CN102546510B

CN102546510B - Method for decreasing peak-to-average power ratio of orthogonal frequency division multiplexing (OFDM) signal

Info

Publication number: CN102546510B
Application number: CN201210006605.2A
Authority: CN
Inventors: 江涛; 李海波
Original assignee: Huazhong University of Science and Technology
Current assignee: Huazhong University of Science and Technology
Priority date: 2012-01-09
Filing date: 2012-01-09
Publication date: 2014-05-07
Anticipated expiration: 2032-01-09
Also published as: CN102546510A

Abstract

The invention discloses a method for decreasing a peak-to-average power ratio (PAPR) of an orthogonal frequency division multiplexing (OFDM) signal. The method comprises the following steps of: clipping a time-domain sequence x, projecting a clipped signal to a linear space consisting of all time-domain sub-sequences, calculating a phase rotation factor al<(v)> of each time-domain sub-sequence x<(v)>, multiplying each time-domain sub-sequence x<(v)> by using a corresponding phase rotation factor al<(v)>, superposing results, calculating a new linear combination signal xSLC, judging whether PAPRSLC is less than or equal to PAPR0 or not according to the PAPRSLC of the xSLC, finishing iteration if PAPRSLC is less than or equal to PAPR0, performing repeated iteration if PAPRSLC is not less than or equal to PAPR0, performing parallel/serial conversion on the time-domain sequence of which the PAPR is decreased, and transmitting the time-domain sequence to a receiver. The phase rotation factors are generated by iteration and projection, the PAPR can be decreased to a certain extent by the iteration of each time, and computational complexity is linearly increased along with the increasing of the number V of the phase rotation factors; and compared with the conventional method, the invention can greatly reduce required computational complexity when the PAPR is required to be decreased to a great extent.

Description

Method for reducing peak-to-average power ratio of OFDM signal

Technical Field

The invention belongs to the technical field of wireless communication by adopting Orthogonal Frequency Division Multiplexing (OFDM) signals, and particularly relates to a method for reducing the peak-to-average power ratio of an OFDM signal based on grouping linear combination.

Background

In a mobile radio channel, time-and frequency-selective fading occurs when a signal passes from a transmitting antenna through a time-varying multipath channel to a receiving antenna. The time-varying nature of the channel causes a broadening of the signal spectrum, resulting in the Doppler (Doppler) effect, causing the signal to fade selectively over time. Orthogonal Frequency Division Multiplexing (OFDM) techniques have been proposed in view of the frequency selective fading characteristics exhibited by multipath channels in the frequency domain. The OFDM technique is a technique of dividing a frequency domain into a plurality of subchannels, and orthogonally overlapping adjacent subchannels with each other to improve spectrum utilization efficiency. Therefore, on one hand, OFDM can overcome frequency selective fading; on the other hand, a time interval smaller than the coherence time is taken as the duration of one OFDM symbol, so that the influence of time-selective fading of the channel on the transmission system can be greatly reduced.

However, there are still some important problems with OFDM technology that are not well solved. One of the difficulties and key techniques is how to control the peak-to-average power ratio (ratio of peak power to average power). Because the OFDM signal is formed by superimposing a plurality of subcarrier signals, the peak-to-average power ratio of the OFDM signal is very high compared to a constant envelope signal such as a single carrier signal. The linear dynamic range of the transmitter power amplifier will require a wide range if the peak-to-average power ratio of the signal is high. A power amplifier with a wide linear dynamic range will increase the cost of the transmitter greatly. Although the power amplifier with low linear dynamic range can reduce the cost, the transmission signal will be seriously distorted, thereby causing the serious reduction of the system performance. Therefore, in order to reduce the requirement of the OFDM signal for the transmission power amplifier, it is necessary to reduce the peak-to-average power ratio of the OFDM signal.

