CN1131629C - Method and device for decreasing peak nean power ratio in multi-carrier communication system - Google Patents

Abstract

The invention belongs to the technical field of communication, and relates to a method and a device for reducing the peak-to-average power ratio of signals in a multi-carrier communication system. It specifically includes the following steps: at the transmitting end, the power of the input signal of the power amplifier is redistributed through numerical transformation: according to the probability distribution of the input signal power, the conversion point of the numerical transformation is determined; according to the requirements for the input signal power of the power amplifier, the numerical transformation is determined The rules; at the receiving end, the received signal is inversely transformed according to the inverse transformation rule corresponding to the transmitting end. The invention can effectively reduce the peak-to-average power ratio of the input signal of the power amplifier with relatively low calculation amount and implementation complexity, and avoid nonlinear distortion caused by the power amplifier when amplifying the signal.

Description

Method and device for reducing peak-to-average power ratio of signals in multi-carrier communication system

Technical Field The present invention belongs to the field of communication technology, and in particular relates to an effective method for controlling the peak-to-average power ratio of transmitted signals in a multi-carrier communication system including an OFDM communication system and an implementation device thereof.

Background Art At present, Orthogonal Frequency Division Multiplexing (OFDM) technology has been widely used as an effective broadband transmission technology in the transmission of Digital Audio Broadcasting (DAB), Digital Video Broadcasting (DVB), and High Definition Television (HDTV). and many other fields. However, one of the main problems of the system using OFDM technology is the ratio of the peak power to the average power of the signal, that is, the peak-to-average power ratio (PAPR) is high, which requires the transmitter power amplifier in the system to be highly linear to avoid amplifying the signal. Linear distortion greatly increases the difficulty and cost of system implementation, and limits the further development of OFDM technology. Especially for mobile two-way communication systems, this shortcoming is even more prominent due to the limited volume and energy consumption of transceivers. Also, the selection of a multi-carrier transmission scheme for a communication system currently does not maximize the use of the system's available power.

In order to reduce the PAPR of signals in OFDM communication systems, people have done a lot of research. Existing methods can be roughly divided into two categories.

One is to process the input data stream before the multiplexer, as shown in Figure 1a, to reduce the possibility of high PAPR, such as encoding, partial transmission sequence and selection of mapping methods. When the number of sub-carriers is large, these methods all require a large amount of calculation or complexity. However, in the current practical OFDM system, the number of sub-carriers is generally large, so its calculation amount or complexity is often unacceptable. Moreover, for the two methods of partial transmission sequence and selective mapping, additional overhead needs to be added to inform the receiving end of the transmission coefficient or combination scheme adopted by the sending end.

The other is to process the output data stream of the multiplexer after the multiplexer, as shown in Figure 1b, to reduce the impact of high PAPR, such as limiting filtering and companding. Limiting the signal will introduce limiting noise, and when the filtering method is used to eliminate these noises, the PAPR of the signal will be increased again, which cannot meet the requirements of the system design. Another method is to use the companding technique in non-uniform quantization to reduce the PAPR of the signal, such as μ-rate companding and A-rate companding. In traditional communication systems, when non-uniform quantization is performed on analog signals, in order to reduce quantization noise, companding technology can be used to compress the quantization interval of small signals at the transmitting end and expand the quantization interval of small signals at the receiving end. In a multi-carrier digital communication system, if the signal is compressed at the transmitter, according to the input and output characteristics of the compressor, the average power of the compressed signal will increase while the peak power remains unchanged, so the PAPR of the signal can be reduced. Although this method is simple to implement, it increases the average power of the input signal of the power amplifier at the transmitting end because it compresses and expands the signal at the transmitting and receiving ends of the system, and does not consider the probability distribution of signal power or amplitude. , making the compressed signal more sensitive to the nonlinearity of the power amplifier, which eventually leads to the deterioration of system performance.

SUMMARY OF THE INVENTION The purpose of the present invention is to overcome the deficiencies of the prior art, and propose a method and device for reducing the peak-to-average power ratio of signals in a multi-carrier communication system, which can achieve a relatively small amount of calculation and complexity , effectively reducing the PAPR of the input signal of the power amplifier, and avoiding nonlinear distortion caused by the power amplifier amplifying the signal.

