TW201112699A - Demodulation module, signal analyzer and signal analyzing method - Google Patents

Abstract

A signal analyzer includes a correlation computing device, a symbol boundary detection device, and a carrier frequency offset estimation device. The correlation computing device receives an input signal, delays the input signal according to different delays to obtain a plurality of delayed signals, computes correlations between a predetermined pseudo noise sequence and the delayed signals to obtain a plurality of correlation computation results, and obtains a correlation sequence according to the correlation computation results. The symbol boundary detection device detects a sample point with a maximum correlation in the correlation sequence, and generates a symbol boundary indication signal according to the correlation value of the sample point. The carrier frequency offset estimation device estimates a carrier frequency offset according to the symbol boundary indication signal.

Description

201112699 6. Technical Description: The present invention relates to a method for estimating carrier frequency drift, and more particularly to a carrier frequency drift estimation method suitable for multipath transmission channels. [Prior Art] In a wireless communication system Due to the inaccurate oscillation frequency and the Doppler Effect, a carrier frequency offset (CFO) problem may occur between the transmitting end and the receiving end. Especially in Orthogonal Frequency Division Multiplexing (OFDM) systems, the effect of carrier frequency drift is more serious. Since the transmission system of orthogonal frequency division multiplexing is a multi-carrier system, it is very sensitive to carrier frequency drift, and the carrier frequency drift will destroy the orthogonality between subcarriers in the orthogonal frequency division multiplexing system, thus causing mutual carriers. Inter-Carrier Interference (ICI) makes the system performance of the orthogonal frequency division multiplexing system worse, and the error rate increases. Therefore, how to accurately estimate the carrier frequency drift to solve the inter-carrier interference (ICI) becomes an important problem that the orthogonal frequency division multiplexing system needs to be solved. SUMMARY OF THE INVENTION According to one embodiment of the present invention, a demodulation module includes an analog to digital converter, a baseband mixer, a timing recovery device, a signal analysis device 201112699, and a decoder. The analog to digital converter converts an analog IF signal to output a digital IF signal. The baseband mixer receives the digital intermediate frequency signal, and down-converts the digital intermediate frequency signal according to a carrier frequency to generate a baseband signal, wherein the baseband mixer is further adjusted according to a first feedback control signal. The

Carrier frequency to compensate for carrier frequency drift of the carrier frequency. The timing recovery device resamples the baseband signal based on a second feedback control signal. The signal analysis device receives the baseband signal, analyzes the correlation between the baseband signal and a predetermined virtual noise sequence to obtain a complex correlation result, and enhances the correlation to generate a correlation sequence, and according to the correlation The sequence of the first feedback control signal and the second feedback signal are generated. The decoder is configured to decode the output signal of the device to generate a decoded output signal. According to another embodiment of the present invention, a signal analysis apparatus includes a correlation calculation device, a symbol margin detection device, and a carrier solution drift estimation W (four) computing device receiving-input (four), according to a complex number = delay delay delay The human signal is used to generate a complex signal, and the delayed signals are mixed with the -like virtual noise sequence to obtain a complex correlation calculation result, and a phase sequence is generated based on the results of the butterfly calculations. The symbol margin detecting means detects and relates to: one sampling point according to the correlation sequence, and sends a marginal indication signal according to the correlation number of the sampling point. The carrier frequency drift estimating means estimates a carrier frequency drift amount based on the marginal 彳a indicator of the μ pay unit. According to another embodiment of the present invention, a signal analysis method includes: an input signal 'where the input signal includes at least one data material and a sequence f delays the input signal according to a complex delay amount to attenuate the delay signal; Correlating the delayed signals with a predetermined virtual 201112699 pseudo-noise sequence to obtain a complex correlation calculation result; generating a correlation sequence based on the correlation calculation results; and estimating the round-in signal according to the correlation sequence A carrier frequency drift amount. [Embodiment] The manufacturing, operation methods, objects and advantages of the present invention will be more apparent and understood. The following detailed description of the preferred embodiments and the accompanying drawings are described in detail below. The Digital Terrestrial multimedia/television broadcasting (DTMB) system is a newly developed digital television broadcasting system specification in recent years. Dtmb's technology involves filling the known pseudo noise (PN) sequence into the guard interval to protect the data signal and further identifying the starting position of the data. The first la_lb diagram shows the data structure of the DTMB. As shown, the data 10〇 and ι〇1 are the dtmb of the frame, where the data 101 is the result of delaying the data 100 by D sampling points. #料100 and 101 respectively include the PN sequence (labeled as pN c〇de; ^ part and data data (labeled as Data) part, where the pN sequence tail part will be repeatedly added to the top of the state sequence (marked as prepN), and the head of the PN sequence is repeatedly added to the end of the pN sequence (8) is Post PN) ° For example, 'DTMB can use the pN42〇 specification, and the number of the pN sequence can be added before the data. , contains a complete Y-point PN sequence, and adds 82 points to the first half of the pN sequence. 201112699

