CN111683032A

CN111683032A - Improved DST-GP peak-to-average power ratio (DST-GP) inhibition method in ACO-OFDM system

Info

Publication number: CN111683032A
Application number: CN202010429013.6A
Authority: CN
Inventors: 袁建国; 李尚晋; 王露; 赵富强; 方小倩; 袁雅琴
Original assignee: Chongqing University of Post and Telecommunications
Current assignee: Chongqing University of Post and Telecommunications
Priority date: 2020-05-19
Filing date: 2020-05-19
Publication date: 2020-09-18

Abstract

The invention relates to a peak-to-average power ratio (DST-GP) restraining method in an ACO-OFDM system, in particular to an improved DST-GP peak-to-average power ratio restraining method based on a Discrete Sine Transformation (DST) method and a General Precoding (GP) method. According to the method, the original DST matrix is expanded into an asymmetric matrix to obtain better PAPR (peak-to-average power ratio) inhibition performance, and then calculation of an expansion part is optimized based on the principle of a GP (GP) method, so that the calculation complexity can be reduced. Simulation results show that compared with the original DST technology, the method and the device have the advantages that better PAPR suppression performance is obtained at lower cost, and good compromise between the PAPR suppression performance and the calculation complexity can be achieved.

Description

Improved DST-GP peak-to-average power ratio (DST-GP) inhibition method in ACO-OFDM system

Technical Field

The invention belongs to the field of asymmetric clipping light Orthogonal Frequency Division Multiplexing (ACO-OFDM) system peak-to-average ratio inhibition, and relates to an improved Discrete Sine Transform-Generalized Precoding (DST-GP) peak-to-average ratio inhibition method.

Background

An asymmetric Clipped Optical-orthogonal frequency Division Multiplexing (ACO-OFDM) system is a multi-carrier system that simultaneously transmits multiple narrowband signals at different frequencies, and when multiple same-phase signals are superimposed, a large Peak is generated, resulting in a high Peak-to-Average Power Ratio (PAPR). Such high peak-to-average ratio signals can degrade system performance when passing through the core components LEDs of the visible light communication. The principle of the Discrete Sine Transform (DST) technology is that a DST matrix is adopted to process signals after constellation mapping, and the maximum peak power of the processed signals is reduced due to the property of the unitary matrix, so that the PAPR of the signals can be suppressed.

However, in the original DST technique, the DST matrix is a square matrix having the same length as the input data sequence, and the PAPR suppression performance is fixed and cannot be flexibly adjusted. The Precoding Matrix (PM) method can extend an original Discrete Fourier Transform (DFT) Matrix into an asymmetric Precoding Matrix, and can flexibly suppress PAPR by adjusting extension, but the computational complexity thereof rapidly increases as the extension portion increases. The General Precoding (GP) method optimizes the calculation in the PM method, reduces the computational complexity while achieving the same PAPR suppression performance, and applies the principle thereof to Discrete Hartley Transform (DHT) and Zaoff-Chu Transform (ZCT) Precoding techniques, but other Precoding techniques, such as DST, cannot directly apply the principle thereof.

The invention provides an improved Discrete Sine Transform-Generalized Precoding (DST-GP) peak-to-average ratio inhibition method based on DST and GP methods. The method not only improves the PAPR suppression performance at lower cost and obtains good compromise between the PAPR suppression performance and the calculation complexity, but also can flexibly adjust the PAPR suppression performance by adjusting and expanding. The analysis shows that the calculation complexity of the DST-GP method is lower than that of the PM method in the document [1] "Reducing the Peak-to-Average Power Ratio of OFDM Signal through decoding [ J ]. IEEE Transactions on vehicle Technology,2007,56(2): 686-. Simulation analysis shows that the PAPR suppression performance of the DST-GP method is superior to that of the original DST technology, and under the same expansion condition, the PAPR suppression performance is similar to that of a DHT-based generalized precoding (DHT-GP) method based on the discrete Hartley transform in the literature [2] generalized precoding method for PAPR reduction with low complexity in OFDM systems [ J ]. IET Communications,2018,12(7):796-808 ].

Disclosure of Invention

In view of this, the present invention aims to provide a DST-GP method based on DST and GP methods, which expands an original DST matrix U into an asymmetric matrix T to obtain better peak-to-average ratio suppression performance. And meanwhile, a coefficient matrix G is constructed based on the coefficient of the DST formula, and a precoding matrix P of the DST-GP method is obtained by multiplying. The method can improve the peak-to-average ratio inhibition performance at lower cost, and can flexibly adjust the expansion size to adjust the peak-to-average ratio inhibition performance. This approach can thus achieve a good compromise between PAPR suppression performance and computational complexity.

In order to achieve the purpose, the invention provides the following technical scheme:

firstly, according to a DST unitary kernel element formula, an N multiplied by N original DST matrix U is expanded into an N multiplied by L asymmetric matrix T, and the purpose is to increase the expansion L to obtain the improvement of PAPR inhibition performance;

secondly, constructing an L multiplied by L diagonal coefficient matrix G based on the coefficients of the DST formula;

and then, optimizing calculation in the pre-coding process according to the rule of L-N row elements behind the asymmetric matrix T so as to achieve the purpose of reducing the calculation complexity.