Currently, many methods have been proposed to reduce the peak-to-average power ratio of OFDM signals. Among them, Partial Transmit Sequences (Partial Transmit Sequences) is an effective method. Assuming that the number of subcarriers of the OFDM system is N, one OFDM frequency domain signal X includes N modulation symbols, X ═ X (0), X (1),.., X (N-1)]. The main idea of a partial transmission sequence can be summarized as: dividing the modulated frequency domain signal X into V mutually disjoint frequency domain subsequences, being { X^(v)V-0, 1., V-1}, each subsequence is still N in length, i.e., each subsequence contains

One modulation symbol being non-zero and the others being filled with zeros, i.e.

Here, the number of subsequences V is greater than or equal to 2, the value of V is determined by the required reduction of the peak-to-average power ratio of the OFDM system signal, and generally, V is 4, 8 or 16. Then, an Inverse Fast Fourier Transform (IFFT) operation of N points, i.e., x, is performed on each frequency domain subsequence^(v)＝IFFT(X^(v)). For each x^(v)Multiplied by a phase rotation factor b^(v)And are summed, then

To facilitate the transmission of the phase rotation factor to the receiving end, b^(v)Typically selected from four discrete values {1, -1, j, -j }, wherein

Traverse all possible 4^VA phase rotation factor b^(v)To obtain different time domain signals x, and selecting the signal with the minimum peak-to-average power ratio in the time domain signals to transmit.

It can be seen that, although the partial transmission sequence method can effectively reduce the peak-to-average power ratio, when a relatively large reduction of the peak-to-average power ratio is required, the number of packets V needs to be increased to generate a large number of candidate signals, and the peak-to-average power ratio of each candidate signal is calculated and the smallest candidate signal is selected as the transmission. This makes the computation complexity high, since it increases exponentially with the number of packets V.

Disclosure of Invention

Aiming at the defect of high complexity of a method for reducing the peak-to-average power ratio in the prior art, the invention provides a novel method for reducing the peak-to-average power ratio of an OFDM system signal, and the method greatly reduces the calculation complexity while reducing the peak-to-average power ratio.

The invention provides a method for reducing the peak-to-average power ratio of OFDM system signals, which comprises the following steps:

(1) coding, interleaving and modulating an input bit stream to obtain a frequency domain data signal;

(2) inserting pilot symbols in a frequency domain data signal according to a comb pattern to obtain a frequency domain sequence X { X (k) }, k 0, 1., N-1}, wherein N is the total subcarrier number of an OFDM system, and k represents a subcarrier sequence number; performing N-point fast Fourier inverse transformation on the frequency domain sequence X to obtain a time domain sequence X ═ X (0), X (1),.., X (N-1) ];

(3) after the frequency domain sequence X is subjected to serial-parallel conversion, the frequency domain sequence X is divided into V frequency domain subsequences { X ] with equal length in an adjacent mode^(v)V-1, where V denotes the number of frequency domain subsequences, and V is 2 ≦ V < N;

X^(v)＝[X^(v)(0)，X^(v)(1)，...，X^(v)(N-1)]，

(4) for each frequency domain subsequence X^(v)Performing N-point fast Fourier inverse transformation to obtain corresponding time domain subsequence x^(v)＝[x^(v)(0)，x^(v)(1)，...x^(v)(N-1)]；

(5) The peak-to-average power ratio reduction processing is carried out on the time domain sequence x, and specifically comprises the following substeps:

(5.1) let x_lX, the iteration number l is 0;

(5.2) for x_lShearing to obtain shearing signal x_c，l＝[x_c，l(0)，x_c，l(1)，...，x_c，l(N-1)]；

(5.3) shearing the signal x_c，lProjection onto all time-domain subsequences { x }^(v)V-0, 1, V-1} in a linear space L (x)⁽⁰⁾，x⁽¹⁾，....，x^(V-1)) Then as followsCalculating by a formula to obtain each time domain subsequence x^(v)By a phase rotation factor of_l ^(v)，

Wherein N-0, 1.., N-1, represents a conjugate;

(5.4) dividing each time domain subsequence x^(v)Are multiplied by the corresponding phase rotation factors a respectively_l ^(v)And the results are superposed to obtain a linear combination signal

