Receiving device and receiving program

Abstract

The present disclosure aims to improve error correction capability when compensating for nonlinear distortion in a modulated signal, particularly when the modulation order is low, without performing unnecessary distortion compensation processing and, on the contrary, without adding nonlinear distortion. [Solution] In the present disclosure, the equalizer 4 provided in the receiving device R controls the degree of freedom of the compensation processing for nonlinear distortion of the modulated signal to be lower the lower the modulation order of the modulated signal, and controls the degree of freedom of the compensation processing for nonlinear distortion of the modulated signal to be higher the higher the modulation order of the modulated signal.

Description

[Technical Field]

[0001] The present disclosure relates to techniques for compensating for nonlinear distortion in modulated signals. [Background technology]

[0002] In the amplifiers of transmitters or repeaters, high power utilization efficiency can be achieved by setting the operating point near the saturation region. However, a modulated signal with nonlinear distortion generates radiation outside the transmission band and degrades the error rate of the received bit.

[0003] Therefore, techniques for compensating for nonlinear distortion of modulated signals are disclosed in Non-Patent Documents 1 and 2 and Patent Document 1. In Non-Patent Document 1, an equalizer that uses a Volterra series is used to compensate for nonlinear distortion of modulated signals. In Non-Patent Document 2, in order to reduce the amount of calculation, an equalizer that approximates the Volterra series with a memory polynomial is used to compensate for nonlinear distortion of modulated signals.

[0004] In Patent Document 1, in order to reduce the amount of calculation, an equalizer that approximates a Volterra series with a memory polynomial is used to compensate for nonlinear distortion of a modulated signal, and the filter coefficients of the equalizer are calculated using known signals with a number of symbols that is smaller than the number of taps of the equalizer. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] International Publication No. 2019 / 171655 [Non-patent literature]

[0006] [Non-Patent Document 1] S. Benedetto and E. Biglieri, “Nonlinear Equalization of Digital Satellite Channels,” IEEE Journal on Selected Areas in Communications, vol. 1, No. 1, pp. 57-62, Jan 1983. [Non-patent document 2] Y. Noda, S. Uehashi, S. Tani, K. Motoyoshi, and A. Okamura, "An adaptive equalization method for single-carrier wideband transmission with nonlinear distortion," IEICE Technical Report, WBS2018-2, vol. 118, No. 51, pp. 7-12, May 2018. [Non-patent document 3] “ETSI EN 302 307-1 V1.4.1 (2014-11)”, [online], EUROPEAN STANDARD, [Retrieved December 4, 2024], Internet <URL:https: / / www.etsi.org / deliver / etsi_en / 302300_302399 / 30230701 / 01.04.01_60 / en_30230701v010401p.pdf> Summary of the Invention [Problem to be solved by the invention]

[0007] By the way, the modulation method and the error correction code rate can be adaptively controlled using the ACM (Adaptive Coding and Modulation) method according to the transmission path condition (CNR: Carrier-to-Noise Ratio). For example, a Low-Density Parity-Check (LDPC) code, which has high resistance to AWGN (Additive White Gaussian Noise), can be applied as the error correction code. Here, in order to calculate a Log-Likelihood Ratio (LLR) or the like in the stage preceding the decoder, it is desirable to compensate for nonlinear distortion of the modulated signal in the stage preceding the likelihood ratio calculation unit.

[0008] However, in Non-Patent Documents 1 and 2 and Patent Document 1, the degree of freedom in compensating for nonlinear distortion of modulated signals is set equal for all modulation methods. Here, when the back-off during amplification is constant, the lower / higher the modulation order, the smaller / larger the nonlinear distortion, making distortion compensation unnecessary / necessary. Therefore, if distortion compensation is performed unnecessarily, particularly when the modulation order is low, nonlinear distortion will be added and the error correction capability will deteriorate.

[0009] Therefore, in order to solve the above problem, the present disclosure aims to improve error correction capability when compensating for nonlinear distortion in a modulated signal, particularly when the modulation order is low, without performing unnecessary distortion compensation processing and, instead, without adding nonlinear distortion. [Means for solving the problem]

[0010] In order to solve the above problem, the equalizer provided in the receiving device controls the degree of freedom of the compensation processing for nonlinear distortion of the modulated signal to be lower the lower the modulation order of the modulated signal, and controls the degree of freedom of the compensation processing for nonlinear distortion of the modulated signal to be higher the higher the modulation order of the modulated signal.

[0011] Specifically, the present disclosure is a receiving device that receives a modulated signal from a transmitting device that transmits the modulated signal, the receiving device comprising: an equalizer that compensates for nonlinear distortion in the modulated signal; a demodulator that demodulates the modulated signal with the nonlinear distortion compensated for; and an error corrector that performs error correction on the demodulated signal, wherein the equalizer controls the degree of freedom of the compensation process for the nonlinear distortion of the modulated signal to be lower the lower the modulation order of the modulated signal, and controls the degree of freedom of the compensation process for the nonlinear distortion of the modulated signal to be higher the higher the modulation order of the modulated signal, the receiving device that is applied to wireless or wired communication.

[0012] According to this configuration, by not performing unnecessary distortion compensation processing, particularly when the modulation order is low, it is possible to improve the error correction capability without adding nonlinear distortion, and by performing sufficient distortion compensation processing, particularly when the modulation order is high, it is possible to improve the error correction capability after sufficiently removing nonlinear distortion.

