calculating means. The product of the baseband signal, the step size function and reduced function in accordance with the lapse of time in each time slot, a receiving device configured to initialize each time the time slot changes. With this configuration, the weight can be quickly converged with a small amount of calculation and a simple algorithm.

According to a third aspect of the present invention, there is provided a quadrature detector that performs quadrature detection on a signal received by each antenna and outputs a baseband signal of an in-phase component and a quadrature component, and an in-phase detector of the baseband signal. Combining means for multiplying the baseband signal by a complex weight with the real and imaginary parts of the complex number as the real and imaginary components, respectively, and determining transmission symbols from the combined baseband signal added and combined by the combining means Determining means for generating a known symbol to be sent as a reference signal, error detecting means for outputting an error signal between a synthesized baseband signal and the reference signal, and an error signal and a baseband signal. Weight calculating means for calculating a weight corresponding to each antenna from the signal, and sequentially updating the weight. The update amount of each weight in the weight calculation means is defined as the product of the error signal, the step size function, and the baseband signal corresponding to each antenna, and the absolute value of the step size function converges to a constant value with time. The step size function is initialized every time the time slot changes, and the absolute value of the initial value is larger than the converged constant value. With this configuration, the convergence speed can be increased by setting the step size of the weight coefficient convergence algorithm to be a variable gain as a step size function instead of being fixed.

According to a fourth aspect of the present invention, the weight calculation means includes a read-only memory, and the step size function is stored in the read-only memory. With this configuration, by storing the step size function in the dedicated memory, the calculation for the step size function becomes unnecessary.

According to a fifth aspect of the present invention, the step size function is a sum of a monotonically decreasing exponential function and a constant. With this configuration, when the weight update algorithm is configured by hardware, a step size function can be configured only by calculating the bit shift and the sum.

In the invention according to claim 6 of the present invention, the step size function is a linear function that monotonically decreases until an appropriate time, and has a constant value after the time. With this configuration,
The step size function operation can be simplified.

According to a ninth aspect of the present invention, it is assumed that the baseband signal is standardized by the total value of the received signal power of each antenna or the maximum value among the received power of each antenna. With this configuration, the convergence speed of the weight update is increased even in a receiver that does not have an automatic gain control (AGC) amplifier in the high frequency amplifier or the intermediate frequency amplifier.

According to a tenth aspect of the present invention, the step size function is configured by multiplying a constant inversely proportional to the total value of the received signal power of each antenna or the maximum value of the received power of each antenna. . With this configuration, the convergence speed of the weight update is further increased.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(Embodiment 1) FIG. 1 is a block diagram of a receiving apparatus according to Embodiment 1 of the present invention, FIG. 2 is an operation diagram of the step size function, and FIG. FIG. 4 is a graph showing an example of a monotonically decreasing exponential function, FIG. 5 is a circuit diagram of implementing a monotonically decreasing linear function, and FIG. 6 is a circuit diagram of a monotonically decreasing linear function. It is a graph showing an example.

In FIG. 1, reference numeral 1 denotes an antenna. A receiving circuit 2 converts a received signal obtained by the antenna 1 for each branch into a complex baseband signal. The received signal is quadrature modulated, and includes an in-phase component and a quadrature component. Here, the complex baseband signal is a signal in which the in-phase component and the quadrature component correspond to the real part and the imaginary part, respectively. In the receiving circuit 2, 21 is a high frequency circuit. 22 is a band pass filter. 23 is an automatic gain control (AGC) amplifier. 25 is a quadrature detector.

Reference numeral 30 denotes a weight calculator for calculating the complex weight of each branch. 100 is the weight calculator 3
This is a complex multiplication unit that multiplies the complex weight generated for each branch by 0 and the complex baseband signal output from each receiving circuit 2. Reference numeral 15 denotes a complex adder that adds a complex baseband signal output from the receiving circuit 2 of each branch after passing through the complex multiplier 100. Reference numeral 16 denotes a determination unit that compares the combined complex baseband signal output from the complex addition unit 15 with an appropriate threshold value and determines transmission data.

Reference numeral 40 denotes a training signal generator for storing a known symbol called a training sequence. The transmitting side periodically inserts a known symbol (training sequence) into a part of the transmission data, and the training signal generator 40 outputs the same data as the known symbol (training sequence). Reference numeral 19 denotes a complex determination error detection unit that outputs a difference between the reference signal D, which is a complex baseband signal, and the complex baseband signal synthesized by the complex addition unit 15 as an error signal E. Here, the reference signal D may be a signal corresponding to the demodulated data determined by the determination unit 16 or a signal corresponding to a known symbol obtained from the training signal generation unit 40.

