TWI269546B - Method and system for impulse noise rejection in OFDM system - Google Patents

Abstract

An impulse noise rejection system for an orthogonal frequency division multiplexing (OFDM) communications system and related methods are disclosed. The impulse noise rejection system detects presence of impulse noise in an OFDM symbol in time domain, and compensates for the impulse noise in frequency domain. Impulse noise corrupted OFDM symbol is ""down-converted"" such that the impulse noise component of the FFT output is a slow-varying waveform. Impulse noise at all subcarriers is estimated by interpolating the estimated impulse noise at pilot subcarriers in the slow-varying waveform. Impulse noise free FFT output is obtained using the estimated impulse noise and an ""up-converting"" process. The apparatus and algorithm for optimizing the down-converting frequency and up-converting frequency are described.

Description

1269546 IX. Description of the invention: [Technical field to which the invention pertains] The present invention relates to an orthogonal frequency division multiplexing (Orthogonal 1 .

Frequency Division Multiplexing (OFDM) communication system method and system thereof, especially a method and system for eliminating pulse noise in an orthogonal frequency division multiplexing communication system. [Prior Art] Orthogonal frequency division multiplexing technology is a popular multi-carrier modulation technology, which has been widely used in many communication standards today, such as: Digital Audio Broadcasting (DAB), digital video broadcasting. Digital Video broadcasting (DVB), wireless local area network, and home-friendly network (HomePlug). The modern orthogonal frequency division multiplexing communication system has the characteristics of plasticity and reliability in channel selection, and has a low carrier-to-noise ratio (C/N ratio). Due to the orthogonal nature between the subcarriers, each subcarrier can be used to employ different transmission rates depending on the transmission environment of the channel in which it is used. In addition, the channel's encoding process often combines the orthogonal frequency division multiplexing modulation mechanism to achieve error correction. Therefore, the Orthogonal Frequency Division Multiplexing technique can be used in extremely harsh communication environments. 1269546 However, the 'orthogonal frequency division multi-m system is particularly susceptible to pulse noise (four), and the so-called pulse noise refers to the high-energy interference signal generated in a short period of time, and the cause of the pulse noise. Usually the odorous equipment passes through electrical equipment (such as truck electrical equipment or household appliances). High-pulse pulse noise is likely to cause the receiver to turn _//bit: the occurrence of a clip or saturation ί / /, pulse noise in the frequency domain = wide bandwidth (wide- Band) of noise, so pulse noise will affect the secondary carrier in all channels and reduce the communication quality of communication equipment. In general, the effect of the analog/digital converter due to pinching is neglected, so the pinching phenomenon of the analog/digital converter is usually not discussed. In European Patent Nos. 1,1,1 and 1,043,874, Dawkins et al. and Haffenden et al. disclose the signals input in the time domain to achieve analog/digital conversion. Zero out is the purpose of the output sample that is affected by the pulse noise. Please refer to Fig. 1'. Fig. 1 shows an orthogonal frequency division multiplexing receiver with a blank input signal function in the time domain. The analog signal input is converted into a digital signal by a type of digital/digital converter 110. If the pulse noise is detected in the input signal, the blank module 120 will send the affected sample to the time domain before transmitting the signal to the Fast Fourier Transform (FFT) module 140. 7 1269546 The processor 130 is such that the operation of the subsequent frequency domain (freqUenCy domain) processor 150 is not corrupted by samples affected by pulse noise. However, the performance of these methods is greatly limited by the fact that they do not employ well-known transmission signals, for example, circuits that transmit signals by pilot subcarriers are not used. Another type of prior art provides some way to eliminate pulsed noise. For example, Jukka is in WIP0 Patent Application No. WO03073683, Asjati is in British Patent No. GB2388500, and Sliskovic is published in July 1995. "7 (IEEE Transactions on Signal Processing) Volume 43 (Volume) Theme 7 (lssue) # 1651~1662 "Signal Processing Algorithm for OFDM Channel with

In the article Impulse Noise, the method of effectively eliminating pulse noise in the frequency domain by guiding the subcarrier is disclosed. In these prior art, the technique of guiding the subcarrier is used to estimate the pulse noise. The estimated pulse noise is transmitted to the time domain for cancellation, and the signal receiver implemented by this method is shown in Figure 2. The noise estimator 280 is guided by the subcarrier technique. To estimate the pulse noise, the estimated miscellaneous A is sent to the Inverse Fast Fourier Transfonn (IFFT) module 270 to make it the time domain §fl number via the inverse fast Fourier transform. A noise cancellation module is used to estimate the noise 1269546: reverse fast Fourier transform from the signal in the blank module 22 。. The noise cancellation mode, the signal of the group 260 will be sent again To a fast Fourier transform module 290 for fast Fourier transform, the output signal of the fast Fourier transform module 29G is further forwarded to the processor 250 for subsequent processing. _, the conventional method of this class Additional hardware and software are required to cope with complex equations. Therefore, communication devices using these methods often have problems with signal extension. [Invention] Therefore, the object of the present invention is to provide an algorithm for eliminating orthogonality. A system and method for pulse noise in a frequency division multiplexing communication system to improve the problems in the prior art described above.

I ... Briefly, the present invention discloses a pulse noise cancellation system for a quadrature frequency division multiplexing system. The pulse noise cancellation system includes a pulse noise signal that is lightly connected to an input end of the pulse noise cancellation system, and a fast Fourier transform mode of the input end of the pulse noise cancellation function and a frequency domain pulse. Noise cancellation module. The pulsed noise symbol, J疋 is used to detect the signal conversion code (wheel) it is in, and the fast Fourier transform __ is used to generate the fast Fourier transform. The frequency domain pulse hybrid A > Xiao divide_group includes _ _ connected to the fast Fourier transform module