The Pre PN portion, and the addition of 83 points to the latter half of the pN sequence, form a Post PN portion, thus forming a sequence of 42 points in length. Since the receiving end knows the content of the PN sequence added by the transmitting end, the conjugate complex number of the DTMB data received by the receiver (for example, the data 100 shown in the figure i) is delayed by D sampling with the DTMB data. The result of the point (for example, the data 101 shown in FIG. 1) is multiplied, and then the operation result is correlated with the PN sequence stored at the local end of the receiver to find the correlation with the greatest correlation. Sampling point _ Xin maX (Q ^ (8)) ' to achieve data synchronization. Among them, the sampling point with the largest correlation represents that the PN sequence stored at the local end matches the received DTMB data at this sampling point. Once the matching of the PN sequence is completed, the starting position of the PN sequence in the received signal can be obtained. Then, the data data synchronization can be achieved by recognizing the starting position of the data data by the position of the PN sequence. Examples of computational correlations can be further referenced by Ling-Long Dai et al. in November 2008 at the Institute of Electrical and Electronic Engineers (IEEE) International Conference on Communications Systems (ICCS). The published technical literature is entitled "A new frequency synchronization algorithm in the TDS-OFDM systems". In addition to achieving the timing synchronization of the data, the receiver can further estimate the carrier frequency drift Δ/ generated by the transmission channel using the obtained maximum correlation calculation result (8). The estimation method of the carrier frequency drift Δ/ is as follows: 201112699

Af=tanrVax(Corrn(n))) J M--fs where (1) where > is a sampling frequency (Fig. 4, analog to the sampling frequency of the digital converter 403). However, in a multi-path transmission channel environment, since the signal received by the receiving end can contain DTMB data from a fixed transmission path, it is theoretically necessary to perform correlation operations. In addition to the maximum correlation (Desired term), a number of related false entries (False term) are additionally generated, and even the error 顼 correlation is equal to or greater than the requested term. The term is the most closely matched sampling point of the PN sequence (with maximum correlation), that is, the result of multiplying the conjugate complex number of the theoretically received DTMB data with the delayed version of the DTMB data and the PN sequence at the local end. The sampling point that should have the greatest correlation after the correlation operation represents the sampling point. The ρβ sequence at the local end is synchronized with the pn sequence in the received DTMB data, while the other sampling points are error items. Figure 2 shows the results of the correlation calculation of the received signals in the multi-path transmission through the soil. The horizontal axis represents the take-up point and the vertical axis represents the calculation result of the correlation. It is assumed that the transmission channel contains 2 paths'. The calculation result should theoretically include two related items 201 with strong correlation. However, once the effect of the channel response of the associated 帙_multipath of the error term 202 is thus higher than the requested term, I causes an error when finding the maximum correlation sampling point max(Cwr/) («)). The result of the execution, in this way, not only causes the wrong data to be synchronized, but also leads to the wrong carrier frequency drift estimation result. 201112699 The 3a - 3b diagram shows the estimated cost of the multi-path transmission channel loop, the horizontal axis of the towel represents the frame index, and the vertical axis represents the wave according to the method described in equation (1). The result of the frequency drift estimation. In more detail, the figure ^ is based on the = term to estimate the carrier frequency drift, to compensate the result, the % figure ^ error item estimates the carrier frequency drift, to compensate the result. : shown as 'if you can find the correct item, you can accurately estimate =