The operation will be described below. First, a signal is received by the antenna 1, and the received signal is converted into a high frequency circuit 21, a band pass filter 22, an automatic gain control (AGC) amplifier 2
3. Pass through the quadrature detector 25 to obtain a complex baseband signal in which the in-phase and quadrature components of the received signal correspond to the real and imaginary parts, respectively. Receiving circuit 2 for each branch
From the complex baseband signal Xk (k = 1,
2,. . . K) is weighted by the complex multiplier 100 with the complex weight Wk from the weight calculator 30.

The complex baseband signals weighted by the complex multipliers 100 of the respective branches are combined by the complex adder 15. By comparing the synthesized complex baseband signal with an appropriate threshold value in the determination unit 16, net data to be transmitted by the transmission side is determined, and the determination unit 16 outputs the data as demodulated data.

The training signal generator 40 generates the same data as the training sequence inserted in the transmission data. The complex decision error detector 19 adds a reference signal D, which is an ideal complex baseband signal corresponding to the demodulated data determined by the determiner 16 or corresponding to a known symbol obtained from the training signal generator 40, to a complex addition. The difference from the complex baseband signal synthesized by the unit 15 is obtained,
Output as an error signal E. The weight calculator 30 calculates the complex weight of each branch by the following (Equation 2).

[0029]

(Equation 2)

Here, * indicates a complex conjugate. G (n) is a step size function. As is apparent from comparison with (Equation 1), when G (n) is a fixed constant μ, the equation becomes the same as that of LMS. Therefore, the amount of calculation is (Equation 1), that is, the same as the conventional LMS algorithm.

Next, the step size function G (n) will be described. In the case of digital communication using the TDMA method,
Regardless of the access method, the transmission data is often divided in time into blocks called time slots and transmitted. FIG. 2 shows an example in which a user (partner station) 1 and a user (partner station) 2 communicate using slot 1 and slot 3, respectively. In general, a known symbol called a pilot signal (training signal) or a preamble is inserted at the beginning of a time slot. Note that the slot 2 is not used in the example of FIG.

As shown in FIG. 2, a step size function G (n) is initialized (n = 0) at the beginning of each slot, and G
Start from (0). G (n) is a function that asymptotically or coincides with a small value enough to obtain the characteristics after convergence in a normal LMS with an increase in n, and decreases during one slot as shown in FIG. Asymptotically or coincides with μ. The initial value G (0) of the function is generally G (0)> μ in order to speed up the convergence in the initial stage. The expression of G (n) in (Equation 2) is an example.

In mobile communication, propagation characteristics often fluctuate due to fading, and in the TDMA system, since the partner station differs in each slot, the propagation environment differs for each slot. As described above, the step size function G (n) is initialized (n = 0) at the head of each slot and starts from G (0), so that even if the partner station differs in each slot, it is possible to cope with it. .

When continuous slots are used for the same partner station and the fading fluctuation is small, the step size function for each slot need not be initialized.

As described above, the step size in the LMS algorithm is not a constant but a function that changes with time, so that the convergence in the initial stage can be accelerated and the characteristics after convergence can be set to desired values. Therefore, the normal L
The convergence speed can be increased with the same amount of computation as MS.

The step size function G of (Equation 2)
(N) may be configured not to directly calculate this, but to store the result of the calculation in advance in a memory and read it from the memory when necessary. In this case, the amount of calculation can be reduced.

Further, instead of the function of (Equation 2), (Equation 3) represented by the sum of a monotonically decreasing exponential function and a constant may be used as the step size function.

[0038]

(Equation 3)

When this function is generated by hardware,
Since the operation of G (n) can be performed by bit shift and sum, FIG.
Can be realized by the configuration shown in FIG. In FIG. 3, reference numeral 50 denotes a right 1-bit shift circuit. 51 is a right 2-bit shift circuit. 52 is a register for delaying one symbol. 53 and 54 are adders.

First, an initial value is input to the register 52. Next, the data is output from the register to the adder 53, added with a constant B, and output as a step size function G (n). Also,
The output of the register 52 is input to the right 1-bit shift circuit 50 and the right 2-bit shift circuit 51, respectively, multiplied by 0.5 and 0.25 by each circuit, added by the adder 54, and input to the register 52. You. FIG. 4 shows a graph of G (n) in this case. Thereby, G (n) represented by the sum of the monotonically decreasing exponential function and the constant can be realized. In this case, G (n) can be generated by a simple circuit as described above, instead of a memory that stores G (n) in advance, so that the circuit scale can be reduced.