a first processing unit of 1269546, an interposer „p 〃 °r coupled to the first processing unit, and a relay unit coupled to the fast Fourier transform module and the second processing unit. The _ processing unit is configured according to the Positive ^^^^^(predicted pil〇t ^ followed by (four), self-mixed Fourier transform module connected (four) signal and: insert (four), to the new 1 waveform (siw but ngwavef_) in the estimated pulse noise The interpolator is configured to insert the estimated pulse noise in the sub-carrier and the sub-carrier to obtain the prediction pulse of all the sub-carriers in the gradual waveform. Miscellaneous δ Κ. 弟二虚: 辟 - 八八相, - 兀 is used to refer to the description symbol and the positive:: all the subcarriers within the hybrid waveform's estimate of the 'fast-varying waveform' : The human carrier does not contain the signal of the estimated pulse noise. The pulse-type (10) orthogonal noise division system of the pulse noise cancellation Γ pulse noise cancellation system includes a signal exchange code connected to the system The input end of φ is used to make the pulse noise to cancel the pulse of the pulse noise connected to the pulse The noise detector, a cyclic shifter that is connected to the input end of the cyclic displacement system, and a light-leaf conversion module are used to generate fast Fourier transform type fast repeat and one frequency domain pulse noise cancellation. When the pulse 1269546 noise detector detects the pulse noise, the cyclic shifter will cyclically shift the distance - the first distance of the tiger to convert the stone horse. The frequency domain pulse noise elimination / ^ And a first processing unit, an interpolator, and a second processing unit are connected to the fast Fourier transform module for using the expected steering times of the orthogonal frequency division multiplexing communication system. The carrier, the signal received from the fast Fourier transform module, and the - description symbol are used to generate the estimated pulse noise in the guided subcarrier of the slow and new waveform. The plug-in is switched to the second fourth unit. The guided sub-skin for inserting the gradual waveform: predicting pulse noise to obtain the orthogonal frequency division multiplexing communication system =:!=: estimated pulse noise of all subcarriers in the picoform. The second place is poor. (8) Fast Fourier The conversion module and the inserter are configured to use the description to turn the missing number, and the orthogonal frequency division, and the estimated pulse hybrid communication system of the signal from the fast Fourier transform module All carriers in the variable waveform do not contain the signal of all the top estimated impulse noises generated in a sharply varying waveform. - The present invention also discloses the use of the signal conversion to eliminate the orthogonal frequency division. The multiplex communication system has the following methods: detecting the orthogonal sub-range = noise. The method includes no pulse noise; the multiplex communication communication system - the signal conversion code is the escaped fast Fourier ^ The signal conversion and the type of the crossover multiplex communication system are based on the orthogonal frequency division multiplexing communication system 1269546: pre (four) to subcarrier, self-fast Fourier transform enchantment reception (four) =: description symbol 'to generate - predicting pulse noise in the guided subcarrier of the slowly varying waveform; and inserting the estimated impulse noise in the guided subcarrier of the graded waveform to obtain the orthogonal frequency division multi-m system within the graded waveform ^All the rest of the hanging pulse 4 and according to the description The symbols and Orthogonal Frequency Division Multiplexing (iv) correspond to the Yu the estimated pulse noise grading of all sub-carriers within the waveform, to be generated - of all sub-carriers within the blast wave signal which does not contain all of the estimated impulse noise. [Embodiment] The present invention discloses a method for performing (b) side and eliminating pulse noise in an orthogonal frequency division multi-predicate system by an efficient (4) single-t-channel structure. The technique of this sub-carrier will be used, and the signal will be down-converted to a gradual waveform before the insertion process.

Uc^vatyingw—M. The method has a monthly b-force for highly pulsed noise and the system of the present invention has a cleaner hardware than the prior art, which has less m consumption. In addition, by the present invention, the problem of delay in signal processing can be avoided by the pulse noise in the Orthodox/Divided-Frequency Communication System. Fig. 3 is a block diagram of the pulse canceling system of the first embodiment of the present invention. The analog input signal is converted into a digital signal by an analog/digital converter (ADC) 310, and the received signal containing Gaussian noise can be expressed by the following equation (1) : r (W) = x(w) ® Λ(«) + ν(«), π e {1,...,TV} (1) where 17 is the sampling index X(8) used to indicate that the transmission signal is 0) Indicates the channel impulse response v (both used to indicate that Gaussian noise # is used to indicate the size of the fast Fourier transform type. ® is used to indicate the convolution operation. When there is pulse noise, it is expressed by the following equation (2): The signal can be rin) + ρ{ή) in the time domain (7) where ρΟΟ is used to represent the pulse noise. 13 1269546 - The fast Fourier transform module 340 is used to reduce the time domain sampling of the analog/digital converter 310 into a Laicheng, minus _+ coffee. The corresponding signal in the frequency domain can be expressed as ·· m +p(k) is , where R(k) = X(k)·H(k) + V{k\ ks{\9..^N} and k is I (nine) to indicate that Han (4) is used to represent the factory (eight) The size of the noise fast Fourier transform (FFT) used to represent the channel gain subcarrier k of the transmission aggregation point subcarrier k of the subcarrier index subcarrier k, and the signal conversion code output by the analog/digital converter 310. (symbol) will be transmitted to a pulse noise detector 320 to detect whether or not there is pulse noise. When the pulse noise is detected from the signal conversion code outputted by the analog/digital converter 310, the signal [Λ(9) + P(1)] output from the fast Fourier transform module 340 is transmitted to a frequency domain. Pulse noise cancellation module 14 1269546 350. The frequency domain pulse noise cancellation module 350 includes a first processing unit 352 coupled to the fast Fourier transform module 340, an interposer 354 coupled to the first processing unit 352, and a coupled to the fast Fourier transform module. Group 340 and second processing unit 356 of plug 354. The frequency domain signal + corpse (4) in the pilot subcarriers will be extracted and according to the predicted pilot subcarriers of the orthogonal frequency division multiplexing communication system and the received fast Fourier transform mode. The signal of group 340 is generated from the predicted pulse noise of the guided subcarrier in a slowly varying waveform. The expected guided subcarrier can be understood by the pilot values and the estimated/expected channel gain, and can be expressed by the following equation (5). · Left (6)|k{directed subcarrier} (5) Impulse noise Eight illusions are not known to the receiver except for the guided subcarriers. Therefore, the pulse noise in the guided subcarrier can be estimated by the following equation (6): Acoustic (moxibustion) = and (4) + iW - left (moxibustion), moxibustion e to subcarrier} (6) In order to obtain all times For pulse noise in the carrier, the estimated impulse noise needs to be inserted into the guided subcarrier [sound (6) oriented subcarrier}]. However, when the waveform 15 1269546 is severely anxious, the above-mentioned insertion action cannot be performed very accurately. Therefore, in the present invention, the first processing unit 352 performs a down conversion operation to down-convert the estimated impulse noise in the phantom, & e{directed subcarrier}] of the guided subcarrier to a gradual change. Among the waveforms·· ή/(10)(幻丨户W·厂, ex-e-subcarriers} (7) 必庄思疋 The above parameter d is a descriptor, which is based on some algorithms. To provide a suitable frequency to predict the pulse noise in the subcarrier for down-conversion, and there are many algorithms for generating and optimizing the descriptor d. Converted to a slowly varying waveform The estimated pulse noise iAi) in the guided subcarrier is turned into the interpolator 354 for interpolation processing, and then the orthogonal frequency division multiplexing is obtained in the above-mentioned slowly varying waveform. Predictive pulse noise in all subcarriers of the communication system: Insertion processing {household-(8) acupuncture e-guided subcarrier} => (8) acupuncture and moxibustion has a secondary carrier} (8) Obtained in the slowly varying waveform The orthogonal frequency division multiplexing communication system is full of 16 1269546, and the estimated pulse noise in the subcarriers will be Unit to the second processing unit 356. The second processing unit unit 356 performs a subtraction operation and an up-conversion conversion, wherein the up-conversion conversion is performed with respect to the down-conversion performed by the previous first processing unit unit 352, and after the subtraction and the up-conversion, It will generate a signal with no pulse noise in each subcarrier: .)))) + corpse (8) one (4), moxibustion G {all subcarriers} (9)