The frequency drifts △/, in this model, the frame index is the day, and the actual carrier frequency of the estimated item is shifted to 1 〇〇 kHz. Two: The carrier frequency drift is divided to compensate the carrier frequency. After _f, the actual carrier frequency drift can converge. However, if 疋 estimates the carrier frequency drift based on the error term, then the helmet ^

The carrier frequency drifts by △/. That is to say, if the carrier frequency drift Δ/ is compensated according to the error term, the actual carrier frequency drift cannot converge. As shown in Fig. 3b, when the frame index is 0, the carrier frequency drift based on the error term is 200 kHZ (instead of the actual carrier frequency drifting 100 kHz), if it is compensated according to the carrier frequency drifting 200 kHz, after the compensation for a period of time, the actual carrier frequency drifts to ^ z ' instead of converge to 0 kHz. In view of this, the present invention proposes a new carrier frequency drift estimation $ method to accurately estimate the correct cable length = this carrier frequency drift estimation method is not only suitable for single path transmission, but is more suitable for multiple paths. The transmission channel environment is used to accurately complete, according to the data synchronization, and the carrier frequency drift is accurately estimated. Figure 4 is a diagram showing the receiver described in one embodiment of Genkang. The receiver 400 includes a tuner and a demodulation module 201112699 402. The tuner 401 converts the radio frequency signal Srf received by the antenna into an analog intermediate frequency signal. The demodulation module 402 can be integrated into a demodulator ic for receiving an analog intermediate frequency signal from the tuner 401 and demodulating the signal to produce an output signal SOUT (Transport Stream). According to an embodiment of the present invention, the demodulation module 402 can include an analog to digital converter (ADC) 403, a base frequency frequency 404, a timing recovery device 405, an equalizer 406, and decoding. The device 407 and the number analyzing device 408. The analog to digital converter 403 is configured to sample the analog intermediate frequency signal according to a sampling frequency (formula (1)') to output the digital intermediate frequency # number. The baseband mixer 404 downconverts the digital intermediate frequency 4 according to a carrier frequency to generate a digital baseband signal Sb. The signal analyzing device 408 is configured to analyze the characteristics of the virtual noise sequence in the digital baseband signal sB, and generate the feedback control signals Sc and ST respectively to the baseband mixer 404 and the timing recovery device 405 according to the characteristics of the virtual noise sequence. The base frequency 匕 匕 404 changes the frequency of the carrier according to the information provided by the feedback control signal sc for the offset carrier frequency offset, and the timing recovery device 405 synchronizes the information according to the timing provided by the feedback control signal ST. The digital baseband signal SB is resampled to restore the signal to the timing synchronized with the transmitting end. The equalizer 406 equalizes the output signal of the timing recovery device 405 to compensate for the frequency response of the transmission channel to remove the effects of the transmission channel. Equalizer 406 is an optional component in this embodiment. The decoder 4〇7 finally decodes the equalized signal to output the signal decoded signal s〇uT. In accordance with an embodiment of the present invention, signal analysis device 〇8 may include a correlation calculation device 4U, a symbol margin detection device 412, a timing error estimation device 413, and a carrier frequency drift estimation device 414. Correlation calculation 201112699 The device 411 is used to multiply the conjugate complex number of the digital baseband signal core obtained from the fundamental frequency mixer 4〇4 and the different delayed version of the digital baseband signal sB to obtain a complex operation result. Then, the result of the complex operation and the PN sequence stored in the correlation calculating means 411 are respectively subjected to a correlation operation to obtain a complex correlation calculation result, and a correlation sequence c is generated