TW445733B - System and method for spectrally shaping transmitted data signals - Google Patents

Abstract

A system for transmitting from a transmitter, on a per frame basis, digital information bits which are encoded into a predefined number of signed symbols per frame for transmission over a network to a receiver, wherein the transmitted signed symbols have a desired spectral shape; the digital information bits being divided into a first predetermined number of magnitude information bits and a second predetermined number of sign information bits per frame, the transmitter including: a magnitude mapping device for mapping the magnitude information bits to the predefined number of symbol magnitudes per frame; a sign bit encoder for encoding the sign information bits into the predefined number of encoded symbol sign bits per frame; and a signal point selector, responsive to the magnitude mapping device and the sign bit encoder, which combines the symbol magnitudes and the encoded symbol sign bits to form the predefined number of transmitted signed symbols per frame; the sign bit encoder comprising; a coset representative generator which generates for each frame, coset representative sign bits for the sign information bits, defining a coset representative element for a convolutional code which identifies a coset of the convolutional code containing candidates of encoded symbol sign bits; and a symbol sign bit selector, responsive to the coset representative sign bits and the symbol magnitudes, which selects the encoded symbol sign bits from the candidates of encoded symbol sign bits that produce the transmitted signed symbols with the desired spectral shape.

Description

44573 3 Printed by A7 B7, Shellfish Consumer Cooperative, Central Bureau of Standards, Ministry of Economic Affairs 5. Description of the invention (1) Related patent documents referred to This invention is related to the following US patent applications, all assignees are assignees of this patent application 'All are incorporated herein by reference. U.S. Provisional Application No. 60 / 042,826, Title " General Spectral Forms, Application, April 8, 1997, inventor M. Vedat EyUboglu, pien: e A · Humblet; Daeyoung Kim, and David Tung. At present, this patent application is based on this patent application and its prerequisites, and therefore has a common patent application theme. U.S. Patent Document No. 08 / 747,840 'Title " Spectrum Shaping Device, System, and Method " filed on November 13, 1996 EI, inventors M. Vedat Eyuboglu, Pierre A. Humblet.

US Patent No. CXO96044P02 ’Title " Apparatus and Method of Precoded Data Signal "

Eyuboglu, Pierre A. Humblet, and Daeyoung Kim. Scope of the invention The present invention is related to high-speed data communications' and particularly to a system and method for spectrally shaping a transmitted data signal. BACKGROUND OF THE INVENTION A public switched telephone network (PTSN) consists of a digital medium-light network and an analog regional circuit. The regional circuit, which connects end users to the China National Network. The normal telephone 'an analog signal sent by a regional user is digitized by the regional central control room' and converted into a 64-kbit / second bit stream, which passes through the digital central network to the remote center The control room converts back to analog signals for transmission to the end user via a remote area loop. Dial -4- This paper size is applicable to China National Standards (CNS) A4 now (210X297 mm) ~-~~-I ------- Cantonese II (Please read the precautions on the back first (Fill in this page)

* tT 44573 3 Duplicate printing of A7 and B7 by the Central Procurement Bureau of the Ministry of Economic Affairs. 5. Description of the invention (2) A connection modem, such as a V.34 modem, can modulate digital information into analog signals for transmission. To communicate via PSTN. The amount of noise generated by the digital-to-analog conversion process at the input points of the digital central network limits the data transmission speed to about 30 kilobits per second. A technology has been developed that can significantly increase the transmission speed to about 56 kilobits per second when users directly connect to a digital network (such as via ISDN or T1). In addition, the existing International Telecommunications Union (ϊτυ) standard v 9〇__ standard convention for such transmissions is expected to be quickly certified. With this technology, a digital pulse cipher modulation (pcM) modem using a channel encoder can encode random digital information into μ-law or Α-law octets (depending on where the earth is > The digital-to-analog (D / A) converter in the central control room where the end user is located maps octets directly to symbols. (Unless otherwise stated ', the following discussions are based on μ-law, Α _Law also applies.) Depending on the load limit of the average power, 'maps all 255 levels of user phones that can use the D / A converter. Since the information is transmitted through the digital network in octets, the encoded data must first be 'Mapped to octets' makes it possible to transmit at a rate of 80 octets per second. The 'central control room where the end user is located' then converts the eight bits with a D / A converter The tuple is converted into the corresponding symbol. The generated 8 kHz symbol sequence will pass the low-pass filter (LPF) and pass through the analog loop to the end user's analog PCM modem. The output information of the d / A converter is visible. Is a sequence of pulses, each of which The amplitude of the pulse sequence corresponds to one of the D / A levels. The analog PCM modem first detects the transmitted symbols, and then maps these symbols in reverse to obtain the original digits. -5- This paper scale applies Chinese national standards (CNS > M size (210X297mm) (Please read the notes on the back before filling this page)