Hk) I pshw(k) · e+j2M/\ ke The frequency domain pulse noise cancellation module can be implemented by any means, as long as the first processing unit can be based on the prescaling of the orthogonal frequency division multiplexing communication system. a period-oriented subcarrier and a signal received from the fast Fourier transform module to generate an estimated pulse noise in the guided subcarrier of the -graded waveform; and the second processing unit is capable of depending on the orthogonality in the graded waveform The estimated pulse noise in all subcarriers of the frequency multiplex communication system generates a signal with no pulse noise in each of the fast-varying wavefom〇 sub-carriers. Please refer to Figure 4, Figure 4 The first processing unit 452 of the frequency domain pulse noise cancellation module 450 includes a 4521 minus 17 1269546 2 for the first embodiment of the present invention. And a converter 4522. The subtractor 4521 is configured to subtract the received signal from the subcarrier to the subcarrier (> PW|k{ to the subcarrier}] to obtain the guided subcarrier. Estimation, pulse chowder (10)), λs pass Face], as shown in Equation ⑹. Leading subcarrier. The = estimated pulse noise will be input to the converter Mu to convert to the slowly changing waveform according to the descriptor 旒d. The interpolator 454 will insert the estimated impulse noise in the subcarrier into the estimated pulse noise in all subcarriers in the gradual waveform.

/V { /mv(Λ) I moxibustion e-guided sub-running tour} f - , 7... ^ 守丨"人戦姒=> {K illusion I k all subcarriers} The first processing unit 456 includes a converter 4561 and a subtractor _ 4562. The converter 4561 converts the estimated pulse noise [{for _ (magic moxibustion all subcarriers)] in all subcarriers in the gradual waveform into all the blast waveforms according to the same descriptive symbol d. Estimated pulse noise in the subcarrier [{female] 14 (magic.f all subcarriers}]. The estimated pulse noise in all subcarriers within the jerk waveform [{fine I te all subcarriers}] will be The received signal and the guided subcarrier in all subcarriers} are extracted to obtain signals without pulse noise in each subcarrier within the sharp waveform: 1269546 In addition to the structure shown in Fig. 3, the present invention The design of the frequency domain pulse noise cancellation module can be slightly changed. Please refer to FIG. 5, which is a block diagram of the frequency domain pulse noise cancellation module 550 according to the second embodiment of the present invention. The first processing unit 552 of the signal cancellation module 550 includes a converter 5521, a subtractor 5522, and a converter 5523. The converter 5521 is configured to send the received signal [{雄)+)|& Carrier}] downconverts to ramping waveform + all subcarriers}], while converter 5523 is used to The period-oriented subcarrier [" (phantom I k {guided subcarrier}] is down-converted into the following slowly varying waveforms: 夂W(4)Θ W f2//E E {all subcarriers} 〇subtractor 5522 self-grading waveform The received signal in the guided subcarrier in the slowly varying waveform extracted from the received signal in all the subcarriers is deducted from the expected steering subcarrier [foot-oriented subcarrier] in the slowly varying waveform to obtain the slowly varying waveform. The estimated impulse noise in the guided subcarrier [is w(A:), k is transmitted to the secondary carrier}]. The estimated impulse noise (8) in the guided subcarrier within the graded waveform is directed to the secondary carrier}] It is sent to the interpolator 554, so that it is inserted into the estimated pulse noise in all the subcarriers in the ramp waveform: 19 1269546 Insertion processing {4^ 幻 I ^ Guided subcarrier} 2 {4^(幻I ^All subcarriers} The second processing unit 556 includes a subtractor 5561 and a converter 5562. The subtractor 5561 is subtracted from the downconverted signal [{U magic + all subcarriers}] by the installer 554. Signal to obtain no pulse noise in each subcarrier within the ramp waveform: Κ Λο〇^) I rs1〇ww+Psi〇^)-psi〇,{k\ {kem^xmm ° The converter 5562 will follow the same descriptive d, and there will be no pulse noise in each subcarrier in the ramp waveform. The signal is upconverted to a signal with no pulse noise in each subcarrier in the turbulent waveform: 永幻丨先 w W f Ά e {all subcarriers}. As shown in FIG. 4 and FIG. 5, one of the innovations of the present invention is that the signal is down-converted into a slowly varying waveform, so that the signal insertion operation for estimating the impulse noise in all subcarriers is obtained. It can be more efficient and efficient than the previous technology implementation 20 1269546. After the signal is inserted, the signal that was previously down-converted will be upconverted and returned to the original frequency to obtain the signal without pulse noise in each carrier. Therefore, any structure or algorithm that achieves the above concepts may be the subject of the invention. In the frequency domain pulse noise cancellation module of the present invention, the description of the mode number d is used to determine the switching frequency at the up-conversion and down-conversion. Describe the model number d is a fixed value that is dynamically generated or preset according to the calculation method to achieve the purpose of optimization. In the present invention, an Impulse Noise Rejection (INR) controller will be proposed in the following description to optimally produce a description of the model number d. Please refer to FIG. 6. FIG. 6 is a block diagram of a pulse noise cancellation system according to a second embodiment of the present invention. The pulse noise cancellation system includes a -pulse noise cancellation controller (4), and the #pulse noise φ is pulsed noise detector 62 (M贞 is detected, the pulse noise cancellation control (4) will rely on the pulse miscellaneous The position of the signal is located in the signal conversion code to generate: The symbol d is used to optimize the noise estimation action. The field is called the 7~1 map, and the 7th to 10th lines are used to indicate the signal conversion. The four different conditions in the clever position, and below; ^ will use these illustrations to illustrate how the pulse noise cancellation controller: Generate the trace number d. Here we follow the previously assumed, orthogonal The 1207546 frequency conversion code contains the output samples of N analog/digital converters. Please refer to Figure 7 first, and Figure 7 shows the first case: in the signal conversion code 700, only the occurrence order A primary impulse event 72, where Ns represents the start position of the event and Ne represents the end position of the event. In this case, the pulse noise cancellation controller may choose to set the description of the payout d to [(Ns+Ne)/2], that is, set to this pulse Event 7 2 midpoint, and when the frequency domain pulse noise cancellation secret uses this description to pay the number, the pulse noise will be converted into DC form. Please refer to Figure 8 and Figure 8 for the second case. There are two pulse noise events, μ, 84 are generated in the signal conversion code 8〇〇. The distance between the two pulse noise events and the material D1 is less than half of the total length of the signal conversion code, so D1 j In the second case, the initial pulse position of the preceding pulse noise event u (4) is Ns and the end position of the subsequent pulse noise event 84 = Ne' is the pulse noise cancellation of the present invention. The controller will set the descriptive symbol d to [(Ns+Ne)/2], which is set to be the midpoint of the two-pulse noise events 82, 84. Please refer to Figure 9, Figure 9 shows In the third case, the rain pulse noise events 92, 94 occur in the signal conversion code _. The pitch D2 of the p-bend λα v 3 of the two-pulse noise events 92, 94 is greater than the signal conversion code 8 〇〇 The total length is -half, so D2 is greater than. In this third case, the front pulse noise magpie ^ ^ i , , u beam position is marked as Ne The starting position on the crepe of the subsequent pulse noise event 94 is denoted as Ns, and the pulse noise 22 1269546 of the present invention eliminates the controller to set the descriptive symbol d to [(Ns+Ne)/2]. Figure 10, Figure 10 shows the fourth case: more than two pulsed noise events 102, 104, 106 occur in the signal conversion code 1〇〇〇. Between every two adjacent pulse noise events The spacing is D3, D4, D5, where D5=D51+D52. The position of !^ and Ne is based on the maximum spacing of the a and 疋, taking the first map as an example, because the maximum spacing is D4, Therefore, Ns is set to the start position of the pulse event 1〇6, and the Bay 6 is set to the end position of the pulse event 104. In this case, the pulse noise canceling controller of the present invention also sets the descriptive symbol d to [(Ns + Ne)/2]. In the case of Figures 9 and 10, the pulse noise cancellation controller cyclically processes the samples in the signal conversion code. The following is an example of a stylized syntax to illustrate how to determine the initial set Ns and the end position Ne in the pulse event in the orthogonal frequency division multiplexing signal conversion code: // Sample_counter: number of samples // // Computation done ··Complete the calculation index // // impulse event : pulse event // //Start: start position; End: end position // 23 1269546