The invention provides a method for reducing the peak-to-average power ratio of signals in a system through numerical transformation. In this method, before the digital/analog conversion or at the same time as the digital/analog conversion, the amplitude value of each point of the input signal is mapped to the amplitude value of the output signal according to a specific functional relationship, and the output signal satisfies the PAPR requirement of the power amplifier in the system Require. In other words, this method is to redistribute the power of the signal through numerical transformation at the transmitting end of the communication system to enhance the small signal while weakening the large signal, which not only reduces the PAPR of the signal but also improves the anti-noise performance of the small signal. It can be understood that for the receiving end of the communication system, in order to make a correct signal decision, it is necessary to perform corresponding inverse transformation on the received signal before the signal decision, that is, to weaken the small signal while enhancing the large signal. Since the power amplifier at the receiving end is usually a small-signal power amplifier, it can have very good linear performance, so the inverse conversion will not have a great impact on the performance of the entire system.

The method specifically includes the following steps:

The first step is to determine the conversion point of the numerical transformation at the transmitting end: according to the probability distribution of the input signal power, the expected value (statistical mean) of the signal power is obtained, and it is used to determine the conversion point in the numerical transformation, that is, for less than this point The signal with a value of 0 is enhanced and transformed, and the signal with a value greater than this point is weakened and transformed; (Usually, for different systems, the signal characteristics required to determine the numerical transformation, including statistical characteristics and instantaneous characteristics, may be different, the simplest of which is It is to use the expected value (statistical mean) and peak value of the signal power to determine the parameters in the numerical transformation.)

The second step is to determine the rules of numerical transformation at the transmitting end: let the input signal be represented by x(t), and the output signal after numerical transformation be represented by y(t), the rule of numerical transformation is x(t) and y( The function mapping relationship between t) is represented by f[·]. In the present invention, f[·] consists of two parts f ₁ [·] and f ₂ [·], which satisfy the following conditions:

in Indicates that the absolute value is taken; m is the conversion point parameter determined by the expected value and peak characteristics of the signal power obtained in the first step, and the function input and output curves corresponding to the conversion rule include linear, piecewise linear, nonlinear symmetric and nonlinear quasi-symmetric transformation curve;

In the third step, at the receiving end, determine the inverse transformation rule for the received signal corresponding to the transmitting end: Similarly, let the input signal be represented by x(t), and the output signal after numerical inverse transformation be represented by y(t), and the numerical value The rule of inverse transformation, that is, the function mapping relationship between x(t) and y(t) is represented by g[·]. In the present invention, g[·] consists of two parts g ₁ [·] and g ₂ [·], which satisfy the following conditions:

in means to take the absolute value; m' is the conversion point parameter corresponding to the parameter m obtained in the second step, and the inverse transformation rule satisfies g[·]=f ^-1 [·].

The method can be used in a system including a multi-carrier communication system that needs to reduce the peak-to-average power ratio of the transmitted signal or is equivalent to reducing the peak-to-average power ratio of the transmitted signal.

The present invention is a device for reducing the peak-to-average power ratio of transmitted signals, which is used in the above-mentioned method in a multi-carrier communication system. The device includes a threshold judgment unit that divides the input signal x(t) into |x(t)|≥m and |x(t)|<m, and a threshold judgment unit that receives |x(t)|≥m Expander, a compressor that receives |x(t)|<m part of the signal and a digital/analog converter that converts the superimposed digital signal after the compressor and expander into an analog signal, where m is according to the input signal The conversion point parameter determined by the expected value of the signal power obtained from the probability distribution of the power, the expander attenuates the signal, and the attenuation rules include linear, piecewise linear, nonlinear symmetric and nonlinear quasi-symmetric attenuation; the compressor performs signal attenuation Amplification, magnification rules include linear, piecewise linear, nonlinear symmetric, and nonlinear quasi-symmetric forms of magnification.