[0013] The present disclosure also provides a receiving device characterized in that the equalizer controls the degree of freedom of the compensation processing for nonlinear distortion of the modulated signal to be lower the higher the coding rate of error correction of the modulated signal, and controls the degree of freedom of the compensation processing for nonlinear distortion of the modulated signal to be higher the lower the coding rate of error correction of the modulated signal.

[0014] According to this configuration, by not performing unnecessary distortion compensation processing, particularly when the error correction coding rate is high, it is possible to improve the error correction capability without adding nonlinear distortion. On the other hand, by performing sufficient distortion compensation processing, particularly when the error correction coding rate is low, it is possible to improve the error correction capability after sufficiently removing nonlinear distortion.

[0015] The present disclosure also provides a receiving device characterized in that the equalizer controls the degree of freedom of the compensation processing for nonlinear distortion of the modulated signal to be lower the smaller the back-off when the modulated signal is amplified, and controls the degree of freedom of the compensation processing for nonlinear distortion of the modulated signal to be higher the larger the back-off when the modulated signal is amplified.

[0016] According to this configuration, by not performing unnecessary distortion compensation processing, particularly when the back-off during amplification is small, it is possible to improve the error correction capability without adding nonlinear distortion.Furthermore, by performing sufficient distortion compensation processing, particularly when the back-off during amplification is large, it is possible to sufficiently remove nonlinear distortion and improve the error correction capability.

[0017] The present disclosure also provides a receiving device characterized in that the equalizer adjusts the weighting coefficients of the first and third or higher order equalization filters so as to minimize an evaluation function including the squared error between the modulated signal in which nonlinear distortion has been compensated and the demodulated signal or known signal, and regularization terms of the weighting coefficients of the first and third or higher order equalization filters, and controls the degree of freedom of the compensation process for the nonlinear distortion of the modulated signal to be lower by setting a larger regularization parameter of the regularization term of the weighting coefficient of the equalization filter, and controls the degree of freedom of the compensation process for the nonlinear distortion of the modulated signal to be higher by setting a smaller regularization parameter of the regularization term of the weighting coefficient of the equalization filter.

[0018] According to this configuration, by including regularization terms for the weighting coefficients of the first-order and third-order or higher equalization filters in the evaluation function, it is possible to control the degree of freedom in the compensation process for nonlinear distortion of the modulated signal.

[0019] The present disclosure also provides a receiving program installed in a computer to cause the equalizer included in the receiving device described above to control the degree of freedom of compensation processing for nonlinear distortion of the modulated signal.

[0020] According to this configuration, it is possible to provide a program having the above-described effects.

[0021] The above-disclosed inventions can be combined as much as possible. [Effects of the Invention]

[0022] In this way, the present disclosure can improve error correction capability when compensating for nonlinear distortion in a modulated signal, particularly when the modulation order is low, without performing unnecessary distortion compensation processing and, in fact, without adding nonlinear distortion. [Brief explanation of the drawings]

[0023] [Figure 1] FIG. 1 is a diagram illustrating a configuration of a transmission / reception system according to the present disclosure. [Figure 2]1 is a diagram showing a procedure of a transmission process and a reception process according to the present disclosure. [Figure 3] 10A and 10B are diagrams illustrating procedures of pre-distortion compensation processing and processing during distortion compensation according to the present disclosure. [Figure 4] FIG. 10 illustrates the control of a regularization parameter according to a modulation order of the present disclosure. [Figure 5] FIG. 10 illustrates the control of regularization parameters as a function of backoff according to the present disclosure. [Figure 6] FIG. 10 illustrates control of a regularization parameter according to a coding rate according to the present disclosure. [Figure 7] 3A and 3B are diagrams illustrating amplitudes and coding rates of modulated signals according to the present disclosure. [Figure 8] 3A and 3B are diagrams illustrating amplitudes and coding rates of modulated signals according to the present disclosure. [Figure 9] FIG. 1 illustrates a table of regularization parameters of the present disclosure. [Figure 10] FIG. 10 illustrates an interpolation curve for the regularization parameter of the present disclosure. [Figure 11] 1A and 1B are diagrams illustrating distortion compensation of a modulated signal according to the prior art and the present disclosure. [Figure 12] 1A and 1B are diagrams illustrating distortion compensation of a modulated signal according to the prior art and the present disclosure. [Figure 13] 1A and 1B are diagrams illustrating distortion compensation of a modulated signal according to the prior art and the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0024]

[0023] The following embodiments of the present disclosure will be described with reference to the accompanying drawings. The embodiments described below are examples of implementation of the present disclosure, and the present disclosure is not limited to the following embodiments.

[0025] (Configuration of the transmission / reception system of the present disclosure) The configuration of a transmission / reception system according to the present disclosure is shown in Fig. 1. The procedures of the transmission process and reception process according to the present disclosure are shown in Fig. 2. The transmission / reception system S is applied to wireless or wired communication and includes a transmission device T and a reception device R. The transmission device T includes an encoder 1, a modulator 2, and an amplifier 3. The reception device R includes an equalizer 4, a demodulator 5, an error corrector 6, a switch 7, and a regularization parameter control unit 8. In particular, to cause the regularization parameter control unit 8 to execute step S4, a reception program for step S4 can be installed in a computer.