Sample—counter = 0;

While (Computation done = false) { //When the calculation is not completed //

If (Impulse noise event detected) { //Detected a pulse noise event //

If (First impulse event detected) { //Detected the first pulse event //

Ns = Start of the first impulse event. //Ns is the starting position of the first pulse event // //Ne is the end position of the first pulse event //} Else {

If (D>N/2) { //The spacing D is greater than N/2 //

Ns = Start of the new impulse event· 24 1269546

Ne = End + Ν·

Computation done = true. //Ns is the starting position of the new pulse event" // Ne = Ne + N // } Else {

Ne ~ End of the new impulse event. //Ne is the end of the new pulse event //

Sample—counter = Sample—counter + 1; If (Sample—counter is N)

Computation done = true; The structure and the use of the pulse noise detector can also have different algorithms for detecting the pulse jamming in the signal conversion code. Figure 11 is a A block diagram of the pulse noise detector 11 〇〇 25 1269546 in the embodiment. The pulse noise detector 1100 includes an absolute value evaluator 1120, a saturation detector 1140, a buffer 116 〇, and a saturation calculation. The 1180° absolute value evaluator 1120 is used to calculate the input signal. The conversion code, the absolute value of r is 丨rl, and the saturation detector 1140 is coupled to the absolute value finder 112 for determining the output Sn according to the input absolute value |r| When the absolute value |r| is greater than a threshold, it indicates that the input signal conversion code of the input is saturated, and the saturation detector 1140 generates a first value; and when the absolute value |r| When not greater than the threshold, the saturation detector 1140 generates a second value. For example, the first value described above can be selected to be 1, and the second value can be selected as:

1, I called >=critical 値 〇, other (11) The buffer 1160 is coupled to the saturation detector 114A for buffering the value data from the saturation detection 1140. The saturation calculator is lightly coupled to the buffer 1160 for calculating the sum of the consecutive Q data generated by the saturation detector 114A. If the sum of the consecutive Q values is equal to p, that is, if at least p saturated samples are included in the consecutive Q samples, the pulse noise detector 11 will detect the pulse noise. The adder 1182 adds the value output by the saturated spare 1140 to the value of the wire and subtracts the value output by the buffer H60. For example, suppose Q is equal to 7 and p is equal to $, 26 1269546 ^ is equal to G, cumulative buffer (four) 84 is temporarily stored as a cumulative sum of 4, and heart, and i^sn are added to the accumulation buffer 1184 for temporary storage. After the accumulation and deduction of the V7 self-accumulation buffer U84, the pulse noise detector 1100 detects the pulse noise. The threshold used by the saturation controller can be a predetermined fixed value, or the value of the signal used in the case of signal detection assistance is the threshold. Please refer to FIG. 12, which is a block diagram of a saturation detecting 1200 according to an embodiment of the present invention. The saturation detector 12A includes a packet detector 1210 for extracting a packet input into the signal of the saturation detector 1200, that is, a packet for detecting the absolute value of the rounded |r| 'To produce a suitable unbounded value. The saturation detector 1200 in this embodiment further includes a multiplexer 1220 for selectively detecting the threshold value of the saturation state. A predetermined threshold or a suitable threshold generated by the packet of the absolute value |r| can be selected to determine whether each sample is saturated. The saturation detecting unit 1230 outputs Sn according to the saturation condition of the sample. In the present invention, a selective module can be utilized to reduce the adverse effects of pulse noise. Please refer to FIG. 13, which is a block diagram of the fast Fourier transform module 1300 of the present invention. Among the fast Fourier transform modules 27 1269546 1300, a pulse noise canceller 1320 is used to previously reduce the power of the pulse balance noise, and a fast Fourier transform unit 134 is used to represent the function block of the fast Fourier transform. . In this embodiment, the input signal conversion code is buffered via the buffer 1322. In the case where no pulse noise is detected, the multiplexer 1326 outputs a signal conversion code from the buffer 1322. However, when the pulse noise detector detects the pulsed noise, the data stored in the buffer 1322 is replaced by the value generated by the substitute sample generator 1324. In a special case, the alternate sample generator 1324 will output zero; in another special case, the substitute sample generator 1324 will output the average of the first few samples. With the above embodiments, the present invention can have the following advantages: the samples output by the analog/digital converter can be automatically and effectively suppressed before or after the detected pulse noise sampling, and these samples are automatically suppressed. May contain undetectable pulse noise energy. As described above, the advantages brought by the present invention and the contribution to the technical field are that the frequency chain is down-converted to a slowly changing waveform before the insertion processing procedure, so that the interpolation processing can be performed efficiently, and at all A signal with no pulse noise is obtained on the secondary carrier. However, this effect can also be achieved by periodically shifting the sampling in the time domain. Please refer to FIG. 14, which is a block diagram of a pulse noise canceling system 14A according to a third embodiment of the present invention. A pulse of miscellaneous detection of the 1420 light connected to a class / digital conversion cry 141 〇 28 1269546 = __ ratio / digital converter 141 () in the signal conversion code in the escape code two t, noise _1420 Detect The signal-to-pulse J-pulse miscellaneous control (4) is just generated according to the position of the signal in the signal conversion code to describe the symbol d, and the estimation process of the signal is optimized. The pulse noise cancellation controller just like the (4) algorithm can also be as described in the previous paragraph ~ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Before the fast Fourier (four) change was made, the Fu group 1440 was transferred to the genus. The cyclic shifter will cyclically shift the sampling of the orthogonal frequency division multiplexing signal to a first distance according to the descriptive symbol d' generated by the pulse noise cancellation control 43G, whereby the midpoint of the pulse will be corrected The first-distance of the above-mentioned first-distance of the cross-frequency __ is determined according to the descriptive symbol d. The signal processed by the cyclic shift can be expressed as ·· rsloM) + PsloM) (12) The step is changed by fast Fourier and the signal processed by the cyclic shift will be re-introduced - after the processor 1440 is processed: 29 (13) 1269546