based on the complex correlation calculation result. Rr(n). The symbol edge detection device 412 receives the correlation sequence from the correlation computing device 411, detects the maximum correlation sampling point in each frame, and records the phase closure value and position of the sampling point to generate a character. The meta marginal indication signal SlND. The timing error estimating device 413 is coupled to the symbol margin detecting device 412 for determining the position and correlation value (ie, the amplitude of the sampling point) of the sampling point having the greatest correlation indicated by the symbol margin indicating signal SIND. The information generation feedback control signal ST causes the timing recovery means 405 to resample the digital baseband signal Sb according to the feedback control signal ST for returning to the timing synchronized with the transmitting end. The carrier frequency drift estimating device 414 is also coupled to the symbol margin detecting device 412 for determining the correlation value of the sampling point having the greatest correlation according to the symbol margin indicating signal Sind (ie, the sampling point) Amplitude) estimates the carrier frequency drift by the method as described in equation (1), and generates a feedback control signal Sc such that the baseband mixer 404 can estimate the carrier frequency drift from the compensated carrier based on the feedback control signal Sc. Frequency offset. As mentioned above, since #唬 received from the multipath's transmission channel environment generates an error item with a large correlation (as shown in Fig. 2) when calculating the correlation, it leads to the search for the maximum correlation. When m3X (C〇?TD(8)) is clicked, an incorrect result is obtained, resulting in erroneous data synchronization, 11 201112699 and incorrect carrier frequency drift estimation results (as shown in Figure 3). Therefore, according to an embodiment of the present invention, the correlation computing device 411 further enhances the correlation of the found items when calculating the correlation, so that the items found in the obtained correlation difference results may have the largest correlation value. Then, the symbol margin detecting means 412 can accurately find the position of the requested item, and the timing error estimating means 413 and the carrier frequency drift estimating means 414 can correctly obtain the estimated values of the timing synchronization information and the carrier frequency drift. Figures 5a-5c show correlation calculation results for a multipath transmission channel environment in accordance with an embodiment of the present invention. The concept of the reinforcement term described in the embodiment of the present invention can be clearly understood by the 5a-5c diagram. Figure 5a shows the delay according to the received DTMB DTMB data 〇 1 sampling point (as shown in Figure 4; == after multiplication operation) and the result of this operation is stored in the local end of the receiver. The pN sequence performs the correlation (C〇rrelati〇n) operation. As shown in the figure, it is assumed that the sampling points n2 and n4 are the requested terms, that is, the most closely matched sampling points in the pN order. At the sampling point nl ' 〇 and Μ are the wrong graphs. It can be seen that the correlation between the error items of the 7-sampling point nl and the correlation between the sampling points/foldings due to the channel response of the multi-path is # Close to 'even larger than the sampling point two = β, so if according to the correlation shown in Figure 5a - to find the exact item very accurately (at the sampling point n2 and two: Gu: Ding two steps and frequency offset Estimated 'There was an error. 5b 2Τ Μ B; expected::, the result of the operation? The result is based on the received %1, the picture:,, "Settings and the delay of the DTMB data D2 sampling points After the Ding version performs the multiplication operation, the operation result 12 201112699 and the receiver local end The result of the operation of the stored p. The effect produced by the channel response of the 5b graph can be made to make the bitmap ^^ out because the multipath is close to the sample point n5 of the sample point n2. The correlation of the phase term of the item of the error point n4, and greater than the point of the sample point.