44573 3 Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs A7 B7 V. Description of the invention (3) Approximate value of information 'to reproduce the original information. If the information is transmitted randomly, the spectral analysis performed after the D / A conversion shows that the sequence output spectrum of the D / A converter is approximately flat. Therefore, when this sequence passes the LPF of the central control room, the signal spectrum shape is the shape of the LPF spectrum. Unfortunately, the energy of this spectrum is quite close to DC (f = 0), which will saturate the transformer in the system and cause unnecessary non-linear distortion of the transmitted signal. This application does not allow this distortion and must be eliminated. In general, PCM needs a structure that can form the signal transmitted by the D / A converter into a spectrum. In addition, in addition to PCM tubes, the spectrum-forming structure must also be suitable for various types of transmission. Figure 1 is a block diagram of a central control room of a general telephone company; Figure 2 is a frequency spectrum of a μ-law symbol output to a schematic i D / A linear converter 'and a spectrum of a low-pass filter of Figure 1 Shape; Figure 3 is a partial diagram of the two frequency spectra; each spectrum has a zero value at DC, one of which suddenly drops to 0 at DC, and the other is more moderate when it drops to 0. Figure 4 Series Figure 5 is a schematic diagram of the central digital modem configured according to the present invention and its transmission; Figure 5 is a schematic diagram of the receiver of the end-user analog pcM modem configured according to the present invention; A simplified diagram of a marker bit encoder; Figure 7 is a simplified diagram of a coset representative generator in Figure 6; -6- This paper size is applicable to China National Elephant Standard (CNS) A4 specification (2! 〇ίΪ97 public epidemic ') 'J ------- < w— (Please read the notes on the back before filling this page) D,-' ° Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs 44573 3 A7 ___ B7 V. Invention Explanation (4) FIG. 8 is a schematic diagram of a symbol tag bit selector in FIG. 6; FIG. 9 is a diagram showing a conventional cipher Fig. 10 is a flowchart showing the generalized logic of the sign bit selector in Fig. 8; Fig. 11 is a simplified view of the tag bit decoder in Fig. 5; and Fig. 12 is an upstream PCM transmitter that converts this Invented as a simplified diagram of the precoder. DESCRIPTION OF THE PREFERRED EMBODIMENTS The system and method for spectrally shaping a transmitted data signal according to the present invention can be applied to various data transmission technologies. For convenience of explanation, the present invention will be described with a PCM transmission system. However, those skilled in the art will understand that the present invention can be applied to other transmission technologies, and the PCM embodiments herein can also be applied to other technologies. Figures 1 and 2 show the signal of an analogue PCM modem transmitted to the end user through 4 analog loops, whose energy is close to do. Figure 1 is a partial view of a typical telephone central control room 10 on the PSTN. The central control room is from The central digital PCM modem with the digital part of its telephone system directly connected receives 12 μ-law octets. These symbols are output through line 16 to a low-pass filter (LPF) 18 ', which is then output through the analog circuit 20 to the receiver on the end-user analog PCM modem' to become a filtered analog signal s (t). . The receiving modem 'demodulates' the signal and outputs it as a digital stream. The digital bit stream is the approximate value β of the original transmitted signal. Figure 2 The symbol Yk from μ_law to linear converter 4 on line 16 has a flat frequency response 22. The spectral shape 24 of the LPF 18 includes considerable energy near DC (f = 0), as shown at point 26. Due to the serial-7- this paper size applies to the Chinese national standard (CNS > A4 size (210x297 public love) --- Ί--------- (W —— {Please read first please note on the back Refill the pages of the matters) Order λ 44573 3 Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs A7 B7 V. Description of the invention (5)

The frequency response of Yk is straight, so that the optical spectrum shape 24 of the signal output by the filter 18 is the same as that of the filter 18, so that when the signal is close to DC, it contains considerable energy. As mentioned above, the energy near DC will saturate the transformer in the system and cause unnecessary nonlinear distortion to the signal received by the data machine. If it is to be applied to PCM, this kind of non-linear distortion D must be reduced to reduce the energy of the transmitted signal close to DC to generate a DC zero value, which can be done. The value of DC zero in Scheme 3 is 2 8. Those skilled in the art understand that in order to establish the zero value of the DC spectrum of the transmitted signal, the total transmission digital value (RDS) of the transmitted symbol yk must be maintained, that is, the total algebraic value of all previous emission levels is maintained close to zero. . The spectral shape of the DC zero value around 28 can be a mildly oblique spectrum 30 'or a DC sharply oblique spectrum 32, and the variation range is quite large-the slope of the zero value depends on the tightness of the RDS control. As described below, the present invention encodes transmitted digital data to maintain R D S near zero. In this way, the non-linear distortion 'caused by transformer saturation can be reduced to establish zero values of DC light. Generally speaking, when necessary, the present invention encodes the transmitted digital data, and can also form the spectrum of the transmitted signal. The transmitter 40 (located in the digital PCM modem) in FIG. 4 will be downloaded from the data terminal device (such as a personal computer). ) Receives a ψ-linked bit stream 42 and encodes the received bits into octets 44 for transmission over a digital network 46. The serial-parallel converter 48 converts the serial bit stream 42 into a parallel format. The transmission / coding structure of the present invention depends on the data frame of η-symbols, where k is the index of the data frame (time). For example, each data frame can have 2, 3, 4, 5, or 6 transmitted symbols. The symbols and choices of these emissions are as follows: • Standards of this paper method ^ National Standard [CNS7A4 specification mm] {Please read the notes on the back before filling this page) J clothing

'ST

Amended whether the substance of this change is permitted. 1¾

Printed by the Central Standards Bureau of the Ministry of Economic Affairs, Shellfish Consumer Cooperatives, $ L 44 57 3 3 No. 8710 ^ 268 Patent Application Chinese Manual Correction Page (January 1990) V. Description of the Invention (6) Law Group Distribution Representing Information Bits Corresponding. The series-parallel converter 48 outputs (n_r) + m information bits to each data hub, where r is the number of redundant bits. The number of redundant bits specified in the V.90 standard can be 0, 1, 2, or 3

It is worth mentioning that the other lowercase variables in this case are scalar, and the uppercase variables are matrices. In addition, 'the row vector is bold lowercase variables and all indices starting with 0' like Xk = (Xk, Q, Xu .. ·) 'to indicate that D is described as the nr bit of Vk, which represents the information of the marked bit (marked information bit); the m bit described as U k is the intensity (intensity information bit) ). The number of bits m can be determined by selecting m that satisfies the following conditions: (1)

f = 0 V J where Mi is the positive group point of the i-th item in the data frame. v.90. This process is explained in more detail. m-intensity information bit uk, through shell mapping (explained by the ITU V.34 standard), or modulus conversion (modulus C〇nverSi〇n) (explained by the ITU V_90 standard), The intensity mapper 50 is provided to map m bits into n symbol intensities Gk. The intensity mapped by the intensity information bit is the intensity of the μ-law point that is used as the cluster point when the information bit is emitted. Those skilled in the art are aware of the above-mentioned intensity mapping device ', and will not repeat them here. The remaining information bits in the data frame, the tag information bit V 'then & supplies the tag bit encoder $ 2 to generate n tag bits Sk (encoded symbol tag bits). The following will There are details. η Characters -9- This paper size is applicable to Chinese National Standard (CNS) A4 specification (210X29? mm) (#Read the precautions of the back and then fill out this page)