Rsl〇w(k) + Psi〇w(k) This cyclic shift processing is performed, and the fast Fourier transform in the forward guided wave is transmitted to the frequency domain pulse. Module 1450. The frequency domain pulse noise cancellation 杈 group 1450 of the first ~ Lu Yi oil to the winter unit 14 ^ 2 according to the expected guided subcarrier to the subcarrier }], and the description symbol generated by the pulse noise cancellation controller 1430 ... 泸, π 理 导向 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前 前In the shape of the first processing unit, in order to slow down the pulse of the 2 wave towel, the pulse is inserted into the slow-changing wave carrier, and all the interpolation processing is performed.

AaDlk all subcarriers} &_ (the illusion A e-directed subcarriers) - the second processing unit 1456 is connected to the fast Fourier transform module and the inserter for the signal from the fast Fourier transform module The gradual change waveform _ all the touch _ estimate _ noise and the description symbol d, to generate the tremor waveform _ all times less than the signal of the estimated impulse noise. Therefore, in the above way, in the time domain The pulse noise cancellation system and the method thereof for the cyclic displacement sampling in advance 30 1269546 can realize 0. The frequency domain pulse noise cancellation module of the pulse noise cancellation system of the third embodiment of the present invention basically has many differences. The structure can be used to implement the frequency domain pulse noise cancellation module, and the implementation condition is as long as the first processing module can be based on the expected guided subcarrier and the fast Fourier transform in the orthogonal frequency division multiplexing communication system. The signal received by the module is used to generate the predicted pulse noise of the guided subcarrier in the gradual change waveform, and the second processing right group can be used for the gradual change wave according to the orthogonal frequency division multiplexing communication system All the predicted pulse noises in the subcarriers are generated to generate signals without pulse noise in all the subcarriers of the turbulent waveform. Please refer to Fig. 15, which is the frequency of the third embodiment of the present invention. A block diagram of the domain pulse noise cancellation module (10) 10. The first processing of the frequency domain pulse noise cancellation module 1550 includes a converter 15521 and a subtractor 15522, and the 15521 system is used for pulse noise. Eliminate the generation, time, and d generated by the controller to reduce the expected preamble subcarrier_丨(4) leading subcarrier down-conversion to #u. In the waveform: 谀 to a slowly varying 4 shoulder thin)·〆·, such as the preamble _ (10) 31 1269546 The step is subtracted by the subtractor to deduct the expected preamble subcarrier in the slow-changing waveform from the received signal from the preamble subcarrier, and the estimated pulse in the preamble subcarrier is obtained in the waveform. Noise: Μ 变

Psio^ik) I Rslow{k) + Pslow(k) - Rsl〇w(kl ke ( χ 5) Inserting Is 1554 will insert the slowly-changing waveform (4) into the subcarrier of the subcarrier to all the slowly varying waveforms. Attack ★ & Estimated Predictive Impulse Noise in Pulse Miscellaneous Carrier: Female Plug-In Processing-Oriented Subcarrier} {Psi〇Jk)\k: All Secondary Carriers} The second processing unit 1556 includes one, 15562. The subtracter 15561 will obtain the sub-cafe from all the sub-carriers from the waveform of the Xuan Xian 1" 15561 and the converter with the estimated pulse in the sub-carrier to subtract the slow-changing waveform. After slowly changing the signal from the pulseless noise·· WI ^W+Pshw(k) -PslUk) k - , the converter 15562 will be on the slowly changing shape according to the same descriptive symbol d of the phase η wave# The estimated impulse noise in all subcarriers that are converted to the sharp-changing waveform in all the signals from the no-pulse noise up/down 32 1269546 (17) {m | Rsl0^ye+jk2^ \ kg