Alternatively, the step size function G (n) is represented by a monotonically decreasing linear function (Equation 4).
May be used as a step size function.

[0042]

(Equation 4)

FIG. 5 shows a configuration in which this function is generated by hardware. 52 is a register for delaying one symbol. In FIG. 5, reference numeral 53 denotes an adder. 55 is a subtractor.

First, an initial value of the register 52 is set. Next, the output from the register 52 to the adder 53 is added with a constant b, and output as a step size function. The output of the register 52 is added with a constant a by a subtractor 55 and is input to the register 52. The graph of G (n) in this case is shown in FIG.
Shown in Also in this case, G (n) can be generated by a simple circuit as described above instead of a memory that stores G (n) in advance, so that the circuit scale can be reduced.

Further, the second term of (Equation 2), that is, the complex baseband signal is normalized by the sum of the received signal power of each antenna or the maximum value among the received power of each antenna (Equation 5) or (Equation 6) may be used as the weight update algorithm.

[0046]

(Equation 5)

[0047]

(Equation 6)

In this case, assuming that the step size function is a fixed constant in (Equation 5), NLMS (Normali)
(Zed LMS) or a learning and identification method. NLMS is known as an algorithm for accelerating LMS, and the present invention uses NLMS as described above.
It is easily applicable to MS. This allows the weight update algorithm to converge at a higher speed.

Further, the step size function G (n) in the second term of (Equation 2) is calculated as the initial value of the total value of the received signal power of each antenna or the maximum value of the received power of each antenna (at the time of starting communication) Alternatively, (Expression 7) or (Expression 8) standardized using the value at the beginning of the slot) may be used as the weight update algorithm.

[0050]

(Equation 7)

[0051]

(Equation 8)

As a result, the weight updating algorithm can converge at a higher speed, and the amount of calculation can be reduced.

As a step size function G (n),
A plurality of linear functions may be used as a composite step size function. FIG. 7 shows an example of G (n) in this case.
In FIG. 7, the step size function is represented by an initial section (n ≦ k
In (1), the first linear function decreases monotonically, and in the intermediate section (k1 <n ≦ k2), the second linear function has a larger change rate than the first linear function. Also (n> k2)
Then, another linear function or a constant value may be used.

[0054]

According to the present invention, a step size function G (n) is used for the LMS algorithm in adaptive array diversity reception, thereby realizing a high convergence speed comparable to that of RLS with the same amount of calculation as that of a normal LMS. it can.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a receiving device according to a first embodiment of the present invention.

FIG. 2 is an operation diagram of a step size function according to the first embodiment of the present invention;

FIG. 3 is a circuit diagram for realizing a sum of a monotonically decreasing exponential function and a constant according to the first embodiment of the present invention;

FIG. 4 is a graph showing an example of a monotonically decreasing exponential function according to the first embodiment of the present invention;

FIG. 5 is a circuit diagram for realizing a monotonically decreasing linear function according to the first embodiment of the present invention;

FIG. 6 is a graph showing an example of a monotonically decreasing linear function according to the first embodiment of the present invention;

FIG. 7 is a graph showing an example in which a plurality of linear functions are combined according to the second embodiment of the present invention;

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 antenna 2 reception circuit 15 complex addition unit 16 determination unit 19 complex determination error detection unit 21 high-frequency circuit 22 band-pass filter 23 automatic gain control (AGC) amplifier 25 quadrature detector 30 weight calculation unit 40 training signal generation unit 100 complex multiplication unit

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04L 1/06 H04B 7/26 K

Claims (10)

