Communication system and method with orthogonal block encoding

Communication system and method with perpendicular block coding

The present invention relates to a known communication system and method with transmission signal coding processing, and more particularly relates to a vertical cross communication method using vertical cross codes.

Reconsidering the frequency band extension through redundant coding is currently adopted, and its advantage is to provide efficient parameters. However, if the communication device is affected by delayed reflection, time spread, or multipath, this advantage will be lost.

Code multiple access, or CDMA, is a known technology and is often used to widen the transmission frequency band. Repetitive transmission techniques such as primary detection on the receiver are extensions of the repetitive coding techniques known in CDMA to combine signal repetitions. In some CDMA applications such as direct sequence spread spectrum, simple repeated mixing, or "false spreading", and error correction coding, or "information spreading" are used to achieve the desired wider frequency band ratio.

It has been known in the prior art that the use of less-information coding and the replacement of false expansion elements with different signal methods will become perpendicular to each other, and then do not interfere with each other. For example, if a signal generates a coded bit stream a₁, A₂, A₃, A₄....., and the second signal generates a coded bit stream b₁, B₂, B₃, B₄…‥。

Then, the first signal is transmitted using four times repeated encoding, such as a₁, A₁, -A₁, -A₁, A₂, A₂, -A₂, -A₂, A₃, A₃, -A₃, -A₃, A₄, A₄, -A₄, -A₄....., and the second signal is transmitted with four repetition codes, such as b₁, -B₁, -B₁, B₁, B₂, -B₂, -B₂, B₂, B₃, -B₃, -B₃, B₃, B₄, -B₄, -B₄, B₄....., and then the symbol format comparison of repeated encoding++-++--++--++--.....The symbol format used for the first signal and repeated encoding, and +--++-- ++--++--+..... is used for the second signal, these lines indicate that half of them are different signs, and the other half are the same. Therefore, once these repetitions are combined with appropriate symbols for boosting a signal, the provision of interference signals will be completely cancelled, and vice versa. These signals are known as "crossovers".

The American digital cellular IS95 system describes the mutual vertical transmission from the cellular base station to the mobile phone. It uses a 64 symbol format selected from a set of 64 mutual vertical Walsh-Hadamard codes. One 64 times repeated encoding. The IS95 system uses non-vertical transmission in the direction from the mobile phone to the cellular base station, which replaces the information error correction code including the convolutional code connected with the vertical Walsh-Hadamard block code. In the direction from the mobile station to the base station, the vertical intersection between different Walsh-Hadamard codes is used to distinguish the difference between the different 6-bit symbols transmitted from the same mobile phone, and it fades to the direction of the mobile station at the base station. Among them, the Walsh-Hadamard codes are used to distinguish between symbols transmitted to different mobile phones.

The disadvantage of the IS95 system with non-vertical transmission in the direction of the mobile station to the base station is that if the power of the mobile station transmitter is not strictly controlled as a function of the distance from the base station, the signals from different mobile phones are roughly at Receive at the same power level. However, when the embodiment of the present invention was disclosed in the US Patent No. 5,151,91 titled "CDAFA Subtractive Demodulation" published on September 29, 1992 in Dent, the requirement for strict power control was reduced, and the decoding The signal has been removed more than once to improve the interference reduction. In the title "Multiple Coding for Radio Communications" of U.S. Patent No. 5,353,3522 issued on October 4, 1994 in Dent and Bottomley, it describes the most suitable spread spectrum access code, which is equivalent to the above-mentioned symbol format. When the vertical cross signal is used in non-vertical transmission between different transmissions, such as using the IS95 uplink in the direction from the mobile station to the base station. The reference patents disclosed above are hereby incorporated all of their references.

There are different reasons between IS95 uplink (mobile station to base station) and IS95 downlink (base station to mobile station). When the prior art communication method is used, its transmission method is to maintain the vertical handover between different transmissions. Have the correct time arrangement. If the first and second signals in the above example have a displacement relative to each other, as shown below: ++--++--++--++--+--++--++-- ++--++As shown in the two example symbol formats provided above, it can be seen that the difference is at the beginning and end of the symbol block. In this way, the handling of the cross is strictly followed.

In the downlink or base station to mobile station, the directions of all signals originate on the same base station, and such time alignment can be ensured. When the signals in the uplink or from the mobile station to the base station originate from different mobile phones located at different distances from the base station, it is not easy to achieve the time arrangement of the signals received on the base station.