The previously described pulse noise canceller can incorporate the pulsed noise cancellation system of the present invention and include a cyclic shift p. The last name will be i^ whistle reference 16, and Fig. 16 is a block diagram of the pulse noise canceling system 1_ of the fourth embodiment of the present invention. The pulse noise cancellation system 16 is a white person ^ ^ 匕3 has a pulse noise canceller 1662' and its function and structure are the same as those of the pulse noise canceller 1320 of the 13th. The analog/digital converter 161 汴 出 出 sample will be transmitted to the pulse noise canceller 1662. When the pulse noise is detected by the pulse noise detector 1620, the pulse noise month 1662 can replace the original sample with the substitute sample, so that the pulse is mixed. The negative impact of high energy will be eliminated in advance. &', in this embodiment, the function and device for cyclic displacement sampling is performed by the U.S. ring displacement unit unit 1664 (4)_. (4) Displacement Displacement is deleted. The core is moved, and the loop is formed. The orthogonal frequency division of the present invention is a7. The performance and performance of the pulse noise cancellation in the y, k-signal system are verified by the actual and non-tested. Please refer to Figure 17 18, Figure 17 for the unexamined mail & ^ V pulse noise cancellation processed signal 33 1269546 Figure 18 is the signal distribution processed by the pulse noise cancellation method and device of the present invention Figure, in which the received signal frequency (DVB-T 2K mode 64 QAM) signal noise ratio (5) (four) -to-Noise Ratio, SNR) is 30dB ' and two of them are affected by pulse noise . It is apparent that the method of eliminating pulse noise of the present invention and its system have succeeded in eliminating the effects of pulse hybridization and greatly improving the performance of the receiver. • Compared with the prior art, the present invention discloses a pulse noise cancellation system for use in a positive-family crossover multi-wei communication system based on a guided subcarrier, first receiving a pirate, and a pulse noise. Can be extracted and has a ground suppression in the time domain and compensated in the frequency domain. In the frequency domain pulsed MIMO group of the present invention, the 'pulse noise is first estimated in the guided subcarrier and then down-converted into a gradual waveform. An interpolator is used to pre-estimate the pulse noise in the non-transient-human carrier, so that the waveform processed by the interpolation can be used to eliminate the pulse noise. In addition, in order to implement the present invention, the complexity of the circuit required is zero, and its power consumption is lower than that of the prior art. Therefore, the present invention can be applied to real-time cancellation of pulse noise in a signal. On the device. The above are only the preferred embodiments of the present invention, and all changes and modifications made in accordance with the present invention should be within the scope of the present invention. 34 1269546 [Simple description of the diagram] Figure 1 is a block diagram of a conventional OFDM crossover multiplex receiver that can be input over the time domain. Figure 2 is a block diagram of another quaternary crossover multi-processor receiver. Figure 3 is a block diagram of a pulse noise canceling system of the first embodiment of the present invention. L is a block diagram of the frequency domain pulse noise cancellation module of the U-first embodiment. " Figure is a block diagram of the U-first-real side frequency domain pulse noise squaring group. Fig. 6 is a block diagram showing the pulse noise of the second embodiment of the present invention. Brother 7 is used to indicate the first condition of the position of the pulse noise in the signal conversion code. The last condition indicating the position of the pulse noise in the signal conversion code. The ninth is used to indicate the third condition in which the pulse noise is located in the signal conversion code. The first case is used to indicate the fourth condition in which the pulse noise is located in the signal conversion code. Figure 11 is a block diagram of the invention - a pulsed noise detector for shooting. Figure 12 is a block diagram of a saturation detector of the present invention. Figure 13 is a block diagram of the fast Fourier transform module of the present invention. Figure 14 is a block diagram of a pulse noise canceling system in accordance with a third embodiment of the present invention. 35 1269546 Figure 15 is a block diagram of the frequency domain pulse noise cancellation module of the second april and second embodiment of the present invention. $ is a block diagram of the pulse noise cancellation system of the fourth embodiment. Figure /7 shows the signal distribution that has not been processed by pulse noise cancellation. Figure 18 is a diagram showing the signal distribution processed by the pulse noise canceling method and apparatus of the present invention. [Description of main component symbols] 72, 82, 84, 92, 94, 1〇2, 1〇4, 1〇6 pulse noise events 110, 210, 310, 610, 1410, 1610 analog/digital converters 120, 220 Blank module 13G' 23〇, time domain processor 140, 240, 290, 340, 640, 1300, 1440, 1640 150, 250 260 270 280 320, 620, 1100, 1420 fast Fourier transform module frequency domain processor Signal elimination module reverse fast Fourier transform module noise estimator 1620 pulse noise detector 350, 450, 550, 650, 1450, 1550, 1650 frequency domain pulse noise cancellation module 36 1269546 352, 452, 552, 1452, 1552 first processing unit 354, 454, 554, 1454, 1554, interpolator 356, 456, 556, 1456, 1556

630, 1430, 1630 700, 800, 900, 1000 1120 1140, 1200 1160, 1322 1180 1182 1184 1210 1220, 1326 1230 1320, 1662 1324 1340 1400, 1600 1460, 1660 Second processing unit pulse noise cancellation controller signal conversion Code absolute value finder saturation detector buffer saturation calculator adder accumulation buffer packet detector multiplexer saturation detection unit pulse noise eliminator alternative sampling generator fast Fourier transform unit pulse noise cancellation system Cyclic Displacer 1664 Cyclic Displacement Unit 4521, 4562, 5522, 5561, 15522, 15561 Subtractor 37 1269546 4522, 4561, 5521, 5523, 5562, 15521, 15562 Converter 38

Claims (1)