Consumption cooperation of employees of the Central Bureau of Standards of the Ministry of Economic Affairs. Du printed 44573 3 A7 _____ _B7 7 Description of the invention (5 ~-strength Gk * n mark bits Sk are provided to the signal point selector w, and combined into η mark After that, the n labeled symbols Yk will be provided to the octet converter 56 to select the octets corresponding to the passed symbols and pass the octets to the digits. Network 46 ° If the marked symbol can be converted to an octet converter compatible with the PSTN digital part format, when other transmission technologies are useless, the signal point selector will directly output the marked symbol to the network The central control room (CO) 60 will receive the octets 44 leaving the digital network 46 (it can be seen that 1 has been modified due to digital loss in the network). The D / A converter in the CO 60 converts the eight bits Group 44 'converts to a symbol and transmits the symbol at a sequence level of 8 kHz to the receiver 64 of the end-user analog PCM modem via the analog loop 62. The receiver front end 66 receives the analog level' and then converts the analog signal Flat digitization for timed reproduction , Equalization, and symbol determination. The receiver front end 66 outputs the received symbol Yk in a serial format to a series-parallel converter 68, which converts the symbol of _link into n parallel labeled symbols Yk. The n parallel labeled symbols Yk are provided to the intensity signal extractor 70 'to extract the symbol intensity Gk and the flag bit Sfc from yk. The symbol intensity Gk is provided to the intensity demapping device 72 (such as a modulus-converted late reflector) ) 'To reproduce the intensity information bit Uk. As those familiar with the technology understand this inverse mapping process, it will not be repeated here. As described below, the flag bit Sfc is provided to the flag bit decoder 74 to Reproduce marker bit information Vk. Decoded information bits can be provided to data terminal equipment (such as personal computers) after further processing. -10- This paper size applies the Chinese National Standard (CNS) A4 specification (210x297 mm) ) · —Τ -------— ＜ Jing Xianwen heard the back of mackerel before filling out this page j

-IT

V Printed by Shellfish Consumer Cooperative of Central Bureau of Standards, Ministry of Economic Affairs ^ 45733 A7 --- B7__ V. Description of Invention (8) Marker Bit Coding Figure 6 further describes the marker bit code H 52. The marker information bit Vk is provided to the coset representative generator 80 to generate n coset representative marker bits Tk for each data frame, and Tk is provided to the symbol marker bit selector 82. The n coset representative mark bits in each frame will define the convolution code G (D) for the defined (used by the sign mark bit selector 82), defining the coset representative elements. As for the entire sequence of coset representative mark bits, the coset representative is defined for the convolutional code. The n cosets represent the flag bit eight. The coset of the convolutional code is also identified. The convolutional code includes the symbol bit to be encoded, which will be further described below. The sign mark bit selector 82 uses n cosets to represent the mark bits Tk, and the bits and bits containing valid non-productive code sequences defined by a lattice structure (such as the lattice structure in Figure 9) The elements perform mutual exclusion or interaction (EXCLUSIVEOR) to modify the co-set representative flag bit Tk to form the symbol flag bit to be encoded, which will be described below. The symbol bit to be encoded is a coset representing the coset element identified by the bit. With the symbol intensity 'symbol marker bit selector 82, the symbol marker bit Sk to be encoded can be selected for each frame from the symbol marker bit 1 to be encoded that generates a desired spectral shape, and these marker bits are provided. The signal point selector 54 in FIG. 4 is given. The symbol tag bit selector 82 can output the entire sequence according to each frame as S (D) = t (D) ㊉C (D); where S (D) is the coded symbol tag bit. Sequence, t (D) is the coset representative of the convolutional code, and c (D) is the code sequence of the components of the convolutional code G (D). It is worth mentioning that, through this selection process, all of the paper can be used and decoded. -11-This paper size applies to Chinese national standards (CNS > A4 specifications (210X297 cm))

(Please read the notes on the back before filling in this page J

445733 A7 B7 — —--— —-— _____ V. Description of the Invention (9) Some of the symbol mark bits to be coded become coded symbol information bits Vk ′ will be described in detail below. In this way, this spectrum forming device will not affect the symbol intensity, nor will it affect the transmission power. In this way, you can easily design a system that can meet the power limit of the FCC transmitter and can complete the spectrum formation. Figure 7. Further describes the coset representative generator 80, which includes a differential encoder 84 and a matrix block 86. Noise generated by the data channel will affect the transmitted marker bits and cause polarity inversion. Differential encoder 84 using exclusive differential encoding and differential decoder 1 32 dedicated to decoding (shown in Figure 1) can process marked bits to specific bit positions (even the following Ητ, HT, G (D) 0, 2, 4) 'In this way, the polarity inversion invariance can be completed. Matrix block 84 is modulo 2 ^ The matrix Η · τ (η_〇Χι „multiplies (ie, filters) the encoded information bits Vk ′ after differentiation to generate n cosets representing the marked bits Tk, provided to the symbol marker bit selector 82. As specified by the ITU V.90 standard, 'this matrix has 6 transmitting symbols and 1 redundant bit per frame: (Please read the back and note first (Please fill in this page again), τ

Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs ο ο ο ο 1 / (1 + D) 〇0 〇1 / (Ι + £)) 〇 〇1 / (1 十 D) 〇〇 * 1 / (1 + £ 〇〇 〇! / (! + £ >) 〇〇 〇〇 〇 Ο Ο (2) where D is the frame delay, which is determined by the frame (time) index k. As shown in Figure 8, the symbol The tag bit selector 82 includes a selector 8S, which can receive n coset representatives from the coset representative generator 80 for each frame. -12- This paper size applies the Chinese National Standard (CNS) Α4 specification (210X 297 (Mm) dd5733 A7 B7 V. Description of the invention (10) C-bit 70 Tk, and receive n symbol intensities from the intensity mapper 50 in Figure 4 and then output bit-coded bits Sk with encoded symbols for each frame output The selector 88 combines the symbol bit and intensity of the symbol to be encoded to form a labeled symbol to be encoded to provide a filter 90. The filter 90 calculates a transmission filter for each of the labeled symbols to be encoded. The total value (RFS) matrix is provided to the selector 88, which selects the coded symbols related to the coded symbols to be coded which reduces the RFS Figures 9 and 10 describe the operation of the sign mark bit selector 82. The selector 88 mutually excludes or acts on a valid code sequence containing a convolution code and a coset representative mark bit to modify η Each coset represents the marker bit T t. The accumulation code is a set of possible sequences defined by the lattice structure; the valid code sequence is a sequence that does not violate the restrictions of the lattice structure. For convenience of explanation, the selector 88 Use a single redundant bit r and a convolutional code G (D) = [1+ D 1 1+ D 1 1 + D 1]. If you want to use a lattice structure, you must use a bimorph lattice structure, like It is the lattice structure 100 in Fig. 9. The restrictions of the lattice structure are described below. The selector 88 will treat each frame k according to the restrictions of the lattice structure and interact with a specific convolutional code sequence. Repel or act to modify η cosets to represent the bit Tk. The sequence of good product codes in this example is as follows: A: 000000 (that is, without any action) b: 111 m (that is, inverting all the frames) Marker bit) c: 1 〇1 〇1 〇 (that is, even mark bits in the inverted frame) D: ο ίο 1 〇1 (also That is, the odd-numbered bits in the frame are reversed.) So, if the state Qk of the selector 88 is 0, the frame starts at frame k. 13- This paper size applies Chinese national standards (CNS > Specifications (2 丨 〇 X 297 mm) τ -------.-,-:,. ¾-- (Please read the notes on the back before filling in this page) 57 3 3 A7 B7 printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 5. When the invention description (11), only the convolutional code sequences A 102 and B 104 in the frame k are valid. Conversely, if the state Qk of the selector 88 is 1, when the frame k starts, only the convolutional code sequences C 106 and B 108 in the frame k are valid. As described above, each valid code sequence is mutually exclusive or interacts with the coset representative mark bit to form a mark bit to be coded, such as {Tk㊉A1 Tk㊉B}. Each suffix code means that the tag bit is a coset element (or a component represented by the coset) that is identified by the tag bit and includes a product code. Then, each bit of the symbol to be encoded is combined with the symbol strength to form a labeled symbol to be encoded to provide the filter 90 in FIG. 8 so that the filter can be used for each of the labeled to be encoded symbols. Symbol, calculate its RFS and return the calculated rfS to the selector 88. The selector 88 outputs the coded symbol mark bits for the frame that can reduce the RFS. The current state Qk is matched with the product code of the coded symbol mark bit selected for the frame, and the next state can be determined according to the limit of the lattice structure 丨 00. For example, if the frame k is selected When the eighth B is to be coded to pay 5 tiger bits, the state of the selector 8 8 is 1 when the frame Q k + 1 starts. Using the antecedent calculation 'can improve the spectrum formation performed by the present invention. In other words, the symbol marker bit selector 82 may not only select the encoded symbol marker bit according to the current frame, but also use the symbol intensity generated by the intensity mapper 50 in FIG. 4 and the coset to represent the marker bit. In order to determine whether the current frame and the future frame can reach the optimal spectrum formation, the coded symbol mark bit = V.90 standard stipulates that with the amount of delay calculated in advance at startup, up to three future messages can be used. frame. This paper scale is applicable to China National Standard (CNS) A4 specifications (210 ^ 297 ^ 57 (please read the precautions on the back before filling this page)

44 57 3 3

Printed by the Consumer Standards Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs⑽ For all paths (or possible sequences) that may pass through the lattice structure to the previously calculated delay or depth △, calculate the spectrum formation matrix RFS based on the furnace wave transfer function . The selector k selects the encoded sequence mark bit corresponding to the possible sequence for the frame k 'that produces the smallest ㈣. For the lattice node m 100, Figure 9 describes the possible sequence of the previous calculation depth]. When the state of the device 88 is 0, at the beginning of the frame k, only the convolutional code sequences A 102 and B 104 are valid for the frame k. However, the code sequence of the frame k + i should also be considered. The code sequences A 102 and b 104 are valid for the frame k, so that the state Q ... when the frame k + 1 can be 0 or 1; thus, the code sequences A 10, B 104_, C 106, D 108, both are valid. As mentioned above, the coset represents the mark bits that are mutually exclusive or act with each valid code sequence to form the symbol mark bits to be encoded. D can be calculated in advance in the lattice structure. Each path in the path uses mutually valid code sequences to mutually exclusive or act on the coset representative mark bits of each frame k and k + 1. The possible sequences in this example are the following four sequences: 丨〗) (Tk㊉ 八 山 + 1㊉ A), 2) (tk ㊉ A 丨 tk + l ㊉ B); 3) (tk® B | tfc + 丨 ㊉ c); 4) (tk ㊉ Bitk + l k its to be edited The sign mark bit. The flowchart 120 in FIG. 10 illustrates the operation steps of the sign mark bit selector 82. The selector 88 in step 1 22 will modify the coset representative mark bit according to the lattice structure to generate the to-be-encoded (If it is calculated in advance, it is a possible sequence). After that, the selector selects the symbol intensity and the symbol bit to be encoded at step 124 to form the marked symbol to be encoded (or (Possible sequence) 'is provided to the filter 90. The filter 90 is in step -15. This paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 mm) (Please read the cautions on the back before filling in this page. J -Binding printed by the Central Bureau of Standards, Ministry of Economic Affairs, Consumer Cooperatives 44 57 3 3 Α7 Β7 V. Description of the Invention (13) Step 126 will determine the RFS of each marked symbol (or possible sequence) to be encoded, and The RFS of the marked symbol (or possible sequence) to be encoded is provided to the selector 88. Finally, the selector 88 selects the RFS down to a smaller bit of the symbol to be encoded at step 128 and sends it. worth it It is mentioned that the present invention can also use various convolutional codes G (D), which are represented by different trellis structures and different sequences of convolutional codes. Those skilled in the art can understand that the present invention is applicable to various Plot code and code sequence β Generally speaking, according to the present invention, by using PCM, the spectrum shaping device can set the response of the chirper 90 in FIG. 8 so that the analog emitted from the C060D / A converter in FIG. 5 The signal forms a spectrum to achieve the desired spectral shape while reducing RFS. The response h (d) of the filter 90 defines the desired spectral shape and can be expressed as:

Nb Σ ^ —]

l / h (D) = B (D) / A (D) = ^ ---- > a〇 = I i = 0 where A (D) and B (D) are functions' a and b are the requirements that can be achieved Real number of spectral shape. Na and NB represent the coefficients of numerator and denominator of h (D). Based on each symbol as% ^ RTFS, the method of raising a brother is as follows: RFSi = Xbiy1_1- £ aiRFS, _1 mi = 0 i = l Depends on the RFS of each frame 'The k-th pivot rfs is calculated as follows ： -16-Chinese paper standard (CNS) applies to this paper size Α4 see the standard (210 × 297 mm) * · = 'β (Please read the precautions on the back before filling this page)

(3)