The European cellular system known as GSM uses dynamic time alignment of mobile transmission, in which individual mobile phones are ordered by one base station to advance or delay the time relationship of signals received by another base. However, the ability to achieve such high accuracy synchronization, such as in less than microseconds, is limited by the phenomenon of multipath signal transmission, which is based on the characteristics of the land-based mobile radio environment.

This multi-path signal transmission phenomenon is caused by reflections of transmitted signals from large objects such as hillsides and tall buildings, and increases the delay of reflections. When it is possible to synchronize with the signal transmitted from a mobile transmitter, so that a selected signal light or reflection will be time aligned, and thus the light from another mobile transmitter is perpendicular, multi-path transmission reflecting light, The path delays of different signal rays, such as reflection or reflection, are not subject to time alignment.

The GSM system uses time-scheduled multiple access (TDMA), and each mobile signal is allocated to a time slot without overlapping transmissions from other mobile stations on the same frequency. The use of the guard time between the longest normal expected reflection time slot and the commanded advance/delay time will reduce the interference between different transmissions caused by multipath transmission. The reflection interference with the original signal is reduced by using an equalizer, which has the advantage of simultaneously increasing the unintended reflection energy of the same signal. Such an equalizer will be described, for example, the patent number 5,331,666 issued in the United States on July 19, 1994, entitled "Adaptive Maximum Likelihood Demodulator", and the patent number issued in the United States on August 2, 1994 5,335,250, titled "Method and Apparatus for Bidrectional Demodulation of Digitally Modulated Signals", which is listed here for reference only. When the use of the first equalizer cannot remove all potential multipath transmission problems, the guard time between time slots will reduce the bandwidth capability of the system.

A need still exists in systems and methods of forming and communicating signals that maintain a large amount of vertical intersections with each other, even if they are delayed by different amounts of time, for example, due to the phenomenon of multipath transmission.

When a communication system and a vertical coding method are implemented according to the present invention, the above-mentioned shortcomings of the prior art will be alleviated. The communication system and method of the present invention are provided by using mutually orthogonal coded symbol blocks to repeatedly transmit coded signals. The symbols in the repeated blocks represent the encoded information. Decoding of the vertical coded repeated symbol block of the transmitted coded signal is provided.

According to an aspect of the present invention, the description of a communication system has vertical block coding and includes a large number of transmitters. Each device uses mutually vertical coded overlapping symbol blocks to repeatedly transmit coded signals. These symbols respectively represent the sampling of the information source signal generated on the transmitter. A receiver is provided to decode the vertical cross-coded repetitive symbol blocks of the transmission coded signal received from all most transmitters. The decoding method is provided by using different vertical cross codes, and these vertical cross codes Different transmitters are combined to divide the received coded signal into relative individual channels.

In another aspect of the present invention, each transmitter of the communication system repeats each information bit generated by a digital source encoder with a preselected first number of times to continuously generate repeated bits Tuple. A symbol change is selectively applied to the repetitive bits of each of the second consecutive repetitive bit groups according to the vertical cross code of the combined transmitter.

The interleaving of the symbol variable bits from the second group number will then be realized to generate a number of blocks continuously, and these blocks are equal to the first number including the second number of symbols, and each block is Including sharing a total of different coded information bits. A modulated signal is transmitted according to the generated blocks, and the generated blocks have a sign change relative to the vertical cross code.

These and other features and advantages of the present invention will become apparent from the following detailed description, drawings, and additional patent applications.

In the following diagrams: Figure 1 is a simplified functional block diagram of the vertical block coding communication system of the present invention; Figure 2 shows the two vertical block coding signals received on the receiving system of Figure 1, which is based on The receiver is completely insensible and out of synchronization by the vertical intersecting block; the description of Fig. 3 is similar to that of Fig. 2, but shows a violation of the concept of vertical intersecting; the description of Fig. 4 is similar to the multipath transmission shown in Fig. 2 Effect; Figure 5 is a functional block diagram of a transmitter according to the present invention; Figure 6 is a functional block diagram of another transmitter configuration according to the present invention; Figure 7 is a functional block diagram of a receiver according to the present invention; Figure 8 (a ) Shows the prior art GSM TDMA burst and data bit format; and Fig. 8(b) depicts the vertical cross CDMA transmission similar to Fig. 8(a) but showing the data bits without perceptual delay according to the present invention .

