CN109889466A - A kind of improved low complex degree PTS peak-to-average force ratio Restrain measurement of joint in optical OFDM system - Google Patents

Abstract

The present invention relates to a kind of peak-to-average force ratio Restrain measurement of optical OFDM system, in particular to a kind of improved low complex degree TS-PTS-Clipping peak-to-average force ratio Restrain measurement of joint.The program is mainly improved by partial transmission sequence technology and Clipping technical tie-up to realize the inhibition to high peak-to-average power ratio.In this scenario, initial input signal is subjected to d grades of IFFT transformation first, then carries out the division of sub-block, remaining n-d grades of IFFT then is carried out to the signal after piecemeal and is converted.Simulation analysis shows that the present invention has preferable peak-to-average ratio inhibition performance, also reduces computation complexity.The present invention is also a kind of scheme for taking into account error rate of system and peak-to-average ratio inhibition performance simultaneously, and the program can meet the inhibition to optical OFDM system high peak-to-average power ratio, further relate to effectiveness of the invention and novelty.

Description

A kind of improved low complex degree PTS peak-to-average force ratio Restrain measurement of joint in optical OFDM system

Technical field

The invention belongs to light orthogonal frequency division multiplexing (Optical Orthogonal Frequency Division Multiplexing, O-OFDM) system peak-to-average than suppression technology field, is related to a kind of improved TS-PTS-Clipping of joint (two-stage Partial Transmit Sequence-Clipping, TS-PTS-Clipping) peak-to-average force ratio Restrain measurement, The program is mainly by being based on partial transmission sequence (Partial Transmit Sequence, PTS) technology and clipping (Clipping) technical tie-up improves to realize the inhibition to high peak-to-average power ratio.

Background technique

Light orthogonal frequency division multiplexing (Optical Orthogonal Frequency Division Multiplexing, O- OFDM) system is by optical communication technique and orthogonal frequency division multiplexing (Orthogonal Frequency Division Multiplexing, OFDM) technology organically combines, and it can be the case where guaranteeing each subcarrier noiseless parallel transmission The lower effective availability of frequency spectrum for improving system.Since it is with availability of frequency spectrum height, anti-intersymbol interference (Inter Symbol Interference, ISI), anti-channel fading and resisting chromatic dispersion ability are strong, inhibit nonlinear effect many advantages, such as, at present by A variety of high speed data transmission systems are as technical standard.However, high papr (Peak-to-average Power Ratio, PAPR) it has always been considered as being the most important disadvantage of coherent light OFDM technology.High PAPR requires high power Amplifier works in nonlinear area, this will will increase inband distortion and out-of-band radiation, so as to cause under O-OFDM system performance Drop.In order to effectively inhibit high PAPR to influence brought by O-OFDM system, current experts and scholars both domestic and external are for height PAPR proposes many solutions, such as clipping (Clipping), selected mapping method (Selective Mapping, SLM), portion Divide the technologies such as transmission sequence (Partial Transmit Sequence, PTS).In current solution, fractional transmission sequence Arranging (PTS) is a kind of promising technology, no any loss in terms of message transmission rate, power and bit error rate (BER), Other than needing to transmit side information, PAPR rejection can be significantly improved.But using a series of quick in PTS scheme Inverse Fourier transform (Inverse Fast Fourier transform, IFFT) thus increases to generate different alternative signals Computation complexity.

Summary of the invention

In view of this, can not only reduce computation complexity the purpose of the present invention is to provide a kind of but also can take into account preferable The peak-to-average force ratio scheme of PAPR rejection.