[0026] When transmitting a modulated signal, the transmitting device T writes the modulation method, the backoff during amplification, and the error correction coding rate into the header of the frame or the like (step S1). The encoder 1 encodes the transmission signal. The modulator 2 generates the encoded modulated signal. The amplifier 3 amplifies the encoded modulated signal (step S2).

[0027] When receiving a modulated signal, the receiving device R reads the modulation method, back-off during amplification, and error correction coding rate from the frame header or the like (step S3). The equalizer 4 compensates for nonlinear distortion in the modulated signal. The demodulator 5 demodulates the modulated signal with the nonlinear distortion compensated for. The error corrector 6 performs error correction on the demodulated signal (step S4). The switch 7 outputs a known signal such as a pilot signal as a reference signal (described later) during the acquisition process, and outputs a hard decision value of the demodulator 5 or the like during the data transmission process. The regularization parameter control unit 8 will be described later.

[0028] That is, the transmission / reception system S can adaptively control the modulation method and the error correction coding rate using the ACM method according to the transmission path condition (CNR). For example, the error corrector 6 can apply an LDPC code, which has strong resistance to AWGN, as the error correction code. Here, the error corrector 6 compensates for nonlinear distortion of the modulated signal before the likelihood ratio calculation unit in order to calculate the log-likelihood ratio (LLR) etc. before the decoder.

[0029] In the prior art, the degree of freedom of the compensation process for nonlinear distortion of the modulated signal is equalized regardless of the modulation method of the modulated signal, the back-off during amplification, and the coding rate for error correction. In the present disclosure, the regularization parameter control unit 8 controls the degree of freedom of the compensation process for nonlinear distortion of the modulated signal according to the modulation method of the modulated signal, the back-off during amplification, and the coding rate for error correction (step S4).

[0030] As a result, by not performing unnecessary distortion compensation processing, particularly when (1) the modulation order is low, (2) the back-off during amplification is small, or (3) the error correction coding rate is high, it is possible to improve error correction capability without introducing nonlinear distortion.

[0031] On the other hand, by performing sufficient distortion compensation processing, particularly when (1) the modulation order is high, (2) the back-off during amplification is large, or (3) the error correction coding rate is low, it is possible to sufficiently remove nonlinear distortion and improve error correction capability.

[0032] (Procedure of processing during distortion compensation according to the present disclosure) The equalizer 4 includes a first-order equalization filter 41, a weighting coefficient calculation unit 42, a cube value calculation unit 43, a third-order equalization filter 44, a weighting coefficient calculation unit 45, an adder 46, and a subtractor 47. As a modification, the equalizer 4 may include a third-order or higher equalization filter.

[0033] The first-order equalization filter 41 receives the input signal y 1、n As the modulated signal and the additive noise x n +z n (See the first equation of Equation 1.) The cube value calculation unit 43 calculates the cube value |x n +z n | 2 (x n +z n The third-order equalization filter 44 calculates the input signal y 2、n The cube of the modulated signal and the additive noise |x n +z n | 2 (x n +z n) (see the second equation in Equation 1).

number

[0034] The first-order equalization filter 41 and the third-order equalization filter 44 may be implemented by a FIR (Finite Impulse Response) filter or the like. n +z n may be sampled at the symbol timing of the aperture point of the eye pattern, or may be oversampled at a sampling frequency P times the symbol timing.

[0035] As will be described later, the weighting coefficient calculation unit 42 calculates a weighting coefficient vector w1 of the primary equalization filter 41 (see the first equation of Equation 2, where 0, 1, . . . , M1-1 indicate tap numbers). The weighting coefficient calculation unit 45 calculates a weighting coefficient vector w2 of the third-order equalization filter 44 (see the second equation of Equation 2, where 0, 1, . . . , M2-1 indicate tap numbers).

number

[0036] The first-order equalization filter 41 is a filter that calculates a weighting coefficient vector w1 and an input signal vector y 1、n The dot product w1 between T y 1、n (See the first equation in Equation 3 and the first term on the right side of Equation 4.) The third-order equalization filter 44 calculates the weighting coefficient vector w2 and the input signal vector y 2、n The dot product w2 between T y 2、n (See the second equation in Equation 3 and the second term on the right side of Equation 4.) The adder 46 calculates the inner product w1 T y 1、n and the inner product w2 T y 2、n The sum x between n ^ (see Equation 4).

number

number

[0037] The demodulator 5 outputs the sum x n ^ and demodulate the hard decision value x n-m The switch 7 outputs the reference signal x n-m In the acquisition process, the known signal x such as the pilot signal n-m In the data transmission process, the demodulator 5 outputs the hard decision value x n-m The subtractor 47 outputs the sum x n ^ and the reference signal x n-m The error value e between n (see Equation 5).

number

[0038] In the present disclosure, when the first-order equalization filter 41 and the third-order equalization filter 44 are not applied with a roll-off filter (ROF), the first-order equalization filter 41 and the third-order equalization filter 44 are applied with a roll-off filter (ROF) of w 1、m and w 2、m is non-zero. Here, when m=0, the reference signal x n-m When m=1, the reference signal x n-m As a modified example, when applying the ROF, the first-order equalization filter 41 and the third-order equalization filter 44 use a value of w at a plurality of m, for example, at the center tap of the ROF. 1、m and w 2、m may be non-zero.