1269546 X. Patent application scope: It is used in the Orthogonal Frequency Division Multiplexing (OFDM) communication system and the pulse noise cancellation system. It includes a · _ _ _ _ _ _ _ _ The pulse noise cancellation input is used to extract the pulse noise in a signal conversion code of the pulse noise cancellation system; the fast Fourier transform (fast F〇urier Gaya (5), the module a handle is connected to the input end of the pulse noise cancellation system for generating a fast Fourier transform type; and a frequency domain pulse noise elimination module, the frequency domain pulse noise cancellation module comprises: - a first processing unit, The fast Fourier conversion module is configured to: according to the predicted Pil subsubcarriers of the orthogonal frequency division multiplexing communication system, the pure 峨 and the description symbol from the fast Fourier transform module, To generate a predicted pulse noise in the guided subcarrier of the slowly varying waveform; - inserting β' minus the first-new unit, and inserting the predicted pulse in the guided subcarrier of the slowly varying waveform Obtaining the estimated cross-talk of the orthogonal frequency division multiplexing communication purely for all the human carriers in the ramp waveform; and 39 1269546 the second processing unit coupled to the interpolator, the symbol and the orthogonal The frequency division is more: according to the description, the estimated pulse multiplexer of all the subcarriers in the embroidery_ice shape is generated in all the subcarriers in a violent waveform, and all the subcarriers are sent to the human 5 with an estimated pulse, The signal of the signal is as follows: 2. The processing unit of the pulse noise cancellation system of claim 1 includes: a subtractor coupled to the fast Fourier transform mode for use in the 詨 orthogonal frequency division multiplexing communication system The expected-oriented subcarrier, from which the fast Fourier transform module receives and is directed to the subcarrier, the tiger deducts to generate the predicted pulse noise for the guided subcarrier, and a The first converter is coupled to the subtractor for downconverting the estimated pulse noise in the guided subcarrier into the gradual waveform according to the descriptive symbol. 3. The pulse noise cancellation system of claim 1, wherein the first processing unit comprises: a converter coupled to the interposer for arranging the gradual waveform according to the descriptive symbol The estimated pulse noise in all subcarriers is upconverted into the sharp waveform; and the 1269546 subtraction is 'connected to the fast Fourier transform module to use for all of the sharp waveforms The 哕μ estimated pulse noise in the carrier, the signal received from the fast Fourier transform module 51 and deducted from all the subcarriers, (10) has no signal in the financial center that does not contain the estimated pulse noise. The pulse noise of the first item cancels the green, wherein the first converter is coupled to the fast-fusing <1 defensive JL leaf conversion module for down-converting the signal into a gradual waveform; Converting the thief' is used to downconvert the estimated impulse noise in the guided subcarrier to the gradual waveform according to the descriptive symbol; and splicing to the first converter and the second conversion , for the waveform, the expected steering subcarrier is subtracted from the downconverted signal of the first converter to obtain the estimated impulse noise of the human carrier in the slowly varying waveform. In the pulse noise cancellation system of the first patent of the patent model, the first processing unit includes: 'salt, coupled to the interpolator, used to pre-determine the 1269546 subcarriers of the sharp waveform. Estimating pulse noise, subtracted from the fast Fourier transform = received signal that is down-converted to the slow = waveform in all sub-senses to obtain no prediction pulses for all subcarriers in the ramp waveform a signal of the noise; and a converter coupled to the subtractor of the second processing unit for upconverting the estimated pulse noise of all the subcarriers of the graded waveform to the descriptor Among the sharp waveforms. A pulse noise cancellation system according to the scope of the patent application scope is further characterized in that the Impulse Noise Rejection (INR) control is 'lightly connected to the pulse noise detector and the frequency domain pulse noise. The elimination module is configured to generate the descriptor according to the position of the pulse noise event in the request conversion code when the pulse noise detector detects the pulse noise in the signal conversion code. η If the pulse noise cancellation system of the patent application scope is less, the balance noise detector includes: ', the value of the value of the 纟巴, coupled to the wheel of the pulse m elimination system 'The absolute value of the sample used to calculate the signal conversion code. · The saturation detector is coupled to the absolute value evaluator to generate a first time when the absolute value of the rounded 42 1269546 is greater than a critical value. a value, and a second value is generated when the absolute value of the input is not greater than the threshold; a buffer coupled to the saturation detector for temporarily storing the value from the saturation detector And the saturation calculation n, which is connected to the buffer, is used to calculate the sum of the continuous Q嗰 value data generated by the saturation sigma. 8. The pulse noise cancellation system of claim 7, wherein the threshold value is preset. 9. If the pulse noise cancellation system of claim 7 is applied, and the critical value is (4) the input to the absolute value detector (4) an envelope ° 10. If the patent application scope i The pulse noise cancellation system, wherein the speed Fourier transform pure group includes a pulse noise canceller, which is encountered by the input end of the noise canceling system and the pulse, and the pulse noise is used by 1 Eliminate the round-in sampling of the system, the fire is used to detect the pulsed noise when the pulse noise is not detected, and when the noise detector detects the pulse The signal is used for round 1 515 temporary sampling to take 43 1269546 buffered samples. Π · The pulse noise cancellation system of claim 10, wherein the pulse noise canceller comprises: a buffer for buffering the processed samples; and an alternative sampling generator for a preset rule for generating an alternate sampling when the pulsed noise detector detects the pulsed noise; and a multiplexer coupled to the buffer and the substitute sampling generator for the pulse The noise detector outputs a sample from the buffer when the pulse noise is not detected, and is used to output an alternative sample from the substitute sample generator when the pulse noise detector detects the pulse noise. 12. If applying for the pulse noise cancellation line of the full-time first item, it is connected to the demodulator, where the demodulator is used to generate the frequency domain pulse noise cancellation module. All of them are pulsed and demodulated and output. 13.-A pulse noise cancellation system for orthogonal frequency division multi-wei ((10)hGg_heart (10) stealing, OFDM) communication system, the package 1269546 contains: a pulse noise detector coupled to the The input end of the pulse noise cancellation system is configured to detect pulse noise in a signal conversion code of the pulse noise cancellation system; a cyclic shifter coupled to the pulse noise cancellation system The input end is configured to cyclically shift the sampling of the signal conversion code by a first distance when the pulse noise detector detects the pulse noise; ❿ a fast Fourier transform (FFT) module, Xiaoma is connected to the cyclic shifter for generating a fast Fourier transform type; and the i'-frequency domain pulse noise canceling module includes: a first processing unit coupled The fast Fourier transform module is configured to: according to the expected steering of the orthogonal frequency division multiplexing communication system, a predicted subcarrier, a signal received from the fast Fourier transform module, and a description symbol. An estimated pulse noise in a guided subcarrier of a slowly varying waveform is coupled to the first processing unit for interpolating the estimated pulse noise in the guided subcarrier of the graded waveform. Selecting the orthogonal frequency division multiplexing communication system to have a subcarrier predicted pulse noise in the graded waveform; and 45 1269546 a second processing unit coupled to the fast Fourier transform module and the inserter And generating, according to the description symbol, the signal received from the fast Fourier transform module, and the estimated pulse noise of all subcarriers in the gradual division multiplexing communication system in the gradual change waveform. All subcarriers within the turbulent waveform do not contain signals for predicting pulse noise. 14. The pulse noise cancellation system of claim 13, wherein the first processing unit comprises: a converter for guiding the orthogonal frequency division multiplexing communication system according to the description symbol Subcarrier down-converting into a ramp-up waveform; and a subtractor coupled to the fast Fourier converter and the interpolator for down-converting the converter to a pilot subcarrier in the ramp-up waveform And subtracting from the signal received by the fast Fourier transform module on the preamble subcarrier to obtain the predicted pulse noise of the guided subcarrier in the graded waveform. 15. The pulse noise cancellation system of claim 13, wherein the second processing unit comprises: a subtractor coupled to the fast Fourier transform module for transforming the buffer 46 1269546 The estimated pulse noise in the subcarrier is deducted from the signal received by the fast Fourier transform module in all subcarriers to obtain that all subcarriers in the graded waveform do not contain estimated pulse noise. And the converter is coupled to the subtractor for upconverting the estimated pulse noise in all subcarriers of the ramped waveform into the sharp waveform according to the descriptive symbol. 16. The pulse noise cancellation system of claim 13 of the patent application, further comprising: an Impulse Noise Rejection (INR) controller that is connected to the pulse noise detector and the frequency domain pulse The signal cancellation module is configured to generate the descriptor according to the position of the pulse noise event in the signal conversion code when the pulse noise detector detects the pulse noise in the signal conversion code. 2. If you apply for the 13th item of the township, the signal is eliminated. The first distance is determined according to the descriptive symbol. 