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

(1) will have N=2 firstⁿThe O-OFDM signal of a subcarrier indicates are as follows:

Wherein X={ X_k, (k=0,1 ..., N-1) represents the symbol sebolic addressing of input, and t represents discrete-time variable.If DefinitionWherein T_iIt is that the symmetrical matrix of a N × N represents the i-stage of IFFT operation, then formula (1) It can rewrite are as follows:

Enable S=[S₀S₁…S_N-1]^T, S_k∈ { 1,2 ..., M } come indicate sub-block segmentation after sequence, wherein working as X_m,k=X_kWhen, S_k=m.Then m-th of sub-block index sequence I_m=[I_m,0I_m,1…I_m,N-1]^TIn I_m,kIt can indicate are as follows:

So m-th of sub-block can indicate are as follows:

WhereinIt is the diagonal matrix of a N × N, diagonal line is by sub-block index sequence I_mIt constitutes.So, it finally obtains Transmission signal can indicate are as follows:

(2) next. is different from PTS-Clipping algorithm and C-PTS algorithm, TS-PTS- used by this programme Clipping algorithm to the O-OFDM signal of input there is no being split processing at once, but first by initial input signal D grades of IFFT transformation are carried out, wherein 1≤d≤n-1, then carries out the division of sub-block again, system block diagram is as shown in Fig. 6.

(3) .TS-PTS-Clipping algorithm is divided into traditional n grade IFFT operation twice to carry out, and first part is Preceding d grades of IFFT operation, remaining n-d grades is second part.In preceding d grades of IFFT operation, input data sequence X_dataPass through IFFT operation forms first part's M signal sequenceThen by the M signal sequenceIt is divided into M A interim subsequence, then n-d grades of IFFT operation is carried out to this M grouping.By using PTS to calculate after two-part IFFT operation Method is chosen the signal all the way that PAPR value is minimum in the signal after being converted and is transmitted.

(4) operation of the Jing Guo (3) step, after TS-PTS algorithm, formula (5) can be indicated are as follows:

(5) carries out amplitude limiting processing to the signal after PTS algorithm, and the signal after clipping can indicate are as follows:

WhereinFor the peak value minimum signal that PTS algorithm obtains, A is clipping thresholding:

A=10CR/20 × σ (8)

Wherein CR is limited amplitude-frequency (clippingratio), and σ is average power signal, σ=E | x (t) |²}。

The beneficial effects of the present invention are:

It is found that compared to only with PTS-Clipping algorithm arrangement, the present invention program from theoretical and Computer Simulation The higher problem of computation complexity is improved to a certain extent, and compared to PTS algorithm, is had in terms of PAPR rejection It is more significant improve, although the present invention inhibits PAPR performance a bit weaker compared to PTS-Clipping algorithm, this Invention achieves good balance between computation complexity, PAPR rejection and the bit error rate.Therefore in actual application, The present invention program has very high utility value and Practical significance.

Detailed description of the invention

In order to keep the purpose of the present invention, technical scheme and beneficial effects clearer, the present invention provides following attached drawing and carries out Illustrate:

Fig. 1 is the Technology Roadmap of the present invention program；

Fig. 2 is the functional block diagram of C-PTS scheme:

Fig. 3 is the schematic diagram of Clipping scheme；

Fig. 4 is the principle flow chart that Clipping reduces PAPR；

Fig. 5 is PTS-Clipping solution principle block diagram；

Fig. 6 is that the present invention program combines improved functional block diagram；

Fig. 7 be the present invention program when sub-block number is identical with the PAPR rejection comparison diagram of other two schemes；

Fig. 8 is the bit error rate performance comparison diagram of the present invention program and other algorithms.

Specific embodiment

1. illustrating in conjunction with attached drawing 2, the main thought of traditional PTS scheme is the O-OFDM signal X that will input according to certain Mode is divided, and is divided into M non-overlapping sub-blocks, each sub-block is multiplied by same phase after being then passed through IFFT operation Location factor, one group of signal that finally minimum PAPR occurs in selection in M group signal are sent.

Traditional PTS scheme is to propose that functional block diagram is as shown in Fig. 2 jointly by S.H.Muller and J.B.Huber.