Another device for reducing the peak-to-average power ratio of the transmitted signal used in the above method of the present invention. The device includes a threshold judgment unit that divides the input signal x(t) into |x(t)|≥m and |x(t)|<m, and a threshold judgment unit that receives |x(t)|≥m An expansion-type D/A converter that converts the signal and a compression-type D/A converter that receives |x(t)|<m part of the signal and converts the signal, where m is according to the power of the input signal The conversion point parameter is determined by the expected value of the signal power calculated by the probability distribution. The extended digital/analog converter attenuates and converts the signal. The attenuation rules include linear, piecewise linear, nonlinear symmetric and nonlinear quasi-symmetric forms. Attenuation; the compression-type D/A converter amplifies and converts the signal, and the amplification rules include linear, piecewise linear, nonlinear symmetric and nonlinear quasi-symmetric amplification.

The effect of the invention is that it not only reduces the PAPR of the signal with relatively small calculation amount and complexity, but also effectively avoids the nonlinear distortion caused when the power amplifier amplifies the signal. Moreover, the anti-noise performance of the small signal is improved, and the performance of the present invention is obviously better than that of the existing method using limiting filtering under the additive Gaussian white noise channel.

Description of drawings

Fig. 1a is a schematic diagram of traditional processing of an input data stream of a multiplexer before the multiplexer.

Fig. 1b is a schematic diagram of processing the output data stream of the multiplexer after the conventional multiplexer.

Fig. 2 is four compression-expansion curves corresponding to four method embodiments of the present invention.

FIG. 3 is a waveform diagram of signals before and after transformation in the first method embodiment of the present invention.

Fig. 4a is a structural block diagram of the first embodiment of the converter proposed by the present invention.

Fig. 4b is a structural block diagram of the second embodiment of the converter proposed by the present invention.

Fig. 5 is a block diagram of a transmitter and receiver of a multi-carrier communication system employing the present invention.

FIG. 6 is a performance comparison diagram of four method embodiments of the present invention and the prior art.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be explained in more detail according to four method embodiments and two device embodiments and their applications in conjunction with the accompanying drawings.

In the present invention, since the small signal is strengthened while the large signal is weakened by compressing the small signal while expanding the large signal, the input-output curve of the converter is called the compression-expansion curve, and the rule f[· ] can be well represented by the input and output curves of the converter, as shown in Figure 2.

In the method for reducing the peak-to-average power ratio of signals in a multi-carrier communication system proposed by the present invention, there are many shapes of compandor curves satisfying transformation rules, corresponding to different compandor curves, the present invention provides four different methods Example.

Embodiment one, referring to Fig. 2, comprises the following steps:

In the first step, at the transmitting end, the conversion point 44 of the compandor curve in the numerical transformation is determined. First, according to the probability distribution of the input signal power, its expected value is obtained. Then use this expected value as the conversion point 44 of companding, and make the ratio of companding the same, because for the input signal, the probability that its power is greater than this expected value and less than this expected value is equal, so the total power of the signal before and after companding is not the same. Change. As for the probability distribution of signal power, according to the central limit law, the amplitude of OFDM signal obeys Gaussian distribution, and its power obeys Chi-square distribution. Since for a Gaussian random variable X with zero mean and variance δ ² , the random variable Y=X ² obeys the Chi-square distribution, and E(Y)=δ ² , ${\mathrm{\&delta;}}_{\mathrm{the\; <figure-callout\; id="2"\; label="y"\; filenames="C0110221100132.png"\; state="\{\{state\}\}">y</figure-callout>}}^2=2{\mathrm{\&delta;}}^4.$ E( ) means to take the mathematical expectation, and δ _y ² is the variance of Y. For OFDM signal, δ ² is its average power, which can be obtained easily. Therefore, ^δ2 is the value corresponding to the transition point 44 of companding.

In the second step, at the transmitting end, determine the rules of the numerical transformation, that is, the shape of the compression-expansion curve in the numerical transformation. Its input and output conversion rules are as follows:

In the formula, the parameters k and b are called slope and intercept respectively, and they respectively satisfy 0<k<1 and b=(1-k)*m, where m is the expected value of the input signal power;

The compression-expansion curve of this embodiment is symmetrical about the origin, and the compression-expansion curve of the positive signal part is shown as the solid line 42 in FIG. 2 . Correspondingly, this transformation is called "linear companding".