Finally, under the same simulation environment, the PAPR performance of the DST-GP peak-to-average ratio inhibition method and the original DST technology provided by the patent is compared, and the PAPR performance of the DST-GP method and the PAPR performance of the DHT-GP method are compared under the same expansion condition.

The invention has the beneficial effects that:

a DST-GP method based on DST and GP methods is presented. According to the method, an original DST matrix is expanded into an asymmetric matrix according to a DST unitary kernel element formula, a coefficient matrix is constructed based on DST formula coefficients, and calculation in a precoding process is optimized according to the property of an asymmetric matrix expansion part to reduce calculation complexity. The method can improve the peak-to-average ratio inhibition performance at lower cost, and can flexibly adjust the expansion size to adjust the peak-to-average ratio inhibition performance. This approach can thus achieve a good compromise between PAPR suppression performance and computational complexity. The analysis shows that the DST-GP method is less computationally complex than the PM method of document [1 ]. Simulation analysis shows that the PAPR inhibition performance of the DST-GP method is superior to that of the original DST technology, and the PAPR inhibition performance is similar to that of the DHT-GP method in the document [2] under the condition of the same expansion

Drawings

In order to make the object, technical scheme and beneficial effect of the invention more clear, the invention provides the following drawings for explanation:

FIG. 1 is a technical roadmap for the process of the invention;

FIG. 2 is a flow chart of the DST-GP process;

FIG. 3 is a comparison graph of PAPR suppression performance simulation of the DST-GP method and the prior DST technology;

FIG. 4 is a comparison graph of PAPR suppression performance simulation of the DST-GP method and the DHT-GP method;

Detailed Description

The present invention will be described in detail below with reference to the accompanying drawings.

1. With reference to fig. 1, the precoding matrix P in the proposed DST-GP method is constructed from the original DST matrix according to a certain rule. And setting U as an original DST matrix of NxN, wherein a unitary core element formula in U is shown as a formula (1):

wherein i and m respectively represent the ith row and the mth column, and 0 is less than or equal to i, and m is less than or equal to N-1.

Expanding U into an L multiplied by N asymmetric matrix T, and deriving that the elements in T satisfy the formula (2):

namely, the odd-numbered column elements of the L-N-1 row behind the matrix T are the periodic expansion of the odd-numbered column elements of the front L-N-1 row, and the even-numbered column elements of the rear L-N-1 row are the periodic expansion after the even-numbered column elements of the front L-N-1 row take negative values.

Constructing an L multiplied by L diagonal coefficient matrix G shown as a formula (3):

and the element formula in the matrix G is shown as (4):

the precoding matrix P in the DST-GP method proposed by the present invention is obtained, and P can be represented by formula (5).

Wherein N is_p＝L-N。

In the formula (5), after N_pThe sign of the Nth column element of row-1 is related to the parity of N, taking a negative value if N is odd and taking a positive value if N is even.

Order to

Representing a sequence of input data, order

Representing the output data sequence, the whole pre-coding process can be represented as:

order to

Let N be even (N parity only with the last N in T_pRow-1 and column-N elements are positive or negative), then

Can be expressed as:

from (7) obtaining the odd-numbered row elements and

middle to odd row element product sum, last N_P-1 line element is top N_P-a periodic extension of 1 row of elements; the first N rows of elements of the sequence Y' are even-numbered elements of each row in U and

middle to even row element product sum, last N_P-1 line element is top N_PPeriodic extension of line 1 elements after taking negative values, and both lines N +1 are 0. According to this principle, the last N_PThe calculation of the rows can be simplified.

From the flow chart shown in fig. 2, the DST-GP method is specifically implemented as follows:

step1 combines the matrix U with the output data sequence

Multiplying to obtain Y and Y' front N line elements, and adding the two to obtain

The first N rows of elements. In this step, N is present²Sub-multiplication and N (N-1) addition.

Step2 obtaining N after Y and Y' according to the principle obtained in (7)_P-1 line element, which are added to obtain

Rear N_P-1 line element, and

the element in row N +1 is 0, which results in

All of the elements of (a). This step has L-N-1 multiplications and L-N-1 additions.

Step3

Multiplying the coefficient matrix G to obtain an output data sequence

There are L multiplications in this step.

2. The complexity analysis of the proposed DST-GP method is described with reference to table 1 and table 2, and m (n) and d (n) respectively represent the multiplication complexity and the addition complexity in the DST-GP method:

M(N)＝N²+2L-N-1 (8)

D(N)＝N²+L-2N-1 (9)

define β as the ratio of redundant data to input data (β generally does not exceed 100%):

β＝(L-N)/N＝N_p/N (10)

for the purpose of comparison of the computational complexity, the multiplication computation complexity reduction rate and the addition computation complexity reduction rate of the DST-GP method with respect to the PM method in document [1] are expressed by CM and CA, respectively, and are defined as equations (11) and (12), respectively:

table 1 shows the complexity comparison between the PM process and the DST-GP process. Taking N as an example 64, table 2 shows the CM and CA corresponding to 25% and 50% β, respectively.