In C-SLM scheme, basic thought is the data vector X={ X to input_k, (k=0,1 ..., N-1) is pressed It is M non-overlapping subvector X according to certain model split_m=[X_m,1X_m,2…X_m,N-1], by each subvector not after The corresponding position for holding original input data symbol is padded to length N with zero.Because they are not overlapped, have:

Wherein subvector X_mIn each subcarrier multiplied by same phase factorWherein W For phase factor space size), also referred to asFor twiddle factor or weight factor, which is all to be independently actuated against every height Carrier wave is typically chosenPhase factor sequence is weighted merging to M subvector, then is had:

Wherein vectorHave after carrying out IFFT transformation to formula (2):

Wherein vectorA referred to as alternative sequence, vector x_mReferred to as x^wA fractional transmission sequence Column.The target of C-PTS scheme is to find that x can be reduced^wThe phase factor of peak-to-average force ratio combines, it is assumed that is obtained by search iteration algorithm The optimum angle factor be(signal of existing minimum PAPR value is represented in the present invention program with the variable with *), and can make In the data sequence | x'_n|²Peak value is minimum, it may be assumed that

All alternative sequences are searched for by the method for exhaustion, find so that the smallest optimum angle of O-OFDM signal PAPR because After subsequence, need optimum angle factor sequence being weighted summation, and the optimum angle sequence that will be obtained to all sub-blocks Column send receiving end to as auxiliary side information.Finally, optimal transmission sequence can indicate are as follows:

It is assumed thatIndicate the optimum angle sequence for making signal PAPR minimum in phase sets.Fig. 2 It show the functional block diagram of C-PTS scheme.

2. illustrating that Clipping is a kind of clipping class technology in conjunction with attached drawing 3, it directlys adopt nonlinear operation and believes in peak value Clipping operation is carried out near number to be effectively reduced PAPR, the signal for converting out by IFFT enters after parallel serial conversion Clipping module.If input signal a_mPeak value than preset threshold value A_maxGreatly, its original phase invariant is kept, and will The amplitude of signal is constrained to preset threshold value, if input signal a_mPeak value be not above threshold value, then this signal It remains unchanged and can pass through.Its basic principle is as shown in Figure 4.

Output signal i.e. after slicing are as follows:

Clipping process can make by the amplitude of signal be limited in preset threshold value A_maxWithin, So as to PAPR control within preset amplitude range, the index of Clipping with limit ratio (Clipping Ratio, CR it) indicates:

Wherein, A_maxFor preset threshold amplitude value, P_inThe average input work of O-OFDM signal before expression Clipping Rate.

3. illustrating in conjunction with attached drawing 4, shown in fig. 5 is the basic flow chart that Clipping process reduces PAPR, and process is adopted With nonlinear operation and realize it is relatively simple, but since Clipping process can introduce serious in-band noise and out-of-band radiation, To make signal generate distortion, spectrum efficiency is reduced.

4. illustrating in conjunction with attached drawing 5, Fig. 5 show the block diagram of PTS-Clipping algorithm.The base of PTS-Clipping algorithm This thought is first to carry out PTS processing to ofdm signal, reduces the probability for high PAPR occur, then carry out at Clipping to signal Reason, further decreases the PAPR of signal.

5. illustrate that in PTS-Clipping scheme, we combine PTS algorithm with slicing algorithm in conjunction with attached drawing 6, Reach good peak-to-average force ratio inhibitory effect, but since there are a large amount of IFFT operation, signal is carried out a large amount of during IFFT Be added with multiple multiplication, the excessively high complexity of PTS-Clipping algorithm is still its inevitable problem, so of the invention It is proposed a kind of second order PTS-Clipping (two-stage PTS-Clipping, TS-PTS-Clipping) algorithm, it will be traditional N grades of IFFT operations are divided into twice carrying out, and first part is preceding d grades of IFFT operation, and remaining n-d grades is second part, Functional block diagram is as shown in Figure 6.As shown in Figure 6 TS-PTS-Clipping scheme the specific implementation process is as follows described:

(1) will have N=2 firstⁿThe O-OFDM signal of a subcarrier indicates are as follows:

Wherein X={ X_k, (k=0,1 ..., N-1) represents the symbol sebolic addressing of input, and t represents discrete-time variable.If DefinitionWherein T_iIt is that the symmetrical matrix of a N × N represents the i-stage of IFFT operation, then formula (8) It can rewrite are as follows:

Enable S=[S₀S₁…S_N-1]^T, S_k∈ { 1,2 ..., M } come indicate sub-block segmentation after sequence, wherein working as X_m,k=X_kWhen, S_k=m.Then m-th of sub-block index sequence I_m=[I_m,0I_m,1…I_m,N-1]^TIn I_m,kIt can indicate are as follows:

So m-th of sub-block can indicate are as follows:

WhereinIt is the diagonal matrix of a N × N, diagonal line is by sub-block index sequence I_mIt constitutes.So, it finally obtains Transmission signal can indicate are as follows:

(2) next. is different from PTS-Clipping (Partial Transmit Sequence Clipping, PTS- Clipping) algorithm and C-PTS (conventional PTS, C-PTS) algorithm, TS-PTS- used by this programme Clipping (two-stage PTS-Clipping, TS-PTS-Clipping) algorithm to the O-OFDM signal of input not It is split processing at once, but initial input signal is subjected to d grades of IFFT operations first, wherein 1≤d≤n-1, then again The division of sub-block is carried out, system block diagram is as shown in Fig. 6.

(3) .TS-PTS-Clipping algorithm is divided into traditional n grade IFFT operation twice to carry out, and first part is Preceding d grades of IFFT operation, remaining n-d grades is second part.In preceding d grades of IFFT operation, input data sequence X_dataPass through IFFT operation forms first part's M signal sequenceThen by the M signal sequenceIt is divided into M A interim subsequence, then n-d grades of IFFT operation is carried out to this M grouping.By using PTS to calculate after two-part IFFT operation Method is chosen the signal all the way that PAPR value is minimum in the signal after being converted and is transmitted.

(4) operation of the Jing Guo (3) step, after TS-PTS algorithm, formula (12) can be indicated are as follows:

(5) carries out amplitude limiting processing to the signal after PTS algorithm, and the signal after clipping can indicate are as follows:

WhereinPeak value minimum signal is obtained for PTS algorithm, A is clipping thresholding:

A=10CR/20 × σ (15)

Wherein CR is limited amplitude-frequency (clipping ratio), and σ is average power signal,

It can be concluded that, step (1) first initializes O-OFDM original signal from implementation steps, and utilizes S sequence To indicate the sequence after sub-block segmentation.In step (2), the present invention program traditional n grade IFFT transformation is divided into come twice into Row, first part are preceding d grades of IFFT operation, and remaining n-d grades is second part, and convert to obtain the output of the road M by IFFT Sequence.And in step (3), guarantee going on smoothly for two-part IFFT operation by introducing M signal sequence.By The signal that step (4) chooses minimum PAPR all the way afterwards is transmitted, in order to enable system can obtain better peak in step (5) Than rejection, the signal of output is subjected to Clipping operation, its peak power is limited within the scope of threshold value, finally The signal after clipping is transmitted again.

6. combining table 1 to illustrate, since TS-PTS-Clipping algorithm is mainly in the transmitting terminal of signal to PTS-Clipping Algorithm is improved, so the part is only with the computation complexity for considering transmitting terminal.In traditional PTS algorithm, it is assumed that son Carrier number is N=2ⁿ, packet count M, phase factor number is W, using random division.When being searched for using traditional optimization algorithm, System needs the IFFT of M N point to convert altogether.And complex multiplication, the quantity minute of addition needed for carrying out the IFFT operation of a N point Not are as follows:

Therefore tradition PTS algorithm carries out the complex multiplication of N point IFFT transformation, additive amount is respectively n_add=MNlog₂N.In addition, in order to from W^M-1Optimal sequence transmission is selected in a alternative signal, it is also necessary to W^M-1(M-1) n times are added with Operation.It is hereby achieved that total complexity of tradition PTS algorithm are as follows:

Since PTS-Clipping is to carry out Clipping processing again after PTS algorithm in signal, PTS algorithm ensure that Integrality, so computation complexity is as traditional PTS method.

And total complexity of the optimization algorithm proposed are as follows:

C_{TS-PTS-Clipping}=C_d+C_n-d (18)

Wherein:

For preceding d grades of computation complexity.