In the third step, at the receiving end, determine the shape of the compression-expansion curve in the inverse transformation, and its input and output transformation rules are as follows:

In the formula, the parameter k' is consistent with the k obtained in the second step, and the parameter b' satisfies b'=(1-k')*m', where m' is the expected value of the input signal power of the inverse converter.

FIG. 3 is an effect diagram of Embodiment 1, showing waveforms before and after linear companding (parameter k=0.1) on the signal at the sending end in Embodiment 1 of the described method. In this embodiment, the small signal is enhanced, while the large signal is weakened, and the peak value of the signal is well suppressed. The compressed and expanded signal reduces the linearity requirement on the post-transmitter power amplifier, and avoids the nonlinear distortion caused when the post-transmitter power amplifier amplifies the signal.

Embodiment 2 of the method of the present invention: other steps of this embodiment are the same as those of Embodiment 1 except that the transformation rules adopted in the second step, ie, the compression-expansion curve and its implementation methods are different. In the described embodiment, the implementation of the conversion rule uses the digital operation circuit to perform the conversion directly, and complete the digital/analog conversion at the same time. The compandor curve of the second embodiment is symmetrical about the origin, and the compandor curve of the part where the signal is positive is shown as the solid line 43 in Fig. 2. Correspondingly, this transformation is called "piecewise linear compandor", which The parameter is k. The input and output conversion rules of this embodiment are as follows:

Among them, k is a normal number less than 1.

Embodiment 3 of the method of the present invention: other steps of this embodiment are the same as Embodiment 1 except that the transformation rule adopted in the second step, that is, the compression-expansion curve and its implementation are different. In the described embodiment, the implementation of the conversion rule uses a storage mapping table to perform one-to-one mapping of the input values, and simultaneously completes the digital/analog conversion. The compandor curve of this embodiment is symmetrical about the origin, and the compandor curve for the part where the signal is positive is shown as the dotted line 40 in Figure 2. Correspondingly, this transformation is called "non-linear symmetric compandor", wherein, The companding rule adopts the universal μ-rate companding, and its parameter is the companding rate u.

Embodiment 4 of the method of the present invention: other steps of this embodiment are the same as those of Embodiment 1 except that the transformation rule adopted in the second step, that is, the compression-expansion curve and its implementation methods are different. In the described embodiment, the implementation of the conversion rule uses the digital operation circuit to perform the conversion directly, and complete the digital/analog conversion at the same time. The compandor curve of this embodiment is symmetrical about the origin, and the compandor curve of the part where the signal is positive is shown as the dotted line 41 in Figure 2. Correspondingly, this transformation is called "non-linear quasi-symmetric compandor", Among them, the companding rule adopts the universal μ-rate companding, and its parameter is companding rate u.

Two embodiments of the converter proposed by the present invention are shown in Figure 4, which are described as follows in conjunction with the accompanying drawings:

Converter device embodiment one of the present invention: for designing compandor and digital/analog converter respectively, as shown in Figure 4a, this converter comprises a threshold judging unit 58, an expander 52, a compressor 53 and a Digital/analog converter 56. First, the output of the IFFT digital signal processor 29 is used as the input digital signal x(t) 30 of the converter, and x(t) is divided into |x(t)|≥m and |x( t)|<m two parts. The threshold judging unit can be realized by a software program or a threshold judging circuit, so as to cooperate with the realization of subsequent devices. Then the portion 50 of |x(t)|≥m is sent to the expander 52 , and the portion 51 of |x(t)|<m is sent to the compressor 53 . (Before the present invention, compressor is usually used in A/D conversion to realize non-uniform quantization, because it compresses the quantization interval of small signal, thereby is called compressor. And in the present invention, compressor is used On the contrary, the expander is used to weaken the large signal. The feasibility of this application is guaranteed by the input and output characteristics of the compressor and the expander. Usually, the compressor 53 and the expander 52 can be stored by a mapping table or digital operation circuits.) Then, the compressed and expanded digital signals are superimposed to form a digital signal y(t)54 with a greatly reduced peak-to-average power ratio. Finally, the digital signal passes through a digital-to-analog converter to output an analog signal 32 with a peak-to-average power ratio meeting the requirements.