Detailed description of the invention

Please refer to FIG. 1, a vertical block coded communication system 10 of the present invention can be seen to include a pair of substantially identical block coded transmitters 11 and 12, which form a plurality of transmitters, which are carried by broadcasting. Information on electromagnetic wave signals S11 and S12. Ideally, these signals S11 and S12 are digital signals, although the present invention will consider them, and can be used with analog signals modulated on a carrier. These signals S11 and S12 are received by a vertical block coded receiver 14, which decodes the vertical block coded signals and divides them into individual output channels. As indicated by the dashed line, the vertical block coded signal S11 from the transmitter 11 reaches the receiver 14 through an indirect path by the reflection of a reflective object 13 on the scene. Because the length of the reflected path is greater than the length of the direct path of the signal S12, the reflected signal S11' will arrive at the receiver 14 at a slower time than the directly received signal S12. Therefore, even if the signal S12 is synchronized so as to arrive at the receiver at the same time as the signal S11, it will not be synchronized with the reflection signal S11'.

Please refer to FIG. 2, which shows a vertical intersecting block coding communication system 10. The first signal S11 includes N information aspect samples b₁, B₂, B₃...B_NIt is repeated many times by the reciprocal indicated by a negative sign on each block or without reversing the reciprocal indicated by a positive sign. Therefore, as shown in Figure 2, when the second block S11-2 is reversed, S11-1, S11-3, and S11-4, the first, third, and fourth blocks will not be reversed. . The inverted/non-inverted format number in Figure 2 is represented by the symbol format +-++.

The second signal S12 is a signal sample including a block of signal a₁, A₂, A₃....A_N, It is also repeated use or not to use a reversal. In the case of the second signal S12, the first, second, and third repetitions are not repeated, but the fourth is inverted, which is represented by the symbol format +++-.

It can be confirmed that the first and second signal formats +-++ and +++- are vertical crosses, which means that they are the same as some of the situations in which they are not consistent.

When the first signal S11 and the second signal S12 are output at the same time, a linear increase in any one of the signal samples will occur. However, as shown in FIG. 2, the two signals S11 and S12, or the signal blocks S11-1 and S12-1 do not need to be time aligned. In the example in Figure 2, when sampling a_iAnd b(_i+2) When aligning and increasing, such sampling a_iAnd b_iDo not align, and then there is no need to increase it.

The receiver 14 is connected to be able to repeatedly receive relative signal samples during the transmission time T portion. The receiver 14 ideally converts the signal samples into an appropriate form, such as digital, which is stored in the receiver sample memory 15. If they have previously received samples, the receiver 14 processes and combines the received relative signal samples in the time period T by reading them from the memory 15. At the four sampling points of the example shown in Figure 2, the sum of the sampling values from the signals S11 and S12 are respectively a₁+b₃, A₁-b₃, A₁+b₃, And -a₁+b₃。

In combining the samples, the receiver 14 uses addition or subtraction according to the symbol format of the combined signal. In the example of FIG. 2, the symbol format +-++ is used to receive the first signal S11. Another way is that the symbol format is used to receive the second signal S12.

In receiving the first signal S11, the receiver 14 will form +(a₁+b₃)-(a₁-b₃)+(a₁+b₃)+(-a₁+b₃)=4b₃, Which describes the sample a from the second signal S12₁And -a₁Interference.

Another way is that the receiver 14 uses the symbol format to combine the received samples to form the second received signal S12, and obtain, +(a₁+b₃)+(a₁-b₃)+(a₁+b₃)-(-a₁+b₃)=4a₁, Which shows the sample b from the first signal S11₃And -b₃Interference.

In this way, the two signals S11 and S12 will intersect vertically, regardless of the improper alignment of the correlation time with the two sampling intervals. The same vertical crossing will be reserved for improper alignment at other times, which is relatively small compared to the block length of the N sampling interval. When some of the repeated bits of the present invention are implemented, the concept of vertical intersection is violated, and the number of bits that occur here is equal to the improper alignment of the time indicated in the sampling interval. Therefore, as shown in Fig. 3, when the duration of the block is larger than that of the improper alignment of time, it violates the concept of vertical intersection and only affects a small part of the bits. The receiver 14 uses the symbol format +-++ to combine the received samples to decode the samples. For b₁In terms of sampling a from'a'₃The interference will be cancelled. However, for decoding b_NIn other words, the receiver 14 will get 4b_N-a₂+a₂'………(1)