Na Σα ^-

44 57 3 3 A7 B7 «Us No. 87105268 Patent Application Revised Chinese Manual (January 1990) V. Description of Invention (Μ) RFS, ni-ΣΜ s nk + j (5) where j is the symbol (time) Exponent. If the marker bit encoder of the present invention is used to establish the zero value of the DC spectrum, its RFS is the total digital value of the drive (RDS), and the response of the filter 90 can be expressed as:

h (D) = l-D (6) The filter 90 calculates the marked symbol Yk transmitted at the symbol time j based on the transmitted marked symbol Yk: ⑺ where j is the symbol (time) index. Depending on the RDS of each frame, the RDS of the kth frame is calculated as follows: RDSk = 艺 娜 4 ⑻ (Please read the notes on the back before filling this page) Staff Consumer Cooperatives, Central Standards Bureau, Ministry of Economic Affairs Where j is the symbol (time) index. As for the antecedent calculation, in terms of the antecedent calculation depth △, the RDS calculation method is as follows: LRDSk = 艺 ⑼⑼ ('= 0 where LRDS is the antecedent calculated RDS. Similarly, we can use the first-17 paper scales to apply Chinese National Standard (CNS) A4 Specification (210X297 mm) 44 57 Revised Chinese Specification of Patent Application No. 87105268 (January 1990) V. Description of Invention (15) A7 B7 Calculate to reduce RFS: LRFSk = 幺 脑, k + i (10) Tag bit decoding Figure 11 shows the tag bit decoder of receiver 64 of Figure 5, which includes a matrix block 110. Matrix area: the matrix Ητ in t ghost 110 (in modulus 2 ) Times the mark bit Sk (that is, it is overheated) to reproduce the mark information bit V of the differential code. Equation 11: Each frame of the Ητ matrix has 6 emission symbols (n = 6 ) And an example of 1 redundant bit (r = 1): Η 1 1 + D 1 1 + D 1 0 0 0 0 1 + D 0 0 0 1 + Z) 0 0 0 \ + D 0 0 0 1 + D 0 0 0 1 + D 0 0 0 0 (11) f Please read the notes on the back before filling in this page) Order the printed matrix Η design method of the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs' Error sk Vk received signal caused by 'not spread to more than one information frame. The reason is that Ητ is a matrix of finite impulse response type with only a single delay. To show the mark bit encoder 52 of FIG. 4 for each frame of the transmitted symbol, how to decode the generated mark bit to be encoded into the same mark information bit, the encoding must be expressed mathematically. Decoding process. The reproduced information bit \ mathematical expression is: 18 This paper size is applicable to China National Skirt Standard (CNS> A4 size (210X297mm)) Central Government Standards Bureau, Ministry of Economic Affairs, Consumer Consumption Du Print 44573 3 A7 --- ----- B7___ V. Description of the invention (16)

SkHT (12) where the mathematical representation of the flag bit Sk (encoding) is:

Sk = VkH'T + TkG (13) If the right part of equation (13) is replaced by equation (12) for St ', the following equation is derived: v, = + Γ) £ 〇ητ (Η ) By selecting 0,1 ^ and h-tS> to satisfy the following conditions: (1) Η · ΤΗΤ = 1 (here, 1 is a bit-taking matrix where ght = 0, regardless of the value of rfc., Vk. Above In this example, the lattice structure 100 in Figure 9 has a single redundant bit U-., The convolutional code g (D) = (I ten D 1 1 + D 1 I + D 1). Since l * rk = rk, and 'makes rkG (D) equivalent to rk (l 1 1 1 1 l) + rk. 丨 (101010). Here the state of the iVi lattice structure Qk, rk passes through this lattice structure. Crossroads or paths. The four convolutional code sequences A to D mapped to iV! And rk are: A: 000000 -rk-i = = 0, rk B: 111111 -fk-i = 〇1 rk C: 101010 ^ fk-i- = 1, rk D: 010101 -fk-r S1 'rk where the serial codes A to D can be regarded as rkG (D). Since the value of rk does not affect the decoding method of the information bits, different valid codes can be used Η marker bit sets generated by the sequence to generate the same decoding information. In this way, _ η marker bit sets that reduce RFS / RDS to low can be selected -19- This paper size is applicable to China National Standards (CNS) Α4 standard (2 dmm) '' t Please first please fill in the weeping matters on the back)