If the progress is _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Second, the position of the occurrence of the occurrence of the case, by using different delays ^ to get = according to this, one of the implementation of the strong correlation of the item, so that θ : = raw calculation results, plus and error items produce significant Separation, such as the correlation between the results of the correlation calculation and the error item 'accumulated, the signal analysis device side can be accurately ^ ^ segmentation'

_ The correct position of the item is evaluated, and the estimated value of the carrier frequency drift is entered. The timing synchronization of the drought and the Fig. 6 show the calculation of each device according to the present invention. According to one embodiment of the present invention, the correlation package, or two or more sets of centroids === can be conceptualized, and the sixth figure shows two sets of operation modules, and the invention can also use two groups. Above, 613. The two sets of operational simulations that are worthy of attention, any ones who are familiar with the technology, and are not limited to the spirit of use, and can make some changes without departing from the scope of protection of the present invention. According to the embodiment of the present invention, each computing module can include two sets of First In First Out (FIFO) registers respectively. The lengths of the first set of FIFO registers are respectively designed according to different delay requirements. For example, the length of the FIFO register 621 can be designed as D1 to delay the input digital baseband signal SB by D1 sampling points. The length of the FIFO register can be designed as D2 1 to delay the input digital baseband signal % by D2 sampling points. After the conjugate complex unit (c〇njugat〇r, abbreviated as C〇nj) 622 and 632 converts the input digital baseband signal % into a conjugate complex number (complex conjUgate), the conjugate complex number of the digital baseband signal % The version with its delay D1 (or D2) is multiplied by multipliers 623 and 633. The multiplied result is further input to the second set of FIFO registers 624 and 634. The second set of FIFOs 624 and (4) are located in correlation calculation units 625 and 635, respectively, to aid in the operation of correlation, and thus the lengths can be designed according to the required correlation lengths C1 and C2, respectively. The result of multiplying the common extinction number of the digital baseband signal SB by the delayed version in the correlation computing units 625 and 635 through the second set of FIFO registers 624 and 634 and the pN sequence stored at the local end of the receiver, respectively Multiply point by point, and separately add the multiplication results to obtain the correlation calculation result c〇rr—to (8) and C〇rr_D2(n). Finally, the correlation calculation results Corr_D1(8) and C〇rr-D2(n) accumulated by the adder 614 are used to further enhance the correlation of the found terms to output the correlation sequence c〇rr(n). As shown in Figures 5a-5c, 'the position where the error term occurs will change with the length of the delay D, and the position at which the term occurs will not change with the delay D. After accumulating different delays The obtained correlation calculation results can enhance the correlation of the obtained items, so that the correlation of the items obtained in the obtained phase 201112699 correlation sequence can be clearly distinguished from the error items. Therefore, the symbol margin detecting means 421 can accurately find the sampling point having the greatest correlation, and generate the feedback control signal core and the 8-bit to the fundamental frequency mixer 404 and the timing recovery means 405 as described above for the carrier. Frequency drift estimation and timing recovery. Figure 7 is a diagram showing a signal analysis method according to an embodiment of the present invention. First, an input signal is received (step S7〇1), wherein the input signal may be a DTMB signal as shown in FIG. 1a, including a virtual noise sequence (PN Code) and a data data (Data), and the input signal may be The baseband signal is down-converted by the baseband mixer 404. Then, the input signal is delayed according to a plurality of different delay amounts (step S7〇2), for example, the input signals are delayed according to two or more different delay amounts to obtain input signals having different delay amounts, respectively. Next, the correlation of the input signals having the different delays with a virtual noise sequence is calculated to pass the complex correlation calculation result (step S7〇3). Next, a correlation sequence is generated based on the result of the correlation calculation (step S704). After the most, the start position of the data data and the carrier frequency drift amount are estimated based on the correlation sequence (step S705). Referring to the structure of the correlation computing device as shown in FIG. 6, step S702 and step S703 may further include: multiplying the conjugate complex number of the input signal by the input signals having different delay amounts to obtain a complex phase Multiplying the result; and calculating a correlation of the virtual noise sequence with the multiplied result to obtain the correlation calculation result, such as the delay amounts D1 and D2 shown in FIG. 6, and the correlation calculation result Corr Dl(n) and Corr_D2(n). In step S7〇4, the correlation sequence corr(8) is generated by accumulating Corr_D1(4) and Corr_D2(n). 15 201112699 In addition, referring to the correlation computing device structure as shown in FIG. 4, step S705 may further include: detecting one of the correlation sequences Corr (n) having the largest correlation; according to the maximum correlation A position of the sampling point and a correlation value generate a symbol marginal indication signal SIND; and estimating a starting position of the data data and a carrier frequency drift amount according to the symbol marginal indication signal. The present invention has been described above with reference to the preferred embodiments thereof, and is not intended to limit the scope of the present invention, and the invention may be modified and modified without departing from the spirit and scope of the invention. The scope of the invention is defined by the scope of the appended claims. 16 201112699 [Simple description of the diagram] The first la-lb diagram shows the data structure of digital TV broadcast data. Figure 2 shows the results of the correlation calculations for the received signals in a multipath transmission channel environment. Figure 3a-b shows the estimation of carrier frequency drift in the multipath transmission channel environment and the corresponding compensation results. Figure 4 is a diagram showing a receiver in accordance with an embodiment of the present invention. Figures 5a-5c show correlation calculation results for a multi-emphasis path transmission channel environment in accordance with an embodiment of the present invention. Figure 6 shows a correlation calculation device according to an embodiment of the present invention. Figure 7 is a diagram showing a signal analysis method according to an embodiment of the present invention. [Main component symbol description] 100, 101~ data;