[0039] The regularization parameter control unit 8 sets regularization parameters λ1 and λ2 of the regularization terms of the weighting coefficient vectors w1 and w2 of the first-order equalization filter 41 and the third-order equalization filter 44 (see the second and third terms on the right-hand sides of Equations 6 and 12). In general, the regularization parameter control unit 8 sets the regularization parameters of the regularization terms of the weighting coefficient vectors of the equalization filters to be larger, thereby reducing the modulation signal x n On the other hand, the regularization parameter control unit 8 sets the regularization parameter of the regularization term of the weighting coefficient vector of the equalization filter to a smaller value, thereby reducing the degree of freedom of the compensation process for the nonlinear distortion of the modulated signal x n This increases the degree of freedom in the compensation process for nonlinear distortion (see the right-hand sides of Equations 6 and 12).

[0040] As a first specific example, the regularization parameter control unit 8 sets the regularization parameters λ1 and λ2 of the regularization terms of the weighting coefficient vectors w1 and w2 of the first-order equalization filter 41 and the third-order equalization filter 44 to be larger, thereby reducing the modulation signal x n On the other hand, the regularization parameter control unit 8 sets the regularization parameters λ1 and λ2 of the weighting coefficient vectors w1 and w2 of the first-order equalization filter 41 and the third-order equalization filter 44 to be smaller, thereby reducing the degree of freedom of the compensation process for the nonlinear distortion of the modulated signal x n This increases the degree of freedom in the compensation process for nonlinear distortion (see the second and third terms on the right-hand sides of Equations 6 and 12).

[0041] As a second specific example, the regularization parameter control unit 8 sets only the regularization parameter λ2 of the regularization term of the weighting coefficient vector w2 of the third-order equalization filter 44 to a larger value, thereby reducing the modulation signal x n On the other hand, the regularization parameter control unit 8 sets only the regularization parameter λ2 of the regularization term of the weighting coefficient vector w2 of the third-order equalization filter 44 to a smaller value, thereby reducing the degree of freedom of the compensation process for the nonlinear distortion of the modulated signal x nThis increases the degree of freedom in the compensation process for nonlinear distortion (see the third term on the right side of Equations 6 and 12). In the following, the first specific example is used, but the second specific example may also be used.

[0042] The weighting coefficient calculation unit 42 calculates the error value e n The expected value of the square of E[|e n | 2 ] and the regularization terms of the weighting coefficient vectors w1 and w2 of the first-order equalization filter 41. ^ (See the first term on the left side of Equation 6.) The weighting coefficient calculation unit 45 calculates the error value e n The expected value of the square of E[|e n | 2 ] and the regularization terms of the weighting coefficient vectors w1 and w2 of the third-order equalization filter 44. ^ (See the second term on the left side of Equation 6.) That is, the weighting coefficient calculation units 42 and 45 calculate the weighting coefficient vector w1 ^ , w2 ^ As the Wiener solution w1 ^ , w2 ^ is calculated.

number

[0043] Weighting coefficient vector w1 of the first-order equalization filter 41 ^ is expressed as in Equation 7. The weighting coefficient vector w2 of the third-order equalization filter 44 ^ is expressed as in Equation 8. Here, the modulated signal x n The nonlinear distortion of is extremely small, and the additive noise z n is assumed to be AWGN.

number

number

[0044] In Equations 7 and 8, P 11 , P 12 , P 22 , r1, and r2 are expressed as in Equations 10 and 11. Here, P x is the modulated signal power, and P z is the additive noise power (see Equation 9). And, E[|x| α ](α≧2) is the modulating signal x n is the expectation of the α-th power of E[|z| 2 ] is the additive noise z n is the expected value of the square of (see Equations 9 to 11).

number

number

number

[0045] In this way, the regularization parameters λ1 and λ2 of the regularization terms of the weighting coefficient vectors w1 and w2 of the first-order equalization filter 41 and the third-order equalization filter 44 are mutually ^ , w2 ^ The regularization parameters λ1 and λ2 may be equal to or different from each other, and while they are 0 in the prior art, they are positive in the present disclosure. As a modified example, for a third-order or higher equalization filter, the regularization parameter of the regularization term of the weight coefficient vector can also be controlled.

[0046] The weighting coefficient calculation unit 42 calculates the error value e n The squared value of |e n | 2 and the regularization terms of the weighting coefficient vectors w1 and w2 of the first-order equalization filter 41 are calculated so as to minimize an evaluation function including the weighting coefficient vectors w1 and w2. ^ (See the first term on the left side of Equation 12.) The weighting coefficient calculation unit 45 calculates the error value e n The squared value of |en | 2 and the regularization terms of the weighting coefficient vectors w1 and w2 of the third-order equalization filter 44 are calculated so as to minimize an evaluation function including the weighting coefficient vectors w1 and w2. ^ (See the second term on the left side of Equation 12.) That is, the weighting coefficient calculation units 42 and 45 may calculate the weighting coefficient vector w1 in accordance with the LMS (Least Mean Square) norm or the RLS (Recursive Least Square) norm. ^ , w2 ^ may be calculated.