TABLE 1 multiplication and addition complexity in PM and DST methods

It can be seen from table 1 that in the DST-GP method, the multiplication complexity and the addition complexity do not increase rapidly with the increase of the extension L.

TABLE 2 complexity comparisons of the PM method and the DST-GP method at 25% and 50%, respectively (N64)

From table 2, it can be seen that when β is 25% and 50%, CM is 18.1% and 31.3%, CA is 19.7% and 32.8%, respectively, which effectively reduces the computational complexity, and when β is 25% and 50%, comparing the computational complexity of the DST-GP method itself, it can be seen that when N is fixed, the computational complexity of the DST-GP method does not rapidly increase with the increase of the extension L, and the larger L, the better the effect of reducing the computational complexity.

3. With reference to fig. 3 and fig. 4, in order to verify that the DST-GP method proposed by the present patent has good PAPR suppression performance, Matlab simulation analysis is performed. The parameter is set to be that the number N of subcarriers is 64, the modulation mode is Quadrature Phase Shift Keying (QPSK) modulation, and the oversampling coefficient LL is 4. In the simulation chart of the peak-to-average power ratio suppression, the horizontal coordinate PAPR₀Represents PAPR threshold value, ordinate (CCDF-Pr PAPR>PAPR₀]) Is a complementary cumulative distribution function that is used to measure PAPR suppression performance.

FIG. 3 is a CCDF plot of the DST-GP method at β of 12.5%, 25%, 37.5% and 50%, respectively.

Compared with the original DST technology, when CCDF is 10-4, the DST-GP method makes PAPR reach β of 12.5%, 25%, 37.5% and 50%₀The reduction is respectively about 0.46, 1.07, 1.14 and 1.66dB, and compared with the original signal, the reduction is respectively about 3.86, 4.47, 4.88 and 5.06dB, and the better PAPR inhibition performance is shown. And the DST-GP process is inWhen N is fixed, the PAPR restraining performance can be flexibly adjusted by adjusting the size of the expansion L, the PAPR restraining performance of the DST-GP method is improved along with the increase of the expansion L, and only a small amount of calculation complexity is increased.

FIG. 4 is a comparison of PAPR inhibition performance of the DST-GP method when beta is 25% and 50%, respectively, with that of the DHT-GP method in document [2 ]. As can be seen from fig. 4, when β takes the same value, the DST-GP method achieves PAPR suppression performance close to that of the DHT-GP method.

In summary, compared to the original DST technology, the present invention achieves better PAPR suppression performance at lower cost, and can also achieve a good tradeoff between PAPR suppression performance and computational complexity.

Finally, it is noted that the above-mentioned preferred embodiments illustrate rather than limit the invention, and that, although the invention has been described in detail with reference to the above-mentioned preferred embodiments, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the scope of the invention as defined by the appended claims.

Claims

1. The invention relates to an improved Discrete Sine Transform-Generalized Precoding (DST-GP) peak-to-average ratio suppression method based on a DST (Discrete Sine Transform, DST) and a Generalized Precoding (GP) method in an asymmetric clipping light orthogonal Frequency Division Multiplexing (ACO-OFDM) system, which expands an original DST matrix U into an asymmetric moment T and aims to obtain better PAPR suppression performance. And meanwhile, a coefficient matrix G is built based on the DST formula coefficient, and a precoding matrix P of the DST-GP method is obtained by multiplying. The method can improve the peak-to-average ratio inhibition performance at lower cost, can flexibly adjust the expansion size to adjust the peak-to-average ratio inhibition performance, and can achieve good compromise between the PAPR inhibition performance and the calculation complexity.

2. The method for suppressing the peak-to-average ratio of DST-GP based on the DST and GP methods according to claim 1, wherein: the method comprises the steps of expanding an N multiplied by N original DST matrix U into an L multiplied by N asymmetric matrix T according to a DST unitary kernel element formula, flexibly adjusting the peak-to-average ratio inhibition performance by changing the size of L, enabling the peak-to-average ratio inhibition performance to be better when the L is larger, and then constructing an L multiplied by L diagonal coefficient matrix G based on the coefficient of the DST formula.

3. The DST-GP peak-to-average ratio suppression method based on the DST and GP methods according to claim 2, wherein: according to the DST unitary kernel element formula, an element of an N +1 row of T is 0, an element of an odd column of a back L-N-1 row of T is periodic expansion of an odd column element of a front L-N-1 row, an element of an even column of the back L-N-1 row is periodic expansion of an even column element of the front L-N-1 row, wherein the even column element of the back L-N-1 row is negative value periodic expansion, and according to the property, calculation of an expansion part in a multiplication process of input data A and a precoding matrix P can be optimized, complexity is reduced, and therefore the method can obtain improvement of peak-to-average ratio inhibition performance at low cost.