201~ to the item; 202~error item; 400~ slave; 401~ tuner; 402~ demodulation module; 403~ analog to digital converter; 404~ basal mixer; 405~ timing Recovery device; 406~ equalizer; 17 201112699 407~ decoder; 408~ signal analysis device; 411, 611~ correlation computing device; 412~symbol margin detection device; 413~ timing error estimation device; 414~carrier Frequency drift estimation device; 612, 613~ computing module; 621, 631, 624, 634, FIFO~ FIFO register; 622, 632~ conjugate complex unit; 623, 633~multiplier; 625, 635~ Correlation calculation unit;

Corr(n), Corr Dl(n), Corr-D2(n)~correlation; D~delay;

Data~data; PN, PN code, PrePN, Post PN~virtual noise sequence; Sb, Sc, Sind, S〇ut, Srf, St~ signal.

Claims (1)

201112699 VII. Patent application scope: 1. A demodulation module, comprising: a class of analog to digital converter, converting an analog IF signal to output a digital intermediate frequency signal; - base frequency mixing ϋ, receiving the number Generating the digital frequency signal according to a carrier frequency to generate a base frequency signal, wherein the base frequency mixer further adjusts the carrier frequency according to the -th feedback control signal to compensate the carrier frequency The carrier frequency drifts; the one-time order recovery device resets the baseband signal according to a second feedback control signal; the % square knives analyzing device receives the fundamental frequency signal and analyzes the fundamental frequency signal and a predetermined virtual pulse The sequence correlation is obtained to obtain a complex correlation result, and the second correlation is performed to generate a correlation sequence, and the first feedback control signal and the second feedback signal are generated according to the correlation sequence; And a decoder for decoding the output of the timing recovery device to generate a decoded output signal. 2. The demodulation module according to the scope of the patent application, In addition, the decoder decodes the output of the timing recovery device, and then 'equalizes the output of the timing recovery device to the output decoder to output the money. 3. If the application The demodulation module according to the first item of the patent scope, the signal analysis device in J1 includes: /, a selective juice device, receiving the fundamental frequency signal, delaying the fundamental frequency signal according to a delay amount of a plurality of different 201112699 Obtaining a correlation between the delayed delay signal and the predetermined virtual noise sequence, calculating the correlation calculation result, and accumulating the correlations = to the correlation sequences; , , 'D to obtain the symbol a marginal_device's sampling point according to the phase correlation, and generating a symbol marginal indication signal according to the most significant value of the sampling point; and a phase-to-one carrier frequency drift estimating device, the root carrier frequency The amount of drift, and according to the carrier's first feedback signal; and, in the Millennium and the generation of the timing error, 四, (4) the second feedback signal. Far * no signal generation 4. If the patent scope is 3 The demodulation module carrier frequency drift estimating device according to the item is based on the -mid to the carrier frequency drift amount of the sampling point. ^, the numerical value is estimated as 5. The demodulation model according to the third application of the patent scope &gt The sequence error estimating device generates the second feedback signal according to the fetching. ^ Location and the correlation value 6. The third aspect of the computing device includes: demodulation group of: & The complex computing module, wherein each computing module comprises: a FIFO register for temporarily storing a baseband signal; and converting the delay amount to the one complex unit for using the baseband signal A total of 20 vehicles · · 201112699 A multiplier, it is the turn of the FIFO to pre-store the number as early as 'based on the first-in first-out ..., the vehicle's complex unit - (four) fruit, and the wait-to-one phase multiplicative-correlation calculation a unit, using a correlation of a virtual noise sequence, τ~phase extraction, and a result of the correlation to calculate the correlation result; and an adder for