[Claims] 1. A receiving apparatus in a time division transmission system in which data is time-divided into blocks called time slots and transmitted, wherein the receiving apparatus is provided for each of a plurality of antennas and received by each antenna. Detecting means for detecting a signal and outputting a baseband signal; synthesizing means for multiplying and multiplying the baseband signal by a complex number weight; and determining a transmission symbol from the synthesized baseband signal obtained by adding and synthesizing by the synthesizing means. Determining means for generating, a reference signal generating means for generating a known symbol to be transmitted as a reference signal, an error detecting means for outputting an error signal between the synthesized baseband signal and the reference signal, and the error signal A weight corresponding to each antenna is calculated from the baseband signal, and the weight is sequentially updated. And a calculation unit, the update amount of the weights in the weight calculation means,
The error signal, the step size function, and the product of each baseband signal, the step size function is a function that becomes smaller as time elapses in each time slot, and is initialized each time the time slot changes. Characteristic receiving device. 2. A receiving apparatus in a time division transmission system in which data is time-divided into blocks called time slots and transmitted, comprising: a plurality of antennas; a signal provided for each antenna; Detecting means for detecting the baseband signal and outputting a baseband signal; combining means for multiplying the baseband signal by a complex number weight; and adding and combining the baseband signal; and determining a transmission symbol from the combined baseband signal added and combined by the combining means. Determining means; reference signal generating means for generating a known symbol to be sent as a reference signal; error detecting means for outputting an error signal between the synthesized baseband signal and the reference signal; A weight corresponding to each of the antennas is calculated from a baseband signal, and the weight is sequentially updated. Calculation means, and the update amount of each weight in the weight calculation means is a product of the error signal, the step size function, and the baseband signals corresponding to the antennas, and the step size function Is a receiver which converges to a constant value over time in each time slot and initializes each time the time slot changes. 3. A receiving apparatus in a time division transmission system in which data is time-divided into blocks called time slots and transmitted, wherein a plurality of antennas, a signal provided for each antenna, and a signal received by the antenna are provided. A quadrature detector that performs quadrature detection and outputs a baseband signal of an in-phase component and a quadrature component, and the in-phase component and the quadrature component of the baseband signal as a real part and an imaginary part of a complex number, respectively. Synthesizing means for multiplying and adding by weight, determining means for determining a transmission symbol from the synthesized baseband signal added and synthesized by the synthesizing means, and generating a reference signal for generating a known symbol to be transmitted as a reference signal Means, error detection means for outputting an error signal between the synthesized baseband signal and the reference signal, Weight calculating means for calculating weights corresponding to the respective antennas from the error signal and the baseband signal, and sequentially updating the weights, wherein an update amount of the respective weights in the weight calculating means is calculated. , The product of the error signal, the step size function, and the respective baseband signals corresponding to the respective antennas, the absolute value of the step size function converges to a constant value with time, and the time slot changes in the step size function Each time, and the absolute value of the initial value is larger than the converged constant value. 4. The apparatus according to claim 1, wherein said weight calculating means includes a read only memory, and said step size function is stored in said read only memory.
The receiving device according to the above. 5. The receiving apparatus according to claim 1, wherein said step size function is a sum of a monotonically decreasing exponential function and a constant. 6. The receiving apparatus according to claim 1, wherein the step size function is a linear function that monotonically decreases until an appropriate time, and has a constant value after the time. 7. The step size function has a first linear function monotonically decreasing until a first time and a second linear function having a larger change rate than the first linear function until the next two times. The receiving device according to claim 1, wherein the receiving device comprises: 8. A receiving apparatus in a time division transmission system for transmitting data by dividing data into blocks called time slots in a time-division manner, comprising: a plurality of antennas; An intermediate frequency signal obtained by frequency-converting a signal or a high-frequency signal is supplied, and the high-frequency signal or the intermediate-frequency signal is quadrature-detected by a local signal having a frequency substantially equal to a nominal carrier frequency of the high-frequency signal or the intermediate frequency signal. A quadrature detector that outputs a baseband signal of a quadrature component and an in-phase component and a quadrature component of the baseband signal as a real part and an imaginary part of a complex number, respectively, and multiplies the baseband signal by a complex weight to perform addition and synthesis. Combining means, and a combined baseband added and combined by the combining means Determining means for determining a transmission symbol from a signal; reference signal generating means for generating, as a reference signal, a known signal portion included in a received signal and a baseband signal corresponding to the determination result of the determining means; Error detection means for outputting an error signal between the signal and the reference signal, and weight calculation means for calculating a weight corresponding to each of the antennas from the error signal and the baseband signal, and sequentially updating the weights The weight update means has an update amount of each weight as the product of the error signal, the step size function, and the respective baseband signals corresponding to the respective antennas, and the absolute value of the step size function is Converging to a constant value over time, the step size function is initialized each time the time slot changes, The absolute value of the period value receiving apparatus, characterized in that larger than a predetermined value the convergence. 9. The reception according to claim 1, wherein the baseband signal is standardized by a total value of received signal power of each antenna or a maximum value among received power of each antenna. apparatus. 10. The step size function according to claim 1, wherein the step size function is obtained by multiplying a total value of received signal power of each antenna or a maximum value among the received power of each antenna by a constant inversely proportional. 9. The receiving device according to 1 or 8.