(5.5) calculating a new linear combination signal

Relaxation factor lambda_lCalculated according to the following formula:

wherein f (n) x_c，1(N) -x (N), N-0, 1.., N-1, I is the set maximum number of iterations, λ^cIs a constant, λ^c∈[1，2]；

(5.6) calculating a new linear combination signal x_SLCPAPR of_SLCDetermine whether the PAPR is present_SLC≤PAPR₀Wherein the PAPR₀If the target peak-to-average power ratio is set, the iteration is terminated, and the step (6) is carried out; if not, let l be l +1, x_l＝x_SLCAnd (5.2) turning to the step;

(6) and after the time domain sequence with the reduced peak-to-average power ratio is subjected to parallel-serial conversion, the time domain sequence is sent to a receiving end.

The invention provides a method for reducing the signal peak-to-average power ratio of an OFDM system, which generates a phase rotation factor sequence through iterative projection, wherein each iteration can have a certain peak-to-average power ratio reduction amount, a larger peak-to-average power ratio reduction amount is obtained after a plurality of iterations, and compared with the traditional partial sequence transmission method, the required calculation complexity is greatly reduced when a larger peak-to-average power ratio reduction amount is required. At the receiving end, the phase rotation factor sequence used by the transmitting end is recovered by using the existing segmented channel estimation method, so that the bit error rate is not influenced. Simulation experiments prove that the calculation complexity of the method can be greatly reduced when a larger peak-to-average power ratio reduction is required.

Drawings

Fig. 1 is a system block diagram at a transmitter end of an OFDM system according to the present invention;

FIG. 2 is a block diagram of an iterative OFDM signal peak-to-average power ratio reduction method described in the present invention;

fig. 3 is a system block diagram at a receiver end of the OFDM system according to the present invention.

Detailed Description

The invention is further illustrated with reference to the figures and examples.

As shown in fig. 1, the total number of subcarriers of the OFDM system is N, k represents a subcarrier number, and a system frequency domain signal is defined as X { X (k) ═ 0,1, and N-1}, where the signal is divided into V subblocks (V is greater than or equal to 2 and less than N), and the value of V is determined by the reduction required by the peak-to-average power ratio of the OFDM system signal, and generally V is 4, 8, or 16.

The method for reducing the average peak power ratio of the OFDM system comprises the following steps:

(2) inserting pilot symbols in comb pattern (comb type) in the frequency domain data signal, i.e. when

When (i ═ 1, 2., 2V), a pilot symbol is inserted to obtain a frequency domain sequence X, and N-point inverse fast fourier transform is performed on X to obtain a time domain sequence X ═ X (0), X (1),.., X (N-1)]；

(3) After the X is converted in a serial-parallel mode, the X is divided into V frequency domain subsequences X with equal length in an adjacent mode^(v)Wherein X is^(v)＝[X^(v)(0)，X^(v)(1)，...，X^(v)(N-1)]，

Thus is provided withFrequency domain subsequence X^(v)Including the data signal, pilot symbols, and 0.

(4) Performing fast Fourier inverse transformation of N points on each frequency domain subsequence to obtain a corresponding time domain subsequence x^(v)Wherein

x^(v)＝[x^(v)(0)，x^(v)(1)，...x^(v)(N-1)]；

(5) As shown in fig. 2, the peak-to-average power ratio reduction processing is performed on the time-domain sequence x according to steps (5.1) to (5.6). Setting PAPR₀The target peak-to-average power ratio is usually set within the range of 5-9 dB; a. the₀The preset shearing threshold is usually set to be 5-15 dB, is determined according to the selected modulation mode, I is the corresponding maximum iteration number and is generally determined by a simulation experiment, and lambda is^cIs a constant number, satisfies lambda^c∈[1，2]；

(5.1) let x_lX, the iteration number l is 0;

(5.2) for the sequence x_lShearing to obtain shearing signal x_c，l，x_c，l＝[x_c，l(0)，x_c，l(1)，...，x_c，l(N-1)]The shearing treatment is as follows:

wherein x is_l(n) is x_lN-1, N represents the total number of subcarriers in the OFDM system, and θ (N) is x_lPhase of (n), A₀Is a clipping threshold;

x_c，land (n) is a time domain sequence after shearing.