summing the operation calculation results to generate the Correlation sequence. _Reading the singularity meter two kinds of signals = analysis device for estimating the carrier frequency drift, including: Sighting & t distracting, receiving - input signal, according to the plural: delay: the amount of delay in the input The signal is used to obtain a correlation between the complex delayed delay signal and the predetermined virtual noise sequence to obtain: a number of different results, and a phase is generated according to the correlation calculation result: a 7L margin detecting device The correlation sequence debt measurement has the most productive: sampling point 'and generates a payout marginal indication signal according to the sampling point correlation value; and ▲ - carrier frequency drift estimating device, according to the symbol marginal indication signal Estimate the amount of carrier frequency drift. ~ ▲ 8. The signal analysis device according to the patent application 帛7, wherein the correlation correlation calculation method accumulates the correlation calculation results to generate the correlation sequence β. 9. As described in claim 7 The signal analysis device, wherein the symbol edge detection device generates the marginal indication signal of the symbol 21 201112699 according to the position of the sampling point, and further comprises: estimating the marginal indication signal of the input ί 'root rib symbol The starting position of one of the eight data files. 10. If applying for the (four) reading device described in the hometown _ 7 item, the correlation computing device includes: a straight eight τ plural number nose module, wherein each computing module includes a · advanced 丨 temporary storage@, for temporary use The input signal of the predetermined delay amount; the mother < Ζ factory coplanar complex unit for performing the complex signal conversion of the input signal; 00 times to ° =, connected to the FIFO register And the plurality of composite orders = for performing multiplication with one of the outputs of the first-in first-out register and one of the common complex units to obtain a multiplication result; and a correlation calculation unit for calculating The multiplication result is phased with the predetermined virtual noise sequence to obtain the correlation calculation result; and an adder is configured to add the correlation calculation result output by the operation module to generate the correlation Sexual sequence. A signal analysis method for estimating carrier frequency drift, comprising: receiving an input signal, wherein the input signal includes at least one data data and a virtual noise sequence; delaying the input signal according to a plurality of different delay amounts to obtain a complex delay signal; calculating a correlation between the delayed signals and a predetermined virtual noise sequence, 22 201112699 to calculate a complex correlation calculation result; generating a correlation sequence based on the correlation calculation results; and determining the correlation according to the correlation The sequence estimates a carrier frequency drift of the input signal. The signal analysis method described in the U.S. patent scope, the step of generating the 4 correlation sequence includes accumulating the correlation calculation results to generate the correlation. Sexual sequence. 13. The signal analysis method according to claim n, wherein the estimating step of the carrier frequency drift comprises: tilting a sampling point having a maximum correlation according to the correlation sequence; and determining a sampling point according to the sampling point One correlation value estimates the amount of carrier frequency drift. Η· The method of signal analysis as described in claim 13 further includes: estimating the starting position of the data resource based on the position of the sampling point and the correlation value. 15. The signal analysis method according to claim 5, wherein the calculating step of the correlation calculation result further comprises: ^ multiplying a conjugate complex number of the input signal by the delay signal to know the complex number Multiplying the result; and calculating a correlation of the predetermined virtual noise sequence with the multiplied results to obtain the correlation calculation results. twenty three