18. The pulse noise cancellation system of claim 13 wherein the % displaced person cyclically shifts the sampling of the signal conversion code such that the midpoint of the pulse noise duration of the signal 1269546 is converted. p The first sample. The stomach falls in the signal conversion code. 19. The pulse noise detector of the 13th item of the patent application scope includes: 〃吼 elimination system, wherein the input of the radix removal system The device is configured to generate a value of one degree when the absolute value of the input is greater than a threshold value, and use an absolute value valuer to lightly connect to the pulse balance noise end to calculate the saturation of the signal conversion code. The detector, _ the absolute value of the absolute value; to generate a second value when the absolute value of the input is not greater than the threshold; a buffer is, the handle is connected to the saturation value detector, The data stored in the saturation detector is temporarily stored; and a saturation calculator is coupled to the buffer for calculating a sum of consecutive Q data generated by the saturated detector. 20. The pulse noise cancellation system of claim 19, wherein the threshold is predetermined. 21. The pulse noise cancellation system of claim 19, wherein the threshold is a packet 48 1269546 (envelope) relative to an absolute value of the input of the absolute value detector. 22. 脉冲Applicable to the pulse noise cancellation system of the 13th patent scope, where the 位移% displacement includes a pulse noise eliminator, which is connected to the pulse noise ' / xiao xiao, 4, the input And the pulse (four) detector for buffering the input sampling of the pulse balance noise cancellation system, and for outputting the buffered sample when the pulse noise detector is not tilted to pulse=noise, and When the pulse noise [贞/则益(四) to pulse noise is used to turn off the alternate sampling to replace the buffered sample. 23. The pulse noise cancellation system of claim 22, wherein the pulse noise canceller comprises: - a buffer: a buffer for buffering processing; and a generation sampling to generate $' According to the preset switch, when the pulse noise is detected, a substitute sample is generated; and ^: lightly connected to the buffer and the substitute sample generator is used for the pulse noise detector The sampling is output from the buffer when no pulse noise is generated, and is used to substitute the output from the alternate sampling generator when the pulse noise detector detects the pulse noise. 49 1269546 24·If the pulse noise of the 13th item of the patent application scope is removed, the system is coupled and outputted to a (4) device, wherein the solution is turned over to generate the pulse pulse elimination module. The signal with no pulse noise on all secondary carriers is demodulated. The signal is used to eliminate the signal conversion code (symbol) of the orthogonal frequency division multiplexing (OFDM) communication system. The method of pulse noise includes the following steps: (a) detecting whether there is a pulse noise in a signal conversion code of the orthogonal frequency division multiplexing communication system; (b) generating the orthogonality The fast Fourier transform of the signal conversion megaphone of the frequency division multiplexing communication system; (c) receiving from the predicted pilot subcarriers of the orthogonal frequency division multiplexing communication system, from a fast Fourier transform module And a descriptive symbol to generate a predictive pulse noise in a slowly varying wave, shaped guided subcarrier; (d) inserting the estimated impulse noise in the guided subcarrier of the graded waveform to obtain the Orthogonal frequency division The estimated pulse noise of all subcarriers in the gradual change waveform, and (e) according to the description symbol and the signal of the orthogonal frequency division multiplexing communication system at 50 1269546 noise r== = ' 26. For example, the 25th item of the patent application scope + the step of the heart (2) armor (C) contains the following steps: • (f) The expected S-divided multi-frequency system is directed to the sub-carrier from the fast Fourier The 35IX received by the conversion module and directed to the secondary carrier generates predicted impulse noise for the secondary carrier; and (g) down-converting the predicted impulse noise in the secondary remainder to the Among the slowly changing waveforms. • 27· The method of claim 25, wherein step (e) comprises the following steps: (〇) upscaling the estimated pulse noise in all subcarriers of the graded waveform according to the descriptive symbol Converting to the turbulent waveform; and (g) subtracting the estimated pulse from all subcarriers of the blast waveform from the signal received by the fast Fourier transform module and subtracting all subcarriers to obtain All subcarriers do not contain the signal of the estimated impulse noise. 5.1 1269546 28. The method of claim 25, further comprising the following steps: (f) when a pulse noise detector converts the signal When the pulse noise is detected in the code, the description symbol is generated according to the position of the pulse noise event in the signal conversion code. 29. The method of claim 28, wherein the signal conversion code includes N Sampling, and step (f) further comprises the following steps: (g) when there is a pulse event in the signal conversion code, a parameter Ns is used to indicate the starting position of the pulse event, and another parameter Ne is used to indicate the pulse The end position of the event; and (h) the descriptive symbol is set to (Ns+Ne)/2. 30. The method of claim 28, wherein the signal conversion code includes N samples, and step (1) The method includes the following steps: (g) when there are a plurality of pulse events in the signal conversion code, and the number of samples between a first pulse event and a second pulse event is greater than N/2, wherein the first pulse event is The second pulse event occurs early, using a parameter Ns to indicate the starting position of the second pulse event, and another parameter Ne to indicate the end position of the first pulse event; and 52 1269546 (h) The descriptive symbol is set to (Ns+Ne)/2. 31. The method of claim 28, wherein the signal conversion code comprises N samples, and the step (1) further comprises the following steps: (g) when the When there are a plurality of pulse events in the signal conversion code, and the number of samples between any two pulse events is not greater than N/2, a parameter Ns is used to indicate the starting position of a first pulse event, and another parameter is utilized. Ne to represent the knot of a second pulse event a position, wherein the first pulse event is adjacent to the second pulse event; and (h) setting the descriptive symbol to (Ns+Ne)/2. 32. The method of claim 25, wherein the step ( a) comprising the following steps: when the absolute values of P samples in consecutive Q samples of the signal conversion code of the orthogonal frequency division multiplexing communication system are greater than a threshold value, determining whether the signal conversion code contains Pulse noise. 33. The method of claim 32, wherein the step (a) comprises the following steps: (f) calculating an absolute value of one of the signal conversion codes; (g) when the absolute value is greater than At the threshold, a first value is generated; 53 1269546 (h) calculating the sum of consecutive Q values produced by a saturation detector; and (i) the sum obtained in step (h) is consistent with one When the rule is preset, it is determined whether the signal conversion code of the orthogonal frequency division multiplexing communication system contains pulse noise. 34. The method of claim 32, wherein the step (a) comprises the steps of: (f) calculating an absolute value of one of the signal conversion codes; (g) when the absolute value is not greater than the threshold a second value is generated; (h) calculating a sum of consecutive Q values generated by a saturation detector; and (i) when the sum obtained in step (h) conforms to a predetermined rule, Then, it is determined whether the signal conversion code of the orthogonal frequency division multiplexing communication system contains pulse noise. 35. The method of claim 25, further comprising the steps of: (f) buffering a sample of the signal conversion code of the orthogonal frequency division multiplexing communication system; and (g) detecting a pulse miscellaneous At the time of the communication, an alternative sampling is generated according to a preset rule; 54 1269546 wherein when the pulse noise is detected, the fast Fourier transform of the substitute sampling is generated in the step (b). 36. The method of claim 25, further comprising the steps of: (f) cyclically shifting the sampling of the signal conversion code by a first distance when detecting pulse noise in the signal conversion code; And wherein the step (1) cyclically shifts the sampling of the signal conversion code by the first distance >, the step (b) generates a fast Fourier transform of the sampling of the signal conversion code. 37. The method of claim 36, wherein the step (c) comprises the following steps: (g) down-converting the guided subcarrier of the orthogonal frequency division multiplexing communication system to the descriptive symbol to And a (h) the downconverted to the pilot subcarrier in the graded waveform, subtracted from the signal received by the fast Fourier transform module on the preamble subcarrier, to The estimated pulse noise of the guided subcarrier is obtained in the ramp waveform. 38. The method of claim 36, wherein the step (e) comprises the following steps: 55 1269546 (g) converting the estimated impulse noise in all subcarriers of the graded waveform from the fast Fourier transform The module deducts the signal received in all subcarriers to obtain a signal that does not contain the estimated pulse noise in all the subcarriers of the ramp waveform; and (h) according to the description symbol, the ramp waveform The estimated pulse noise in all of the subcarriers is upconverted into the sharp waveform. I ^ 39. The method of claim 36, further comprising the step of: (g) delimiting the first distance in accordance with the descriptive symbol. 40. The method of claim 36, wherein the first distance is a predetermined distance. XI. Schema: 56