Embodiment 2 of the converter device of the present invention: jointly design the compandor and the digital/analog converter. As shown in FIG. 4 b , the converter includes a threshold judging unit 58 , an expansion type D/A converter 56 and a compression type D/A converter 57 . In this embodiment, a digital operation circuit is used to directly convert the input value according to a predetermined functional relationship, and simultaneously complete the digital/analog conversion. The threshold judging unit can be realized by a software program or a threshold judging circuit, so as to cooperate with the realization of subsequent devices. Then the part 50 of |x(t)|≥m and the part 51 of |x(t)|<m are sent to the expansion type D/A converter 56 and the compression type D/A converter 57 respectively, and their output By superposition, an analog signal 32 with a peak-to-average power ratio meeting the requirements can be obtained.

The embodiment structure of the transceiver adopting the multi-carrier two-way communication system of the present invention is shown in Figure 5, but it can be understood that the four method embodiments of the present invention are not limited to be used in two-way communication equipment, so the multi-carrier communication transmission The machine can be considered as a device physically separate from the multi-carrier communication receiver.

The converter 31 designed according to the present invention can be used in the transmitter 21 of the multi-carrier communication system as shown in the upper block of FIG. 5 . The transmitter 21 includes a source 23 generating an information stream 24, an encoder 25, a serial-to-parallel converter 27, a modulator 28, an inverse fast Fourier transform (IFFT) digital signal processor 29 and a converter 31 .

In this embodiment, the data symbol stream 24 carrying information generated by the information source 23 is sent to the encoder 25, and the encoder 25 performs source encoding and channel encoding on the information in sequence. The purpose of source coding is usually to compress the redundancy of information carried by the data symbol stream 24, so as to transmit more information in the same time. The channel coding usually enhances the system's anti-interference or error control during information recovery by increasing the correlation between data. The binary serial data stream 26 generated by the encoder 25 is converted into a binary parallel data stream through a serial-to-parallel converter 27 . The modulator 28 performs the first modulation on the binary parallel data stream, and maps the binary data on each parallel channel to points on the signal constellation diagram, that is, data in complex form. In a multi-carrier communication system, in order to adapt to a channel whose frequency response changes with time, different modulation modes can be used for the output data stream of each parallel channel. In the described embodiment, considering that the frequency response of the general channel does not change too drastically, the same modulation method is used for each channel, and the mapping relationship is shown in Table 1 below, which is the so-called quaternary phase shift key control (QPSK), where $i=\sqrt{-1}.$

Table 1 Quadrature Phase Shift Keying enter 00 01 10 11 output -1-i -1+i 1-i 1+i

Then, the IFFT digital signal processor 29 performs a second modulation on the complex-valued parallel sequence, and at the same time merges the data streams of each channel into a serial data stream. Since performing IFFT processing on a digital signal is equivalent to frequency modulating the original signal with a set of mutually orthogonal sine (cosine) signals, this process is also called orthogonal frequency division multiplexing. Next, the modulated serial data stream 30 is sent to a converter 31 . According to the requirements of the system and the available equipment, the converter 31 can map the input values one by one according to the predetermined functional relationship by storing the mapping table, so as to meet the required PAPR. Then, digital/analog conversion is performed on the output value of the storage mapping table, and the mapping precision is determined by the capacity of the storage mapping table. Or the converter 31 uses a digital operation circuit to directly convert the input value according to a predetermined functional relationship, so as to meet the required PAPR, and complete the digital/analog conversion at the same time. The analog signal 32 output by the converter is finally sent to a post-stage circuit (not shown in the figure) including a power amplifier. At this time, the PAPR of the analog signal 32 output by the converter meets the requirements of the subsequent power amplifier, that is, the signal will not enter the nonlinear region of the power amplifier, so that it will not cause nonlinear distortion.