number

[0047] Weighting coefficient vector w1 of the first-order equalization filter 41 ^ may be expressed as in Equation 13. The weighting coefficient vector w2 of the third-order equalization filter 44 ^ may be expressed as in Equation 14. Here, w in Equation 13 1、n ^ and w in Equation 14 2、n ^ is the estimated value at time n, and Δw in Equation 13 1、n ^ and Δw in Equation 14 2、n ^ is the update amount at time n, and μ1 in Equation 13 and μ2 in Equation 14 are parameters indicating the step size in the LMS model.

number

number

[0048] In this way, the regularization parameters λ1 and λ2 of the regularization terms of the weighting coefficient vectors w1 and w2 of the first-order equalization filter 41 and the third-order equalization filter 44 are respectively and independently adjusted by the weighting coefficient vectors w1 and w2 of the first-order equalization filter 41 and the third-order equalization filter 44. ^ , w2 ^The regularization parameters λ1 and λ2 may be equal to or different from each other, and although they are 0 in the prior art, they are positive in the present disclosure. As a modified example, for a third-order or higher equalization filter, the regularization parameter of the regularization term of the weight coefficient vector may be controlled.

[0049] (Procedure for pre-processing of distortion compensation according to the present disclosure) The procedures for pre-distortion compensation processing and during-distortion compensation processing of the present disclosure are shown in Figure 3. In pre-distortion compensation processing, prior to during-distortion compensation processing, the regularization parameters λ1 and λ2 are optimized for each modulation method of the modulated signal, back-off during amplification, and coding rate for error correction (see the left column of Figure 3). In during-distortion compensation processing, after the pre-distortion compensation processing, the regularization parameters λ1 and λ2 are controlled to optimized values ​​for each modulation method of the modulated signal, back-off during amplification, and coding rate for error correction (see the right column of Figure 3).

[0050] First, the distortion compensation preprocessing will be described. The modulator 2, amplifier 3, and encoder 1 respectively fix the modulation method of the modulated signal, the back-off during amplification, and the coding rate for error correction to various methods and values ​​(step S11). The regularization parameter control unit 8 changes the regularization parameters λ1 and λ2 to various values, and then the equalizer 4 compensates for the nonlinear distortion of the modulated signal, the demodulator 5 demodulates the modulated signal with the nonlinear distortion compensated, and the error corrector 6 performs error correction on the demodulated signal and evaluates the error rate of the demodulated signal (step S12). The regularization parameter control unit 8 optimizes the regularization parameters λ1 and λ2 for each modulation method of the modulated signal, the back-off during amplification, and the coding rate for error correction (step S13). Details will be described with reference to Figures 4 to 10.

[0051] Next, the distortion compensation process will be described. The receiving device R reads the modulation method of the modulated signal, the back-off during amplification, and the error correction coding rate from the frame header or the like (step S14). The regularization parameter control unit 8 controls the regularization parameters λ1 and λ2 to optimized values ​​for each modulation method of the modulated signal, the back-off during amplification, and the error correction coding rate. Then, the equalizer 4 compensates for the nonlinear distortion of the modulated signal, the demodulator 5 demodulates the modulated signal whose nonlinear distortion has been compensated for, and the error corrector 6 performs error correction on the demodulated signal (step S15).

[0052] Control of the regularization parameter according to the modulation order of the present disclosure is shown in Fig. 4. The amplifier 3 and the encoder 1 fix the back-off and error correction coding rate when amplifying the modulated signal to one value, and the modulator 2 changes the modulation method of the modulated signal to various methods (step S11).

[0053] Then, when the modulator 2 sets the modulation order of the modulated signal lower and the regularization parameter control unit 8 changes the regularization parameters λ1 and λ2 to various values, the error rate of the demodulated signal becomes lowest with larger regularization parameters λ1 and λ2 (steps S12 and S13). Therefore, the lower the modulation order of the modulated signal (step S14), the larger the regularization parameter control unit 8 sets the regularization parameters λ1 and λ2 (the regularization parameters of the third-order or higher equalization filters are also larger), and controls the degree of freedom of compensation processing for nonlinear distortion of the modulated signal to be low (step S15).

[0054] On the other hand, when the modulator 2 sets the modulation order of the modulated signal higher and the regularization parameter control unit 8 changes the regularization parameters λ1 and λ2 to various values, the error rate of the demodulated signal is lowest with smaller regularization parameters λ1 and λ2 (steps S12 and S13). Therefore, the higher the modulation order of the modulated signal (step S14), the smaller the regularization parameters λ1 and λ2 are set by the regularization parameter control unit 8 (the regularization parameters of the third-order or higher equalization filters are also set smaller), thereby controlling the degree of freedom of compensation processing for nonlinear distortion of the modulated signal to be high (step S15).

[0055] When the regularization parameters λ1 and λ2 are excessively large, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is excessively low, resulting in an excessively high error rate for the demodulated signal, regardless of the modulation order of the modulated signal.When the regularization parameters λ1 and λ2 are excessively small, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is excessively high, resulting in an excessively high tuning of the weighting coefficient vector w2, resulting in an error rate for the demodulated signal that takes an asymptotic value higher than the minimum value, regardless of the modulation order of the modulated signal.