(5.3) shearing the signal x_c，1Projection onto all time-domain subsequences { x }^(v)V-0, 1, V-1} in a linear space L (x)⁽⁰⁾，x⁽¹⁾，....，x^(V-1)) Then, the phase rotation factor of each time-domain subsequence is calculated as follows:

wherein N-0, 1.., N-1, represents a conjugate; | represents an absolute value;

(5.4) dividing each time domain subsequence x^(v)Are multiplied by the corresponding phase rotation factors a respectively_l ^(v)And are superposed to obtain a linear combined signal

Namely, it is

(5.5) calculating a new linear combination signal

Relaxation factor lambda_lCalculated according to the following formula,

wherein f (n) x_c，1(n)-x(n)，n＝0，1，...，N-1，λ^cIs a constant, λ^c∈[1，2]；

(5.6) calculating a new linear combination signal x_SLCPAPR of_SLCDetermine whether the PAPR is present_SLC≤PAPR₀If yes, stopping iteration and entering the step (6); if not, let l be l +1, x_l＝x_SLCAnd (5) transferring to the step (5.2).

The following describes an information processing method at the receiving end, as shown in fig. 3.

(a) After the received signal is subjected to serial-parallel conversion, Fast Fourier Transform (FFT) and parallel-serial conversion, a receiving end frequency domain sequence Y is obtained,

Y＝{Y(k)，k＝0，1，...，N-1}；

(b) dividing Y into V receiving end frequency domain subsequences { Y) in the same way as in the step (3)^(v)，v＝0，1，...，V-1}，

Wherein, h (k) represents a channel frequency domain response value on the k-th sub-carrier, and w (k) is a frequency domain value of gaussian white noise on the k-th sub-carrier;

(c) and V takes values from 0 to V-1, and all data symbols in the frequency domain subsequences of the V receiving ends are calculated according to the steps (c1) to (c 2):

(c1) defining virtual channel frequency response

Is composed of

Calculated using equation (c 1):

wherein,

is shown in

The frequency domain signal values on the sub-carriers,denotes the ith pilot symbol (i ═ 1, 2.,. 2V) in the vth receiving end frequency domain subsequence, and the positions and values of all pilot symbols are known at the receiving end;

calculating the virtual channel frequency response at all pilot frequency symbol positions in the v-th receiving end frequency domain subsequence by using the formula (c1), obtaining the virtual channel frequency response at the subcarrier where the data symbol of the v-th receiving end frequency domain subsequence is located by interpolation by using the obtained virtual channel frequency response value at the pilot frequency symbol position, and obtaining the virtual channel frequency response on the data symbol subcarrier of the v-th receiving end frequency domain subsequence by the following formula if linear interpolation is adopted, namely the virtual channel frequency response on the data symbol subcarrier of the v-th receiving end frequency domain subsequence is obtained by the following formula

= a_{I}^{(v)} [H (k_{i}^{(v)}) + (\frac{H (k_{i + 1}^{(v)}) - H (k_{i}^{(v)})}{L}) (k^{(v)} - k_{i}^{(v)})], (k_{i}^{(v)} < k^{(v)} < k_{i + 1}^{(v)})

Wherein k is^vIs the position of the data symbol in the v-th receiving end frequency domain sub-sequence,

if other interpolation methods are adopted, the calculation formulas are different, and the interpolation method is not required;

(c2) receiving end using virtual channel frequency responseTo calculate the data symbol D' (k) of the v-th receiving end frequency domain sub-sequence^(v))：

D′(k^(v))＝Y^v(k^(v))/H′(k^(v))；

(d) Combining the data symbols in the V receiving end frequency domain sub-sequences obtained in the step (c) into a sequence according to the size of the sub-carrier serial number, and then demodulating, de-interleaving and decoding the sequence to obtain an output bit stream.