For rear n-d grades of computation complexity, N'=2 in formula (20)^n-d。

For the ease of intuitive comparing calculation complexity, computation complexity reduced rate (CCRR) is defined herein and is come to calculating The reducing effect of complexity is quantified.The CCRR of TS-PTS-Clipping and PTS-Clipping two schemes is defined as:

By the definition of CCRR it is found that the more former algorithm complexity reduction of the new algorithm of the higher proposition of CCRR is then more.

The computation complexity CCRR in W=2 of table 1 is compared

Table 1 illustrates the corresponding CCRR when N and M gets representative value.From the point of view of the data presented in table, the present invention program It further reduced the computation complexity of PTS-Clipping algorithm.In addition, can also be seen that from table: (1) when M, N are constant When, with the increase of remaining series n-d, that is, the increase of the number of IFFT operation, the computation complexity of algorithm are carried out after being grouped again It can be gone up；(2) when N, n-d are constant, as the reduction of its computation complexity of the increase of M is more.Sufficiently demonstrate the present invention Scheme has better validity and superiority for computation complexity reduction.

7. illustrating that scheme is to a certain extent for PAPR's in order to further illustrate the present invention in conjunction with attached drawing 7 and attached drawing 8 Rejection has more significant raising, has carried out an experiment below.Fig. 7 and Fig. 8 is the simulation result under identical parameters.

Experiment:

The parameter setting used when 2 analogue simulation of table

Parameter	Size is set
		Sub-carrier number N	256
Maximum symbolic number	105
		Oversample factor L	4
Modulation system	QPSK
		Slicing ratio γ	4
Sub-block quantity M	4
		Partitioning scheme	Random division
Phase factor bm	{1,-1}

The simulation result of the peak-to-average ratio inhibition performance that the present invention program (TS-PTS-Clipping) is realized as shown in fig. 7, Used simulation parameter is as shown in table 2.In addition to this, since Clipping algorithm generates certain signal to O-OFDM system Distortion phenomenon causes to increase the difficulty that original signal is restored in receiving end.So also simulating PTS- simultaneously in an experiment " Hong Shanyan, Zhang Chaoyang are for dropping for details, reference can be made to document by the BER of the bit error rate (BER) of Clipping scheme and the present invention program PTS-Clipping unified algorithm [J] Circuits and Systems journal of low ofdm system peak-to-average power ratio, 2009,14 (04): 105- 110. " and document " Verma R, Tharani L.Constant modulus algorithm for PAPR reduction using PTS and clipping hybrid scheme in MIMO OFDM/A[C]//2016 International Conference on Micro-Electronics and Telecommunication Engineering(ICMETE) .IEEE,2016:337-342.”。

It can be seen from figure 7 that TS-PTS-Clipping algorithm is compared with original OFDM signal when number of sub carrier wave is identical And tradition PTS algorithm, the probability for all occurring high PAPR significantly reduce, and as the increase algorithm of remaining series n-d is to system The rejection of PAPR can be promoted further, gradually close to PTS-Clipping curve.It is 10 in CCDF^-4Place, n-d=5's TS-PTS-Clipping algorithm makes PAPR on the basis of original signal₀About 4.82dB is reduced, is promoted compared with PTS algorithm About 1.81dB, also only poor compared to PTS-Clipping scheme about 0.13dB.

As shown in Figure 8, the bit error rate (BER) performance of the present invention program is better than PTS-Clipping scheme, but will be much The BER performance being better than in Clipping scheme.It, can be with this is because partial transmission sequence (PTS) technology is non-distortion class technology Original signal is preferably recovered, and Clipping process is distortion class technology, and certain signal is produced to O-OFDM system Distortion phenomenon causes to increase the difficulty that original signal is restored in receiving end, to produce the higher bit error rate (BER).

In conclusion although the PAPR rejection of the present invention program is slightly poorer than PTS-Clipping algorithm, with The increase of remaining series n-d its PAPR rejection will be further to PTS-Clipping algorithm, and in remaining series n-d =5, packet count M=4, sub-carrier number can be reduced 71.18% computation complexity when being 256.Simultaneously it can also be seen that the present invention Scheme is also a kind of peak-to-average force ratio Restrain measurement that can take into account peak-to-average force ratio (PAPR) inhibitory effect and error rate of system (BER).