The inverse converter 33 designed according to the present invention can be used in the receiver 22 of the multi-carrier communication system as shown in the lower box of FIG. 5 . The receiver 22 includes an inverse transformer 33, a fast Fourier transform (FFT) digital signal processor 34, a demodulator 35, a parallel-to-serial converter 36, a decoder 37 and a sink 39 for receiving information.

The inverse converter 33 converts the signal sent from the front-end circuit (not shown in the figure) of the receiver 22 inversely to the operation performed by the converter 31 in the transmitter 21 on the signal 30 . The design principle of the inverse converter is basically the same as that of the converter, that is, the analog/digital conversion can be performed first, and then the input values can be mapped one by one according to the predetermined functional relationship through the storage mapping table, or the digital operation circuit can be used to The input value is directly transformed according to the predetermined functional relationship, and the analog/digital conversion is completed at the same time. For the inverse converter 33, the selection of these two implementation methods does not have to be consistent with the converter 31, but should consider the requirements of the receiver and the equipment that can be provided. The output data stream of the inverse converter 33 first passes through the FFT digital signal processor 34 to complete the first demodulation, that is, orthogonal frequency division multiplexing. Then the output of the FFT digital signal processor 34 is sent to the demodulator 35 to complete the second demodulation. Due to the influence of channel interference and noise, there is a certain error between the input signal of the demodulator 35 and the output signal of the modulator 28 of the transmitter 21, which is manifested in the signal constellation diagram as the offset and rotation of the signal constellation points, Therefore, the demodulator 35 will perform signal judgment and output a signal that matches the constellation point of the original signal. Parallel converter 36 converts the parallel data output from demodulator 35 into a serial data stream. Then the serial data stream is sent to the decoder 37, and the decoder 37 completes channel decoding and source decoding in sequence. Finally, the decoder 37 outputs received information 38 that is substantially the same as the original information 24 to meet the requirements of the sink 39 .

Compared with the encoding, partial transmission sequence and selective mapping methods introduced above, the calculation amount and implementation complexity of the present invention are obviously much smaller.

FIG. 6 is a comparison chart of the performance of the system adopting the four method embodiments of the present invention and the performance of the prior art. In all embodiments, without loss of generality, the multi-carrier communication system uses 128 subcarriers and uses QPSK modulation. By setting parameters reasonably, the four embodiments of the present invention can reduce the average power without increasing Signal PAPR: Linear companding (parameter k=0.5) and nonlinear quasi-symmetric companding (parameter u=1) reduce about 4.1dB, nonlinear symmetric companding (parameter u=4) reduce about 2.9dB, and segment Linear companding (parameter k=0.5) reduces about 5.6dB. It can be seen that the present invention can greatly reduce the peak-to-average power ratio of the signal, and prevent the signal from entering the nonlinear region of the power amplifier, thereby avoiding the nonlinear distortion caused by it.

Under the same parameter setting, for an additive Gaussian white noise channel, when the bit error rate is required to be equal to 10 ^-5 , the linear companding, piecewise linear companding, nonlinear symmetric companding and nonlinear quasi-symmetric companding require The signal-to-noise ratios are 14dB, 13dB, 13.5dB and 12.2dB respectively, and for the same PAPR improvement, the signal-to-noise ratio required by the existing limiting filtering method is 18dB. Compared with it, linear companding, piecewise linear companding, nonlinear symmetric companding and nonlinear quasi-symmetric companding have 4dB, 5dB, 4.5dB and 5.8dB improvement respectively. It can be seen that the performance improvement of the system applying the present invention is obvious.

It will of course be understood that the foregoing description has been given by way of example only and that modifications may be made within the scope of the invention. For example, other transformation rules, that is, other compandor curves and parameter setting schemes, and other physical implementation methods including hardware and software may be used.

In the current multi-carrier communication system, the present invention can fully replace the existing methods of limit filtering and companding and expansion. At the same time, for systems with high real-time requirements such as wireless multimedia communication and the receiver should not be too complicated, the present invention can be used as a candidate solution to reduce the peak-to-average power ratio of multi-carrier signals, and maintain the multi-carrier communication system itself. has advantages.

It should be pointed out that the multi-carrier communication technology of the present invention can also be applied to optical fiber communication systems.

