CN102244636B - A kind of partial transmission sequence method - Google Patents

Abstract

The invention belongs to the field of mobile communication systems and discloses a partial transmission sequence method. Aiming at the problem of high complexity of the traditional random partial transmission sequence, the method of the present invention sorts the phase factor sequence according to the correlation from low to high, and only needs to search for the previous signal with lower correlation when looking for the candidate signal with lower PAPR. A low part of the candidate signals, instead of the traversal search of the traditional random PTS, reduces the number of optimal signal searches, so that the performance similar to that of the traditional PTS can be achieved through lower complexity operations; and the candidate signals of the interleaved PTS The correlation of is only determined by the phase factor and has nothing to do with the specific signal data, so the sorting only needs to be calculated and stored once, and the system complexity is not increased.

Description

A Partial Transmission Sequence Method

technical field

The invention belongs to the field of mobile communication systems, and in particular relates to a partial transmission sequence method.

Background technique

Orthogonal Frequency Division Multiplexing (OFDM, Orthogonal Frequency Division Multiple) technology has become the core technology of the new generation of wireless communication due to its high spectrum utilization rate, good anti-multipath fading and anti-interference performance. One of the main disadvantages of the OFDM system is that the ratio of signal peak power to average power (PAPR, Peak-to-average PowerRatio) is relatively high. Here PAPR is expressed as: Among them, x _n represents the output signal obtained after the IFFT operation, max{·} represents the maximum value, and E{·} represents the average value. However, when the high-peak signal enters the saturation region of the power amplifier at the transmitting end of the system, it will generate large in-band distortion and out-of-band radiation, causing signal distortion and adjacent-band interference, resulting in system performance degradation. In order to transmit OFDM signals without distortion, the transmitter is required to have a large linear range, which increases the cost of system design and limits the wide application of OFDM technology.

Partial Transmit Sequence (PTS, Partial Transmit Sequence) is a common method to reduce the PAPR of OFDM signals. It uses multiple sequences to represent the transmission of the same set of information to reduce the probability of high-power signals. The method description can refer to Literature: Muller.SH, Huber.JB, A novel peak power reduction scheme for OFDM. The 8th IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 1997, 1090-1094. Specifically: divide the N subcarriers into V groups according to a certain grouping method, perform IFFT operation on each group to obtain x _v =IFFT{X _V }, multiply different phase factors b _v , and then multiply each Group signal superposition, by selecting different phase factors {b _v , v=1, 2, ..., V} combination to obtain multiple candidate signals Finally, the signal with the smallest PAPR is selected for transmission. Here the phase factor b _v is a complex factor with an absolute value of 1, φ _v ∈ [0, 2π], theoretically speaking, the phase φ _v of b _v is generally selected from P phase factor values, without loss of generality, usually the phase factor of the first group is fixed to 1, then the PTS generated The number of candidate signals U=P ^V-1 . Usually the phase factor b _v is selected from {+1, -1, +j, -j} or {+1, -1}, representing the phase rotation of the signal. The specific working principle is shown in Figure 1, and the specific process is shown in Figure 2.

But its disadvantage is that the complexity of the algorithm increases exponentially with the increase of the number of groups and phase rotation factors. For the OFDM system, the calculation complexity is relatively high, and the system burden is relatively large.

Contents of the invention

The object of the present invention is to propose a partial transmission sequence method in order to solve the problem of high complexity of the existing PTS method.

The technical scheme of the present invention is: a kind of partial transmission sequence method, comprises the following steps:

Step 1. Divide the N-point frequency-domain data vector X to be transmitted in a subcarrier block into V (V≥2) sub-vectors, each sub-vector has N data, wherein, N/V data and the corresponding position in X The data are the same, the other data are all zero, and satisfy X _V (v=0,..., V-1) is the sub-vector of division;

Step 2. Carry out IFFT operation to each sub-vector;

Step 3. According to the size of the cross-correlation function variable variance Δ(u, i) of any two candidate signals, according to the decision formula Sort the phase factor sequences corresponding to all U channel candidate signals from low to high, where argmin{ } represents the judgment condition when the function obtains the minimum value, U=P ^V-1 , and P is the number of phase factors First, all the phase factors of the first phase factor sequence are set to be the same; then, from the remaining U-1 phase factor sequences, select the Ω(2) that has the lowest correlation with the first phase factor sequence as the first 2 optimal phase sequences; from the remaining Uk (k=2,..., U-1) phase factor sequences, select the Ω(k+1) with the lowest average correlation with the previous k phase factor sequences as k +1 optimal phase factor sequence; finally get the phase factor sequence with correlation from low to high;

Step 4. Utilize the phase factor sequence of correlation from low to high that step 3 obtains, generate the U path candidate signal of correlation from low to high in turn;

Step 5. Select the L candidate signals with lower correlation in front of the U channel candidate signals generated in step 4 to calculate their PAPR values respectively, and select the candidate signal with the lowest PAPR as the optimal candidate signal output.