Finally, it should be noted that the above preferred simulation example is only used to illustrate the technical scheme of the present invention and not to limit it, Although the present invention has been described in detail by above-mentioned preferred simulation example, those skilled in the art should be managed Solution, can make various changes to it in the form and details, be limited without departing from claims of the present invention Range.

Claims (2)

1. the present invention relates to a kind of smooth orthogonal frequency division multiplexing (Optical Orthogonal Frequency Division Multiplexing, O-OFDM) system peak-to-average force ratio Restrain measurement, in particular to the improved low complex degree TS- of a kind of joint The peak PTS-Clipping (two-stage Partial Transmit Sequence-Clipping, TS-PTS-Clipping) Compare Restrain measurement.The program is mainly by being based on partial transmission sequence (Partial Transmit Sequence, PTS) skill Art and clipping (clipping) technical tie-up improve to realize the inhibition to high peak-to-average power ratio.In low complex degree scheme, O-OFDM Initial input signal is carried out d grades of IFFT (Inverse Fast first there is no being split at once by signal Fourier Transform, IFFT) operation, then the division of sub-block is carried out, remaining n-d then is carried out to the signal after piecemeal Grade IFFT operation, then utilizes PTS algorithm search best match phase factor and the smallest signal all the way of PAPR, in conjunction with Clipping technology by signal limiting within the scope of threshold value, to obtain optimal peak-to-average ratio inhibition performance.

2. scheme according to claim 1, the present invention is comprised the steps of:

(1) will have N=2 firstⁿThe O-OFDM signal of a subcarrier indicates are as follows:

Wherein X={ X_k, (k=0,1 ..., N-1) represents the symbol sebolic addressing of input, and t represents discrete-time variable.If definitionWherein T_iIt is that the symmetrical matrix of a N × N represents the i-stage of IFFT operation, then formula (1) can be with It rewrites are as follows:

Enable S=[S₀S₁…S_N-1]^T, S_k∈ { 1,2 ..., M } come indicate sub-block segmentation after sequence, wherein working as X_m,k=X_kWhen, S_k= m.Then m-th of sub-block index sequence I_m=[I_m,0I_m,1…I_m,N-1]^TIn I_m,kIt can indicate are as follows:

So m-th of sub-block can indicate are as follows:

WhereinIt is the diagonal matrix of a N × N, diagonal line is by sub-block index sequence I_mIt constitutes.So, the transmission finally obtained Signal can indicate are as follows:

(2) next. is different from PTS-Clipping (Particle Swarm Optimization-Clipping, PTS- Clipping) algorithm and C-PTS (conventional PTS, C-PTS) algorithm, TS-PTS- used by this programme Clipping (two stage PTS-Clipping, TS-PTS-Clipping) algorithm to the O-OFDM signal of input not It is split processing at once, but initial input signal is subjected to d grades of IFFT transformation first, wherein 1≤d≤n-1, then again The division of sub-block is carried out, system block diagram is as shown in Fig. 6.

(3) .TS-PTS-Clipping algorithm is divided into traditional n grade IFFT operation twice to carry out, and first part is IFFT fortune Preceding d grades calculated, remaining n-d grades is second part.In preceding d grades of IFFT operation, input data sequence X_dataIt is transported by IFFT Calculation forms first part's M signal sequenceThen by the M signal sequenceIt is divided into M a interim Subsequence, then n-d grades of IFFT operation is carried out to this M grouping.By using PTS algorithm, choosing after two-part IFFT operation The signal all the way for taking PAPR in the signal after being converted (peak-to-average power ratio, PAPR) value minimum carries out Transmission.

(4) operation of the Jing Guo (3) step, after TS-PTS algorithm, formula (5) can be indicated are as follows:

(5) carries out amplitude limiting processing to the signal after PTS algorithm, and the signal after clipping can indicate are as follows:

WhereinFor the peak value minimum signal that PTS algorithm obtains, A is clipping thresholding:

A=10CR/20 × σ (8)

Wherein CR is limited amplitude-frequency (clipping ratio), and σ is average power signal, σ=E | x (t) |²}。