Claims (6)

1. A method for reducing the peak-to-average power ratio of signals in a multi-carrier communication system, characterized in that the power of the input signal of the power amplifier at the transmitting end is redistributed through numerical transformation, and the small signal is enhanced while the large signal is weakened, specifically including the following step: The first step is to determine the conversion point of numerical conversion at the transmitter: according to the probability distribution of the input signal power, the expected value of the signal power is obtained, and it is used to determine the conversion point in the numerical conversion. Enhanced transformation, for the signal greater than the value of the point, the weakened transformation is performed; The second step is to determine the rules of numerical transformation at the transmitting end: let the input signal be represented by x(t), and the output signal after numerical transformation be represented by y(t), the rule of numerical transformation is x(t) and y( The function mapping relationship between t) is represented by f[·]; where, f[·] is composed of f ₁ [·] and f ₂ [·], which satisfy the following conditions: In the formula

Indicates that the absolute value is taken; m is the conversion point parameter determined by the expected value and peak characteristics of the signal power obtained in the first step, and the function input and output curves corresponding to the conversion rule include linear, piecewise linear, nonlinear symmetric and nonlinear quasi-symmetric transformation curve; In the third step, at the receiving end, determine the inverse transformation rule for the received signal corresponding to the transmitting end: Similarly, let the input signal be represented by x(t), and the output signal after numerical inverse transformation be represented by y(t), and the numerical value The rule of inverse transformation, that is, the functional mapping relationship between x(t) and y(t) is represented by g[·]; where, g[·] is composed of g ₁ [·] and g ₂ [·], They meet the following conditions: In the formula

means to take the absolute value; m' is the conversion point parameter corresponding to the parameter m obtained in the second step, and the inverse transformation rule satisfies g[·]=f ^-1 [·]. 2. The method according to claim 1, characterized in that, in said second step, said input-output conversion rules are as follows:

In the formula, the parameters k and b are called slope and intercept respectively, and they satisfy 0<k<1 and b=(1-k)*m respectively; 3. The method according to claim 1, characterized in that the input-output transformation rules in the second step are as follows:

Among them, k is a normal number less than 1. 4. The method according to claim 1, characterized in that, in said third step, at the receiving end, determine the shape of the compression-expansion curve in the inverse transformation, and its input and output transformation rules are as follows:

In the formula, the parameter k' is consistent with the k obtained in the second step, and the parameter b' satisfies b'=(1-k')*m', where m' is the expected value of the input signal power of the inverse converter. 5. A device for adopting the method according to claim 1, characterized in that it comprises a device for dividing the input signal x(t) into two parts |x(t)|≥m and |x(t)|<m The threshold judgment unit, an expander that receives |x(t)|≥m part of the signal, a compressor that receives |x(t)|<m part of the signal and a digital that will be superimposed after the compressor and expander A digital-to-analog converter that converts signals into analog signals, where m is the conversion point parameter determined by the expected value of the signal power obtained from the probability distribution of the input signal power, and the expander attenuates the signal. The attenuation rules include linear and piecewise linear , nonlinear symmetric and nonlinear quasi-symmetric attenuation; the compressor amplifies the signal, and the amplification rules include linear, piecewise linear, nonlinear symmetric and nonlinear quasi-symmetric amplification. 6. A device for employing the method as claimed in claim 1, characterized in that it comprises an input signal x(t) divided into |x(t)|≥m and |x(t)|<m Part of the threshold judgment unit, an extended digital/analog converter that receives |x(t)|≥m part of the signal and converts the signal, and a receiving |x(t)| Converted compression-type digital/analog converter, where m is the conversion point parameter determined by the expected value of the signal power obtained according to the probability distribution of the input signal power, and the expansion-type digital/analog converter performs attenuation and digital-to-analog conversion on the signal, and the attenuation The rules include linear, piecewise linear, nonlinear symmetric and nonlinear quasi-symmetric attenuation; the compression type D/A converter amplifies and converts the signal, and the amplification rules include linear, piecewise linear, nonlinear symmetric and non-linear Amplification in linear quasisymmetric form.

Images