44 57 3 3 A7 B7 V. Description of the invention (17)-

The spectrum forming device with the PCM transmission ^ L 'can also be used to wait & like a μ r ... this: the transmitter is used by the upstream PCM transmission to execute β, in response to the h (D) central control room (C0) The channel response between the converters transmitted in the line card == ⑽) most often includes the transmitted = according to the movement's "analog regional loop" and the wave effect in the co line card using the principles of the present invention , The resulting channel output sequence x (n) (if there is a front chirp, it is z (n)) 'can generate a sequence y (D) at the input of the A / D converter, the signal point is close to A / D quantization level. The goal at this point is to make it transmitted. The energy of the number x (D) »y (D) / h (D) is minimized, and at the same time, a low group deployment at the input of the A / D converter can be maintained. Group deployment is not available at this time because the larger group deployment will increase the quantization noise and other problems caused by reflected waves. Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs (please read the precautions on the back before filling this page) At this time, the 'channel response h (D) is usually judged by the receiving modem alone, or by the received and transmitted data Based on the channel measurement when the data machine is started, the machines jointly determine. The transmitted data opportunity then maps the received bits into a transmission sequence χ (D) during data transmission. After passing through the channel, x (D) is converted into a channel output sequence y (D). Usually the smallest phase is selected as the channel) 'This is not difficult (such as through the additional filtering step in the transmitter), the transmitter 40 in Fig. 12 is analogous to ρ (: M [transmission in the modem 1-line upper-domain PCM transmission transmitter 40 'uses the spectral shaping device precoder 140' of the present invention to pre-encode the received data bit stream 42 ,. Application 20- This paper scale applies to the Chinese national sample rate ( CNS) 8 specifications (2 丨 〇'κ 297 mm) Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 44 573 3 A7 _____B7__ 5. Description of the invention (18) US Patent No. CX096044P02 dated December 29, 1997 , The inventors M. Vedat Eyuboglu, Pierre A. Humblet; Daeyoung Kim, the title "Precoding Asset Signal" is based on the subject of PCM upstream precoding. The precoder 140 can perform multidimensional precoding. In other words, one by one Box precoding symbol. This is different from single-dimensional precoding in practice, but its concept is similar. The above phenomenon will be further explained below. The precoder 140 includes a series-parallel converter 481 and an intensity mapper 50 '.The marker bit encoder 52f and the signal point selector 54. The configuration and operation of these components are similar to those of the same numbered components in Figure 4. For example, as described below, the marker bit encoding is modified as described below. The pre-encoder can perform pre-coding; and its signal point selector will re-encode n frames for each frame, and the pre-coded levels xkM η corresponding to the marked symbols yk. The encoded level Xk is provided to the parallel-series converter 142 to output the pre-encoded levels in series to the front wave filter 144. The front wave reducer 144 passes these levels and passes through the wave The level is output to the digital-to-analog converter 146, which transmits these pre-coded levels through the analog channel 148. The analog channel will modify the pre-coded level xk. For the central control room (C0 The quantizer 150 in) generates a level corresponding to the marked symbol yk. In other words, the precoder responds to the analog channel 148 (to be precise, it is the target channel response h (n) ) To select the pre-encoded level xk_, which The level produced by the quantizer corresponds to the desired labeled symbol yk_. The target channel response h (n)) is equivalent to the response g (n) of the pre-wavelet 144, which is similar to the response c of the analog channel 148. (n) Convolution 'where the η symbol time refers to -21-This paper size applies Chinese National Standard (CNS) Α4 specification (210_ 乂 297 mm) "' (Please read the precautions on the back before filling in this Page) W. Order printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 44 573 3 A7 ____ B7______ V. Description of the invention (19) The number 'and h (0) = 1. The relationship is as follows: y (n) = h (0) x (n) + h (l) x (nl) + ... h (Nh) x (n-Nh) (15) Since h (0) equals 1 'So the equation can be simplified as follows: x⑻ = 少 (《)-Σ A (,) x (w-,) (1 6) / = 1 The value of h (n) for special time is the same as that before x (n) The values are all known. The oscillator 90 in FIG. 8 calculates the total term of Equation 16 as the RFS and provides it to the selector 88. The previous value of χ (η) is determined from the previous symbol y (n) based on the following relationship x (D) = y (D) / h (D) 'and then stored in the filter 90. Thereafter, the first operation mode of the selector 88 is shown in the flowchart 120 in FIG. 10, and the coded symbol s which has the lowest x (n) can be selected. Instead of transmitting the marked symbol y (n), the transmitted coded level is transmitted. In addition to the above mapping operation, the reflected wave cancellation system also includes the reflected wave cancellation to separate the transmission direction; it also includes a timing interpolation filter 'to determine that the time of symbol transmission is synchronized with the network clock. This timing interpolator is mostly driven by the clock recovery circuit used by the downstream receiver. The transmitter can also include a linear filter whose role is mainly to limit the transmission bandwidth to around 4 kHz and provide the necessary pre-filtering functions to provide the minimum phase of the h (D) of the entire channel response. In addition, the “actual system” can include a trellis-coded organization to increase resistance to noise. For example, U.S. Patent No. 08 / 74〇040, filed attorney file CX〇96〇50, filed on η 14th, 1996, -22- This Λ-sheet scale is applicable to China National Standard (CNS) A4 specifications (21〇 > < 297mm) (Please read the notes on the back of Jing before filling out this page) Order printed by the Central Standard Bureau of the Ministry of Economic Affairs and printed by the Shellfish Consumer Cooperatives 44 57 3 3 A7 _-_____ B7 _____ V. Invention Explanation (20) Question " According to the trellis code of a baseband signal selected from a fixed set of baseband signal points, its communication system, device, and method ", this trellis code is explained and can be used. It is integrated here and Reference is made to this patent. The operation of the system is basically not affected by the grid coding. 6 It is worth mentioning that the invention is applicable to both hardware and software, such as being stored in a computer-usable medium (computer disk or memory chip). The present invention can also be used for computer data signals in a carrier wave, and the signals are transmitted electronically (such as through the Internet) as in the case of software and hardware. As long as it does not exceed the spirit and field of the patent application scope of the present invention, Modify and replace the above examples , Or some modifications. The above description is only an example, not a limitation. Therefore, the present invention is defined by the patent application description, not by the above description. Any change as long as it does not exceed the meaning and field of the patent application circle is all the invention -23- This paper size it JU f Miscellaneous standard (CNS) from the specifications (210X297 mm) (Please read the precautions on the back before filling in this page)

Claims (1)