[0056] In this way, by not performing unnecessary distortion compensation processing, particularly when the modulation order of the modulated signal is low, it is possible to improve the error correction capability without adding nonlinear distortion.Furthermore, by performing sufficient distortion compensation processing, particularly when the modulation order of the modulated signal is high, it is possible to improve the error correction capability after sufficiently removing nonlinear distortion.

[0057] Control of the regularization parameter according to the back-off of the present disclosure is shown in Fig. 5. The modulator 2 and the encoder 1 fix the modulation method and error correction coding rate of the modulated signal to one method and value, and the amplifier 3 changes the back-off during amplification of the modulated signal to various values ​​(step S11).

[0058] Then, when the amplifier 3 sets a smaller back-off during amplification and the regularization parameter control unit 8 changes the regularization parameters λ1 and λ2 to various values, the error rate of the demodulated signal becomes lowest with larger regularization parameters λ1 and λ2 (steps S12 and S13). Therefore, the smaller the back-off during amplification (step S14), the larger the regularization parameters λ1 and λ2 are set by the regularization parameter control unit 8 (the regularization parameters of the third-order or higher equalization filters are also larger), and the degree of freedom of the compensation process for nonlinear distortion of the modulated signal is controlled to be low (step S15).

[0059] On the other hand, when the amplifier 3 sets a larger back-off during amplification and the regularization parameter control unit 8 changes the regularization parameters λ1 and λ2 to various values, the error rate of the demodulated signal becomes lowest with smaller regularization parameters λ1 and λ2 (steps S12 and S13). Therefore, the larger the back-off during amplification (step S14), the smaller the regularization parameters λ1 and λ2 are set (the regularization parameters of the third-order or higher equalization filters are also set smaller), thereby controlling the degree of freedom of compensation processing for nonlinear distortion of the modulated signal to be high (step S15).

[0060] When the regularization parameters λ1 and λ2 are excessively large, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is excessively low, resulting in an excessively high error rate for the demodulated signal, regardless of the back-off during amplification.When the regularization parameters λ1 and λ2 are excessively small, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is excessively high, resulting in the weighting coefficient vector w2 being over-tuned, resulting in an error rate for the demodulated signal that takes an asymptotic value higher than the minimum value, regardless of the back-off during amplification.

[0061] In this way, by not performing unnecessary distortion compensation processing, particularly when the back-off during amplification is small, it is possible to improve the error correction capability without adding nonlinear distortion.And, by performing sufficient distortion compensation processing, particularly when the back-off during amplification is large, it is possible to improve the error correction capability after sufficiently removing nonlinear distortion.

[0062] Control of the regularization parameter according to the coding rate of the present disclosure is shown in Fig. 6. The modulator 2 and amplifier 3 fix the modulation method of the modulated signal and the backoff during amplification to one method and value, and the encoder 1 changes the coding rate for error correction of the modulated signal to various values ​​(step S11).

[0063] Then, when the encoder 1 sets a higher error correction coding rate and the regularization parameter control unit 8 varies the regularization parameters λ1 and λ2 to various values, the error rate of the demodulated signal becomes lowest with larger regularization parameters λ1 and λ2 (steps S12 and S13). Therefore, the higher the error correction coding rate (step S14), the larger the regularization parameter control unit 8 sets the regularization parameters λ1 and λ2 (the regularization parameters of the third-order or higher equalization filters are also larger), and controls the degree of freedom of compensation processing for nonlinear distortion of the modulated signal to be low (step S15).

[0064] On the other hand, when the encoder 1 sets a lower coding rate for error correction and the regularization parameter control unit 8 varies the regularization parameters λ1 and λ2 to various values, the error rate of the demodulated signal becomes lowest with smaller regularization parameters λ1 and λ2 (steps S12 and S13). Therefore, the lower the coding rate for error correction (step S14), the smaller the regularization parameters λ1 and λ2 are set by the regularization parameter control unit 8 (the regularization parameters of the third-order or higher equalization filters are also set smaller), thereby controlling the degree of freedom of compensation processing for nonlinear distortion of the modulated signal to be high (step S15).

[0065] The amplitude and coding rate of the modulated signal of the present disclosure are shown in Figures 7 and 8. In the constellation of the 16APSK signal shown in Figure 7, the amplitude of the modulated signal becomes slightly smaller as the error correction coding rate increases in the order of R = 2 / 3, 3 / 4, 4 / 5, 5 / 6, and 8 / 9. In the constellation of the 32APSK signal shown in Figure 8, the amplitude of the modulated signal becomes slightly smaller as the error correction coding rate increases in the order of R = 3 / 4, 4 / 5, 5 / 6, 8 / 9, and 9 / 10. The constellations of the 16APSK and 32APSK signals shown in Figures 7 and 8 use 5.4.3 and 5.4.4 of the DVB-S2 standard described in Non-Patent Document 3 as examples.

[0066] Therefore, when the coding rate for error correction is higher, the regularization parameter control unit 8 sets the regularization parameters λ1 and λ2 in the same way as when the back-off during amplification is smaller, thereby controlling the degree of freedom in the compensation process for nonlinear distortion of the modulated signal. On the other hand, when the coding rate for error correction is lower, the regularization parameter control unit 8 sets the regularization parameters λ1 and λ2 in the same way as when the back-off during amplification is larger, thereby controlling the degree of freedom in the compensation process for nonlinear distortion of the modulated signal.