Beneficial effects of the present invention: the method of the present invention sorts the phase factor sequence according to the correlation from low to high, when looking for a candidate signal with a lower PAPR, it only needs to search for a part of the candidate signal with a lower correlation in front, Instead of the traversal search of the traditional random PTS, the number of optimal signal searches is reduced, so that the performance similar to that of the traditional PTS can be achieved through lower complexity operations; and the correlation of each candidate signal of the interleaved PTS is only determined by the phase factor The decision has nothing to do with the specific signal data, so the sorting only needs to be calculated and stored once, which does not increase the complexity of the system.

Description of drawings

Fig. 1 is the working principle diagram of the transmitter of PTS.

Fig. 2 is a schematic flow chart of searching for the best candidate signal by using the traditional stochastic PTS method.

Fig. 3 is a working principle diagram of random PTS and interleaved PTS.

Figure 4 is a schematic flow diagram of the method of the present invention.

detailed description

Specific embodiments of the present invention will be given below in conjunction with the accompanying drawings. It should be noted that the parameters in the examples do not affect the generality of the present invention.

For ease of understanding of the present invention, before setting forth specific embodiment, at first introduce the term used wherein:

Random PTS: Each subcarrier is randomly assigned to V PTSs, and the specific working principle is shown in Figure 3.

Interleaved PTS: Allocate subcarriers with an interval of V in one PTS. The specific working principle is shown in Figure 3.

Alternative signal correlation analysis: equipment x ^u and x ⁱ are the u-th and i-th alternative signals in the interleaved PTS, Indicates the phase factor of the corresponding packet on the kth subcarrier in the uth candidate signal, where When the kth subcarrier is in the vth group, N is the number of subcarriers. Therefore, the cross-correlation function between any two candidate signals (denoted as u, i) at any time point (denoted as m, n) is defined as:

Assuming that the system input data X _n is an independent and identically distributed random variable with an average power of 1, and defining τ=mn, the above formula can be simplified as:

Since R _u,i (τ) is symmetric about τ, it is only necessary to analyze the part where τ is positive, therefore:

Define cross-correlation function variables The variance Δ(u,i) of :

in, for mean value. Cross-correlation function random variables in all candidate signals The lower the arithmetic mean of the variance, the lower the correlation between each candidate signal, and the better the PAPR suppression effect of the system. At the same time, it can be seen that the correlation of each candidate signal is only determined by the phase factor, which is different from the specific Signal data is irrelevant.

The method flow diagram of the present invention is as shown in Figure 4, and concrete steps are as follows:

Step 1. Divide the N-point frequency-domain data vector X to be transmitted in a subcarrier block into V (V≥2) sub-vectors, each sub-vector has N data, wherein, N/V data and the corresponding position in X The data are the same, the other data are all zero, and satisfy X _V (v=0,..., V-1) is the sub-vector of division;

Step 2. Carry out IFFT operation to each sub-vector;

Step 3. According to the size of the cross-correlation function variable variance Δ(u, i) of any two candidate signals, according to the decision formula Sort the phase factor sequences corresponding to all U-channel candidate signals from low to high, where argmin{ } represents the judgment condition when the function obtains the minimum value, that is, when selecting the k-th phase factor sequence for sorting, Calculate the average cross-correlation value of the uth (k≤u<U) phase factor sequence and the previous k-1 phase factor sequence, and select the original uth phase factor sequence that minimizes the decision formula as the sorted kth phase Factor sequence, U=P ^V-1 , P is the number of phase factors, specifically: first set all phase factors of the first phase factor sequence to be all the same, here you can choose a sequence whose phase factors are all 1; then From the remaining U-1 phase factor sequences, select the Ω(2) that has the lowest correlation with the first phase factor sequence as the second optimal phase sequence; from the remaining Uk(k=2,..., U-1) in the phase factor sequence of the road, select the Ω(k+1) with the lowest average correlation with the phase factor sequence of the previous k roads as the optimal phase factor sequence of the k+1 road; finally obtain the phase of the correlation from low to high factor sequence.

Step 4. Utilize the phase factor sequences with low to high correlations obtained in step 3 to sequentially generate U-channel candidate signals with low to high correlations.