ABCD 44 57 3 3 Patent No. 87105268 Patent Application Chinese Amendment to Patent Scope (January 1990) Patent Application Scope One by One Frame The digital information bits transmitted from the transmitter through the Gangluo are exclusively transmitted to the receiver's system, where the The digital information bits have been edited. 焉 A fixed number of fixed symbols in the fixed frame are marked and emitted; they have a sufficient spectral shape, and the digital information bits are divided into a first predetermined number of intensity information bits and each A second predetermined number of information bits are marked for each frame, and the transmitter includes:-an intensity mapping device that can map the intensity information bits to a predetermined number of symbol intensities for each frame; a marker bit encoder , Which can encode the marker information bits into a predetermined number of coded symbol marker bits per frame; and a signal point selector corresponding to the intensity mapping device and the marker bit encoder, which can The coded symbol mark bits are combined to form. A predetermined number of mark symbols have been transmitted for each message; The 7G encoder includes: a coset representative generator 'can generate each light coset representative flag bit for the tag information bit, to define the coset representative element of the convolution code, the identified convolution code The coset includes the symbol mark bit to be encoded; and a symbol mark bit selector corresponding to the representative mark bit of the coset, which can select the encoded symbol mark bit from the symbol mark bit to be encoded. 7C To produce an emitted marker symbol with the desired spectral shape. 2. If the system of item 帛 i of the patent application is applied, where the first pre-number of intensity information bits is m bits, and the second predetermined number of papers marked with information bits are from the applicable Chinese National Standard (CNS) from Threat (21GX297 public fantasy --- (please read the notes on the back before writing this page) Order · Printed by the Central Standards Bureau of the Ministry of Economic Affairs, Shellfish Consumer Cooperatives, 8 8 8 ABCD Cooperative printed 44 57 3 3, the scope of the patent application is nr bits, where η corresponds to the predetermined number of symbols of each frame 'r corresponds to the number of redundant bits used by the mark bit encoder. 3. For example, the system of claim 2 of the patent scope, in which the coset representative generator includes a differential encoder, which can differentially encode the predetermined bit position of the nr bit and provide it to the marker bit encoder to complete the constant polarity inversion. 4. The system of item 3 of the patent application park, in which the intensity mapper uses a modulus conversion mapping: mapping device 'for each frame, mapping m intensity information bits into n symbols. 5. If applied Patent Scope Item 4 In the system, the coset representative generator also includes a matrix block. The matrix block multiplies the η-ι label information bit by the matrix ΗΓΤ to generate n coset representative label bits for each frame. 6 For example, the system of claim 5, wherein the symbol bit selector includes a selector and a filter, and the selector includes: logic. The logic can generate the to-be-encoded symbol from the coset on behalf of the marker bit. Symbol mark bit; logic 'this logic can combine the symbol mark bit to be coded with the symbol strength' to form the marked symbol to be coded; wherein the waver includes: logic 'this logic and the marked symbol to be coded Correspondence to determine the RFS of each marked symbol to be encoded, and the logic that can provide the selector to the RFS of each marked symbol to be encoded; and the logic 'this logic is related to each of the marked symbols to be encoded Corresponds to the RFS of the symbol in order to select the related to the marked symbol to be encoded which can reduce the RFS (please read the precautions on the back before filling this page; > 44573 3 A8 B8 C8 D8 Printed by the Consumers' Cooperatives of the Central Bureau of quasi-branch of the Ministry of Economic Affairs 31 6. The scope of the patent application has been edited with the symbol mark bit D 7. If the system for applying for the fifth item of the patent park, the symbol mark bit selector Use pre-calculation and include a selector and filter 'where the selector includes:' logic that generates a sequence of symbol tag bits to be encoded from the coset representative tag bits; logic, the logic The coded symbol mark bit sequence and symbol strength are combined to form a marked symbol sequence to be encoded; the filter includes: logic, and the logic corresponds to the marked symbol sequence to be encoded to determine each of the The RFS of the marked symbol sequence, and the logic that can provide the selector with the RFS of each marked symbol sequence to be encoded; and wherein the selector also includes: 1 logic, which is related to each of the marked symbols to be encoded The RFS of the sequence corresponds to selecting the coded symbol flag bit from the marked symbols to be encoded which can reduce the RFS. 8. If the system of item 7 of the scope of patent application, the receiver includes an intensity, signal extractor, the extractor can separate the transmitted marked symbols into coded symbol mark bits and symbol intensity, the receiver It also includes the tag bit decoder 'can decode the coded symbol tag bit into the tag information bit. "9. If the system of the patent application item number 8, the tag bit decoder includes a matrix block' this area The block has a matrix Ητ. With this matrix -3- not paper size, the country's national standard (CNS) A4 size (210X297 mm) (Please read the precautions on the back before filling this page) 445733 Α8 Β8. C8 D8 6. Scope of patent application Multiply the coded symbol mark bit to reproduce the mark information bit. 10. The system according to item 9 of the scope of patent application, further comprising a differential decoder 'for differentially decoding a predetermined bit position of the reproduced marker information bit. 1L The system according to item 10 of the patent application scope further includes an octet converter corresponding to the signal point selector to transmit the octets corresponding to the marked symbols through the network. . 12. For example, the system of claim 10 of the patent scope, where η is equal to 6, and r is equal to the matrix Η * τ is defined as follows: 0 0 〇〇0 1 / (1 + D) 'τ 0 〇00 1 / ( 1 + /)) 〇H'T = 0 〇0 1 / (1 + D) ο ο 0 〇1 / (1 + D) 〇 〇〇.0 1 / (1 + β) 〇 〇 〇〇ο 13. If the system of item 12 of the patent is requested, the definition of the convolutional code is as follows: G (D) = [1 + D1 1 + D1 1 + D1]. 14. If you apply for the system of item 3 of the patent scope, the definition of the matrix is as follows (please read the precautions on the back before filling this page) The paper standard printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs applies to Chinese National Standards (CNS) M specifications (2 mm) X 297 mm) 44573 3 AS B8 C8 D8 VI. Patent application scope 1 1 + D 1 1 + D 1 0 0 0 0 1 + D Ητ = 0 0 0 1 + D 0 0 0 1 1 + Z) 0 0 0 1 + D 0 0 0 1 + D 0 0 0 0 15. As described in item 7 of the scope of patent application System, where the RFS is the Drive Digital Totals (RDS). … 16. —A kind of frame that transmits the digital information bits transmitted from the transmitter to the receiver's square-by-frame through the network, wherein the digital information bits have been encoded into a predetermined number of marked symbols per frame and The transmitted marked symbols have the desired spectral shape, and the digital information bits are divided into a first predetermined number of strong night information bits and a second predetermined number of marked information bits per frame. The method includes: '' Map the intensity information bits into a predetermined number of symbol intensities per frame; encode the marker information bits into a predetermined number of coded symbol marker bits per frame; and print by the Consumer Cooperative of the Central Standards Bureau of the Ministry of Economy (Please read the notes on the back of the performance first and then fill out this page.) Combine the symbol strength and the symbol bit to be encoded to form a predetermined number of marker symbols that have been transmitted: The encoding steps include: For the marker information bit Generate a coset representative bit for each frame to define a coset representative component of a convolutional code. The coset identified by the convolutional code includes the code to be encoded. And the use of cosets to represent the sign bit and the sign strength. From the sign to be coded ______-5-This paper is again suitable for Ningguo Family Standard (CNS) y ----- 57 3 3 A8 B8 C8 D8 Patent Application Scope Printed by the Central Bureau of Standards of the Ministry of Economic Affairs® C Industry Consumer Cooperatives =: Bit Select the coded symbol mark bit to generate the emitted mark symbol with the desired spectral shape. … Please: the method of the 16th item of the profit range, in which the first pre-number of the intensity information bits is m bits, and the second predetermined number of marked information bits = η-Γ bits', where n and each If the predetermined number of symbols of the frame corresponds, r corresponds to the number of redundant bits used in the encoding step. 18. According to the method of claim 17 in the scope of patent application, the generation step includes: differentially editing the binary unit position of the binary to complete the polarity inversion. 