[0067] When the regularization parameters λ1 and λ2 are excessively large, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is excessively low, resulting in an excessively high error rate for the demodulated signal, regardless of the error correction coding rate.When the regularization parameters λ1 and λ2 are excessively small, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is excessively high, resulting in the weighting coefficient vector w2 being over-tuned, resulting in an error rate for the demodulated signal that takes an asymptotic value higher than the minimum value, regardless of the error correction coding rate.

[0068] In this way, by not performing unnecessary distortion compensation processing, particularly when the error correction coding rate is high, it is possible to improve the error correction capability without adding nonlinear distortion.And, by performing sufficient distortion compensation processing, particularly when the error correction coding rate is low, it is possible to improve the error correction capability after sufficiently removing nonlinear distortion.

[0069] A table of regularization parameters according to the present disclosure is shown in Fig. 9. A table of regularization parameters λ1 and λ2 (λ1 = λ2 or λ1 ≠ λ2) is stored in the regularization parameter control unit 8 when the modulation scheme and error correction coding rate of the modulated signal are QPSK (R = R1_1, . . . , R1_N1), 16APSK (R = R2_1, . . . , R2_N2), or 32APSK (R = R3_1, . . . , R3_N3), and the back-off during amplification of the modulated signal is X_1 [dB], X_2 [dB], . . . , X_L [dB].

[0070] The interpolation curve of the regularization parameter of the present disclosure is shown in Fig. 10. The modulation method and error correction coding rate of the modulated signal are fixed to one method and value, while the back-off during amplification of the modulated signal is discretely changed to X_1 [dB], X_2 [dB], , X_L [dB]. A curved interpolation formula or linear interpolation formula based on a table of regularization parameters λ1 and λ2 (λ1 = λ2 or λ1 ≠ λ2) is stored in the regularization parameter control unit 8.

[0071] In the embodiment, the regularization parameter control unit 8 sets the regularization parameters λ1 and λ2 according to the modulation scheme of the modulated signal, the coding rate of the error correction, and the back-off during amplification, and controls the degree of freedom of the compensation process for nonlinear distortion of the modulated signal. As a variation, the regularization parameter control unit 8 may set the regularization parameters λ1 and λ2 according to the transmission path condition (CNR) and the back-off during amplification of the modulated signal, and control the degree of freedom of the compensation process for nonlinear distortion of the modulated signal.

[0072] The reason why this variation is possible is that the transmission / reception system S adaptively controls the modulation method of the modulated signal and the coding rate of the error correction according to the transmission path conditions (CNR), and therefore the conditions of the transmission path conditions (CNR) can be said to be equivalent to the conditions of the modulation method of the modulated signal and the coding rate of the error correction.

[0073] Here, the regularization parameter control unit 8 sets the regularization parameters λ1 and λ2 larger (and also the regularization parameters of third-order or higher equalization filters larger) as the transmission path conditions (CNR) deteriorate, thereby controlling the degree of freedom of compensation processing for nonlinear distortion of the modulated signal to be lower.

[0074] On the other hand, the better the transmission path conditions (CNR), the smaller the regularization parameter control unit 8 sets the regularization parameters λ1 and λ2 (and also the regularization parameters of third-order or higher equalization filters), thereby controlling the degree of freedom of compensation processing for nonlinear distortion of the modulated signal to be higher.

[0075] The receiving device R may execute a process of estimating the transmission path condition (CNR). The regularization parameter control unit 8 may store a table of the regularization parameters λ1 and λ2 shown in Fig. 9 using the transmission path condition (CNR) and the back-off during amplification of the modulated signal as variables, and may store a curved interpolation formula or linear interpolation formula of the regularization parameters λ1 and λ2 shown in Fig. 10.

[0076] (Distortion compensation of modulated signals in the prior art and the present disclosure) Fig. 11 shows distortion compensation of modulated signals according to the conventional technology and the present disclosure. In Fig. 11, there is no nonlinear distortion and a QPSK scheme with a lower modulation order is applied. The upper left and upper right columns of Fig. 11 show the ideal and transmitted QPSK signal constellations, respectively (the same applies to the conventional technology and the present disclosure). The lower left and lower right columns of Fig. 11 show the distortion-compensated QPSK signal constellations according to the conventional technology and the present disclosure, respectively.

[0077] In the prior art, the regularization parameters λ1 and λ2 are 0, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is excessively high, and the weighting coefficient vector w2 is over-tuned, resulting in significant distortion of the signal points in the amplitude direction in the distortion-compensated QPSK signal. In the present disclosure, the regularization parameters λ1 and λ2 are larger positive values, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is appropriately low, and the signal points in the distortion-compensated QPSK signal are not distorted in the amplitude direction and are in a state almost identical to that at the time of transmission. Thus, when compensating for distortion in a QPSK signal with a lower modulation order, in the present disclosure, the regularization parameters λ1 and λ2 are larger than in the prior art, resulting in a low degree of freedom in the compensation process for nonlinear distortion of the modulated signal, and therefore nonlinear distortion is not actually added.