According to the process of generating the phase twiddle factors in the method, the method provided by the invention directly calculates the value of the V phase twiddle factors, the complexity of the method is linearly increased along with the increase of V, and through analysis, the complexity required by each iteration is not very high, the iteration times are not very large, and thus the total calculation complexity is much lower than the complexity of the traditional partial sequence transmission method which is exponentially increased along with the increase of V.

The embodiment of the invention adopts the following specific parameter scheme: the number of subcarriers of the OFDM system is 64, the number of packets is 8, the number of pilots is 16, the input signal is a Quadrature Phase Shift Keying (QPSK) modulated signal, the clipping threshold is set to 2.0, λ is set to^c1.95, target peak-to-average power ratio (PAPR)₀6.0dB, corresponding to a maximum number of iterations I of 20.

The simulation result shows that the complementary cumulative distribution function CCDF is 10^-3When the peak-to-average power ratio of the OFDM signal is reduced to 4.3dB, the required real number addition times and the required partial sequence transmission method ratio of the method are reduced by 99.68%, and the required real number multiplication times and the required partial sequence transmission method ratio are reduced by 97.66%.

The above is an example of the present invention, but the present invention should not be limited to the disclosure of the example and the drawings. Therefore, it is intended that all equivalents and modifications which do not depart from the spirit of the invention disclosed herein are deemed to be within the scope of the invention.

Claims

1. A method for reducing the peak-to-average power ratio of a signal in an OFDM system, the method comprising the steps of:

(2) inserting a pilot symbol in a frequency domain data signal according to a comb pattern to obtain a frequency domain sequence X (k), wherein k is 0,1, N-1, and k represents a subcarrier sequence number of an OFDM system; performing N-point fast Fourier inverse transformation on the frequency domain sequence X to obtain a time domain sequence X ═ X (0), X (1),.., X (N-1) ];

(3) after the frequency domain sequence X is subjected to serial-parallel conversion, the frequency domain sequence X is divided into V frequency domain subsequences { X ] with equal length in an adjacent mode^(v)V-1, where V is equal to or greater than 2 < N,

X^(v)＝[X^(v)(0),X^(v)(1),...,X^(v)(N-1)]，

(4) for each frequency domain subsequence X^(v)Performing N-point fast Fourier inverse transformation to obtain corresponding time domain subsequence x^(v)=[x^(v)(0),x^(v)(1),...x^(v)(N-1)]；

(5) The peak-to-average power ratio reduction processing is carried out on the time domain sequence x, and comprises the following substeps:

(5.1) let x_lX, the iteration number l is 0;

(5.2) for x_lShearing to obtain shearing signal x_c,l＝[x_c,l(0),x_c,l(1),...,x_c,l(N-1)]；

(5.3) shearing the signal x_c,lProjection onto all time-domain subsequences { x }^(v)V-0, 1, V-1} in a linear space L (x)⁽⁰⁾,x⁽¹⁾,....,x^(V-1)) Then, each time domain subsequence x is obtained by calculation according to the following formula^(v)By a phase rotation factor of_l ^(v)，

Wherein N-0, 1.., N-1, represents a conjugate;

(5.4) dividing each time domain subsequence x^(v)Are multiplied by the corresponding phase rotation factors a respectively_l ^(v)And the results are superposed to obtain a linear combination signal x_SLC；

(5.5) calculating a new linear combination signal x_SLC＝x_l+λ_l(x_SLC-x_l) (ii) a Relaxation factor lambda_lCalculated according to the following formula,

wherein f (n) x_c,1(N) -x (N), N-0, 1.., N-1, I is the set maximum number of iterations, λ^cIs a constant, λ^c∈[1,2]；

2. The method of claim 1, wherein the time domain sequence x is represented by the following equation_l(n) carrying out shearing,

wherein N is 0,1,.., N-1; n represents the total number of subcarriers of the OFDM system, and theta (N) is x_lPhase of (n), A₀Is a clipping threshold;

x_c,land (n) is a time domain sequence after shearing.