19t The method of claim 18 in the scope of patent, wherein the mapping step, The system uses a modulus conversion missing device to map m intensity information bits into n symbols for each frame. 20. The method of claim 19, wherein the generator step 其中 also includes multiplying the matrix block r by n-r mark information bits to generate n coset representative mark bits for each frame. 21. The method of claim 20 in the patent application scope, wherein the selection steps include: generating the symbol mark bit to be encoded from the coset representative mark bit; combining the symbol mark bit to be encoded with the symbol strength to form The marked symbols to be encoded; determine the RFS and logic of each marked symbol to be encoded to provide the selector with the RFS of each marked symbol to be encoded; and to select and reduce the RFS to be encoded The coded symbol tag bit associated with the marked symbol of. 22. For the method of applying for the scope of patent No. 20, the selection steps are -6- The silver scale of this paper is applicable to China National Standard (CNS) A4 (210X297 mm) (Please read the notes on the back first (Fill in this page again) The Ministry of Economic Affairs and the Central Government Bureau of Standards printed 44573s AbI ____S and applied for the scope of patents' a — '— pre-calculation'. This step also includes: generating the symbol bit to be encoded from the coset representative bit bit sequence. Combining the symbol bit sequence to be encoded with the symbol strength to form the marked symbol sequence to be encoded; determining the RFS and logic of each to-be-encoded marked symbol sequence to provide the selector with RFS of the encoded sequence of marked symbols; and selecting the encoded symbol mark bits from the marked symbols to be encoded that reduces the RFS. 23. The method according to item 22 of the patent, further comprising the following steps: The receiver separates the transmitted marked symbols into coded symbol mark bits and symbol strength 'and decodes the coded symbol mark bits into Tag information bits. 24. The method of claim 23, wherein the decoding step includes the following steps: the matrix Ητ is multiplied by the coded symbol mark bits to reproduce the mark information bits. 25. The method of item 24 of the scope of patent application, also includes the following steps: differentially decoding the predetermined bit position of the reproduced marker information bita. 26. The method of item 25 of the patent scope, further comprising: The following steps 骡 · Transmit the octets corresponding to the marked symbols through the network. 27. For the method of applying for the 25th item in the patent application park, where n is equal to 6, Γ is equal to t, the definition of the matrix Η_τ is as follows: This paper size applies the Chinese National Standard (CMS) A4 specification (210X297 mm) ----- -------- V., .. '/ u (please read the note on the back and fill in this page first) Order. 44573 3 A8 B8 CS D8 VI. Application scope of patent' 0 0 0 〇Λ AAA 0 U 1 / (1+ D) Ding U u 0! / (! + £)) 0 Η · τ = 0 0 0 1 / (1 + D) 0 〇0 0 1 / (1 + D) 0 0 0 0 _ _0 1 / (1 + D) 0 0 0 0 28 · If the method of the 27th scope of the patent application, the definition of the convolutional code is as follows: G (D) = [1 + D 1 l + DI l + D 1]. 29. If the scope of the patent application is as follows: The method of 28 items 'where the definition of the matrix HT is as 1 l + D 1 1 + D 1' 0 0 0 0 1 + D HT = 0 0 0 1 + D 0 0 0 \ + D 0 0 0 l + D 0 0 0 1 + D 0 0 0 0 0 (please read the note on the back before filling out this page) Order. Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs 30. If you apply The method of item 22 of the patent, wherein the RFS is a transmission digital total (RDSP 31. — a system for precoding digital information bits in the transmitter one frame at a time into a predetermined number of precoded levels per frame , And the digital information bit is divided into a first predetermined number of intensity information bits and a second predetermined number of information flag bits for each frame, the system includes: an intensity mapping device that can convert the intensity information bits Maps to a predetermined number of symbol intensities per frame; a marker bit encoder 'encodes the marker information bits into each ** 8 paper music scales using the Chinese national standard < CNS > A4 specification (210X297 Mm) 8 888 ABCD 445733 VI. Patent Application Fan Box A predetermined number of coded symbol mark bits;-a signal point selector corresponding to the intensity mapping device and the mark bit encoder, which can combine the symbol strength and the coded symbol mark bits to form a mother frame A predetermined number of marker symbols have been transmitted, and a precoded level corresponding to the marked symbols is output; the marker bit encoder includes: a coset representative generator, which can generate a The coset represents the flag bit, and the coset of the convolution code identified by the coset representative element of the convolution code includes the symbol flag bit to be encoded; and a symbol corresponding to the coset flag bit The marker bit selector can select the coded symbol marker bits from the symbol marker bits to be coded. 32. A kind of pre-coding digital information bits into a predetermined number of each frame in the transmitter Precoding level method, and the digital information bits are divided into a first predetermined number of intensity information bits and a second predetermined number of information bits for each frame marker The method includes: mapping the intensity information bits into a predetermined number of symbol intensities per frame; encoding the information bits into a predetermined number of coded symbol mark bits in each frame; The symbol mark bits are combined to form a predetermined number of mark symbols that have been transmitted in each frame, and the pre-coded levels corresponding to the marked symbols are output. The encoding steps include: -9-This paper uses the Chinese country Standards (CNS > A4 Wash (210 > < 297 Gongchu) (Please read the notes on the back first, also this page ⑤) Order printed by the Ministry of Economic Affairs t Central Standards Bureau employee consumer cooperatives 4 S. ^ 573 3 A. The scope of the patent application A8 B8 C8 D8 generates the coset of each frame. The coset of the product code includes the sign bit of the h to be edited: and 33. -------- ^-(read the notes on the back before reading 4 and " · :: · this page) Use cosets to represent marker bits and symbol strengths. Select bit encoded symbol mark bit. A type of digital information bits transmitted from a transmitter received from the network > χ frame by frame: 疋 receiver 'where the digital information bits have been coded into a pre-sufficient number of marked symbols per frame and the transmitted marked The symbol has the desired spectral shape, and the digital information bits are divided into a first predetermined number of intensity information bits and a second predetermined number of rejection information bits. The receiver includes: •? Τ An intensity and signal extractor that can distinguish the transmitted marked symbols into encoded symbol mark bits and encoded symbol intensities; an intensity anti-mapping device that can decode the encoded symbol intensity into intensity information Bits; — a marker bit decoder that can decode the strength of the coded symbol into marker information bits; printed by the consumer cooperation of the Central Bureau of Standards of the Ministry of Economic Affairs; where the marker bit decoder includes a matrix block, the area The block has a matrix Ητ that can be multiplied by the coded symbol marker bits to reproduce the marker data bits. 34. If the receiver of the scope of patent application is No. 33, the definition of the matrix HT is as follows: -10 This paper size applies to China National Standard (CNS) A4 specification (21 OX297 mm) 5 4 4 3 3 ABICD VI. Application Patent scope 1 0 0 1 + D 1 + D 1 1 + D 1 0 0 0 1 + Z) 0 0 1 + D 0 0 1 + D 0 0 1 + D 0 0 0 0 0 0 0 (Please read the back first Please note this page before filling in this page) i / ρΛ Order printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs -11-This paper size applies to China National Standard (CNS) A4 (210 × 297 mm)