[0078] FIG. 12 also shows distortion compensation of modulated signals in the conventional technology and the present disclosure. In FIG. 12, a QPSK scheme with a lower modulation order is applied due to nonlinear distortion. The upper left and upper right columns of FIG. 12 show the ideal and transmitted QPSK signal constellations, respectively (similar in the conventional technology and the present disclosure). The lower left and lower right columns of FIG. 12 show the distortion-compensated QPSK signal constellations in the conventional technology and the present disclosure, respectively.

[0079] In the prior art, the regularization parameters λ1 and λ2 are 0, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is excessively high, and the weighting coefficient vector w2 is over-tuned, resulting in some distortion-compensated signal points being distorted in the amplitude direction in the QPSK signal. In the present disclosure, the regularization parameters λ1 and λ2 are larger positive values, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is appropriately low, and the signal points in the distortion-compensated QPSK signal are not distorted in the amplitude direction and are somewhat similar to those at the time of transmission. Thus, when compensating for distortion in a QPSK signal with a lower modulation order, in the present disclosure, the regularization parameters λ1 and λ2 are larger than in the prior art, resulting in a lower degree of freedom in the compensation process for nonlinear distortion of the modulated signal, and therefore, nonlinear distortion is not actually added.

[0080] FIG. 13 also shows distortion compensation of modulated signals according to the conventional technology and the present disclosure. In FIG. 13, a 32APSK scheme with a higher modulation order is applied due to nonlinear distortion. The upper left and upper right columns of FIG. 13 show the ideal and transmitted 32APSK signal constellations, respectively (the same applies to the conventional technology and the present disclosure). The lower left and lower right columns of FIG. 13 show the distortion-compensated 32APSK signal constellations according to the conventional technology and the present disclosure, respectively.

[0081] In the prior art, the regularization parameters λ1 and λ2 are 0, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is moderately high, and in the distortion-compensated 32APSK signal, the inner constellation points (excluding the outer constellation points) are in a better state than at the time of transmission. In the present disclosure, the regularization parameters λ1 and λ2 are smaller positive values, the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is moderately high, and in the distortion-compensated 32APSK signal, the inner constellation points (excluding the outer constellation points) are in a better state than at the time of transmission. Thus, when compensating for distortion in a 32APSK signal with a higher modulation order, in the present disclosure, the regularization parameters λ1 and λ2 are approximately equal, and the degree of freedom in the compensation process for nonlinear distortion of the modulated signal is approximately equal, compared to the prior art, so that nonlinear distortion can be sufficiently removed. [Industrial Applicability]

[0082] The receiving device and receiving program disclosed herein can improve error correction capability when compensating for nonlinear distortion in a modulated signal, particularly when the modulation order is low, without performing unnecessary distortion compensation processing and, in fact, without adding nonlinear distortion. [Explanation of symbols]

[0083] S: Transmitting and receiving system T: Transmitter R: Receiver 1: Encoder 2: Modulator 3: Amplifier 4: Equalizer 5: Demodulator 6: Error corrector 7: Switch 8: Regularization parameter control section 41: 1st order equalization filter 42: Weighting coefficient calculation unit 43: 3rd power calculation section 44: 3rd order equalization filter 45: Weighting coefficient calculation unit 46: Adder 47: Subtractor

Claims

1. A receiving device that receives a modulated signal from a transmitting device that transmits the modulated signal, an equalizer that compensates for nonlinear distortion of the modulated signal; a demodulator that demodulates the modulated signal whose nonlinear distortion has been compensated for; and an error corrector that performs error correction on the demodulated signal; The equalizer controls the degree of freedom of the compensation process for nonlinear distortion of the modulated signal to be lower as the modulation order of the modulated signal is lower, and controls the degree of freedom of the compensation process for nonlinear distortion of the modulated signal to be higher as the modulation order of the modulated signal is higher.

1. A receiving device adapted for wireless or wired communication.

2. The equalizer controls the degree of freedom of the compensation process for nonlinear distortion of the modulated signal to be lower as the coding rate of error correction of the modulated signal is higher, and controls the degree of freedom of the compensation process for nonlinear distortion of the modulated signal to be higher as the coding rate of error correction of the modulated signal is lower.

2. The receiving device according to claim 1, wherein:

3. The equalizer controls the degree of freedom of the compensation process for nonlinear distortion of the modulated signal to be lower as the back-off during amplification of the modulated signal is smaller, and controls the degree of freedom of the compensation process for nonlinear distortion of the modulated signal to be higher as the back-off during amplification of the modulated signal is larger.

3. The receiving device according to claim 1 or 2.

4. the equalizer adjusts weighting coefficients of the first-order and third-order or higher equalization filters so as to minimize an evaluation function including a squared error between the modulated signal, in which nonlinear distortion has been compensated, and the demodulated signal or a known signal, and a regularization term for weighting coefficients of the first-order and third-order or higher equalization filters; By setting a larger regularization parameter of the regularization term of the weighting coefficient of the equalization filter, the degree of freedom of the compensation process for the nonlinear distortion of the modulated signal is controlled to be lower, and by setting a smaller regularization parameter of the regularization term of the weighting coefficient of the equalization filter, the degree of freedom of the compensation process for the nonlinear distortion of the modulated signal is controlled to be higher.

2. The receiving device according to claim 1, wherein:

5. 10. A receiving program installed in a computer to cause the equalizer included in the receiving device according to claim 1 to control the degree of freedom of compensation processing for nonlinear distortion of the modulated signal.

Images