Step 5. Select the L candidate signals with lower correlation in front of the U channel candidate signals generated in step 4 to calculate their PAPR values respectively, and select the candidate signal with the lowest PAPR as the optimal candidate signal output.

In this embodiment, the number of subcarriers N=256, b _v is selected from {+1, -1}, that is, the number of phase factors P=2, the number of divisions V=8, and U=P ^V-1 =128 Group alternative signal generation. For the traditional random PTS, the number of complex multiplications required for the IFFT operation of V groups is Complex addition is Nlog ₂ N. According to the fact that 1 complex number multiplication is equivalent to 18 real number additions, and 1 complex number addition is equivalent to 2 real number additions, it can be concluded that the total equivalent real number addition complexity of the IFFT operation in V divided random PTS is: 11VNlog ₂ N .

Here, interleaved division is adopted in step 1. For the interleaved PTS, the Cooley- TukeyFFT algorithm, which can reduce the complexity of complex multiplication to Complex addition reduces to The entire IFFT process needs to be performed addition of real numbers. If there are L candidate signals, (V-1)NL complex number additions are required for PTS phase factor combination, and 2NL real number multiplications and NL real number additions are required to search for the candidate signal with the lowest PAPR. The number of equivalent real number additions of the total complexity of the entire random PTS algorithm is C _R-PTS = 11Nlog ₂ N+(2V+7)NU, and the equivalent number of additions of real numbers of the total complexity of the interleaved PTS algorithm is

For the method proposed by the present invention, if L=64 candidate signals are selected from all generated U=128 candidate signals, the total computational complexity C _{I-PTS of the interleaved PTS} =526336, and the complexity at this moment is only The random PTS total complexity CR _-PTS = 53.4% of 933888. Through complexity analysis, it can be seen that the complexity of the interleaved PTS is much lower than that of the random PTS. Under the same computational complexity, compared with random PTS, interleaved PTS can use more sub-blocks V and candidate signals L to achieve better PAPR suppression performance.

The above example is only a preferred example of the present invention, and the use of the present invention is not limited to this example. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention should be included in this document. within the scope of protection of the invention.

Claims (4)

1. A partial transmission sequence method, comprising the steps of:

step 1, dividing N point frequency domain data vectors X to be sent in a subcarrier block into V sub-vectors, wherein V is larger than or equal to 2, each sub-vector has N data, wherein N/V data are the same as data of corresponding positions in X, other data are zero, and the requirement of meeting the requirement of the condition that the N/V data are the same as the data of corresponding positions in X is metX_VIs a divided sub-vector, V is 0, …, V-1;

step 2, performing IFFT operation on each subvector;

step 3, according to the magnitude of the cross-correlation function variable variance delta (u, i) of any two paths of alternative signals, according to a decision formulaSorting the phase factor sequences corresponding to all the U-path alternative signals from low correlation to high correlation, wherein a cross-correlation function variable is definedVariance Δ (u, i) of (a):

$\mathrm{\&Delta;}(u,i)=\frac{1}{N}\underset{\mathrm{\&tau;}=0}{\overset{N-1}{\&Sigma;}}{\left(\right|R_{u,i}\left(\mathrm{\&tau;}\right){|}^2-\mathrm{\&eta;}\left(u,i\right))}^2$

wherein,is composed ofThe mean value of (a);

argmin {. cndot } represents a decision condition when the function takes a minimum value, U ═ P^V-1And P is the number of phase factors, firstly setting all the phase factors of the 1 st path of phase factor sequence to be the same; then selecting omega (2) with the lowest correlation with the 1 st path phase factor sequence from the rest U-1 path phase factor sequences as a 2 nd optimal phase sequence; selecting omega (k +1) with the lowest average correlation with the former k paths of phase factor sequences from the rest U-k paths of phase factor sequences as k +1 paths of optimal phase factor sequences, wherein k is 2, … and U-1; finally, a phase factor sequence with the correlation from low to high is obtained;

step 4, sequentially generating U-path alternative signals with the correlation from low to high by using the phase factor sequence with the correlation from low to high obtained in the step 3;

and 5, selecting the L alternative signals with lower correlation in the front among the U-path alternative signals generated in the step 4, respectively calculating the PAPR values of the L alternative signals, and selecting the alternative signal with the lowest PAPR from the L alternative signals as the optimal alternative signal to be output.

2. The method of claim 1, wherein the partitioning of step 1 is an interleaved partitioning.

3. The method according to claim 1 or 2, wherein the IFFT operation in step 2 is performed by using Cooley-Tukey FFT algorithm.

4. The method of claim 3 wherein the phase factor is selected from { +1, -1 }.