Sample from'a' to b_NThe interference will not be completely cancelled, because a₂Are repeated samples from the next set of blocks, not necessarily equal to<img file="TW387175B_D0001.tif" />. However, when the number of repetitions is larger, it is greater than four in the example, the b_NThe value will be boosted by a larger multiplier, and the interference from the'a' sampling will be almost cancelled. Moreover, any basic error correction code will allow some'b' values destroyed by irreversible interference from the'a' signal value without causing transmission errors in the basic information. So in practice, with larger blocks, such as many repetitions of 64 repetitions used in IS95, and the use of further error correction codes, the scope of the patent application of the present invention is that the signal vertical is essentially It will be maintained, even if the time between the signals at different many sampling intervals is not properly aligned.

When a signal, such as the'a' element of the second signal S12, is transmitted from a transmitter to a receiver on multiple transmission paths of different lengths, the signal will have a complex number C₀Received, and C₀It represents the phase and amplitude changes on the first path, and is received by a complex factor C, which represents the phase and amplitude of a delay path. Figure 4 illustrates the relative path delay for a sampling interval. Therefore, when the decoding method samples a₂At the time, except for the first transmission path C₀.a₂In addition to the phase change and amplitude, it will be determined by the second path factor C₁The increase in the amplitude and phase changes of the sample is further destroyed. As shown in Figure 4, the output of the receiver 14 is then 4(C₀.a₂+C₁.a₁), which is four times the output of the receiver without duplication. The output of the receiver 14 will thus be continuous: 4C₀.a₁+C₁.a_N-C₁.a_N".........(2)4C₀.a₂+4C₁.a₁ 4C₀.a₃+4C₁.a₂ 4C₀.a₄+4C₁.a₃...4C₀.a_N+4C₁.a(_N-1) Where a_N"It represents the Nth symbol of the previous block of N symbols. All outputs except the first one are determined in the two transmitted samples.

The output sequence can be processed by an equalizer, as described in the reference, which is designed to handle one or more sample delay path delays. Such an equalizer will correctly process all samples except for the edges between the two blocks, such as the first example provided by equation (2) above. The sampling at the edge of the block is appropriately processed by such an equalizer. When the number of repeated'M' combined is greater than 4, the degree of approximation will be better, so that the first output will become MC₀.a₁+(M-3).C₁.a_N-a_N"=M.(C₀.a₁+C₁.a_N-C₁.(a_N"+3.a_N)/M).........(3) When M becomes larger, error C₁.(a_N"+3.a_N)/M is compared to C₀.a₁+C₁.a_NWill lean towards zero. However, in the above example of Figure 4 based on three samples a₁, A_N, And a_NAfter the combination of ", the dependence of the receiver can be effectively modularized, and the first equalizer is constructed, which uses this module to decode a₁, Or use only one module determined on the two transmitted samples. Such an equalizer needs to maintain a larger number of decoding states or "Viterbi" states in order to rely on additional symbols to resolve the fixed signal.

In a CDMA system, the receiver 14 of the present invention thus includes non-spreading followed by a traditional equalizer for multi-path delivery. According to the present invention, the receiver 14 includes an equalizer or a Viterbi maximum possible sequence estimator form of a decision feedback equalizer (DFE). Another way is that an appropriate RAKE receiver is described in the non- Multi-path delivery in extended processing. A suitable RAKE receiver is described in the title "Quantized Coherent Receiver" of US Patent No. 5,305,349 of April 19, 1994, and is listed here for reference only.

A transmitter 16 according to the present invention is ideally structured as shown in FIG. 5, which includes a block interleaver 18, which after the last orthogonal spread spectrum coding operation is executed by the circuit 20, Operate on this signal. The circuit 20 includes a one-bit repeater 22, a direct sequence orthogonal code generator 24, and a modulo-2 adder 26.

An information source 28 provides information such as voice or transmission signals to a digital source encoder 30, which converts the information into digital form. The output of the digital source encoder 30 will be used in an error correction encoder 32, which will more tolerate the transmission of noise and interference. The output bit stream of the encoder 32 (b₁, B₂, B₃...) is expanded by the bit repeater 22, which samples each bit M times, where M is the intended expansion factor. Then by bitwise addition, the modulo-2 adder 26 adds bits to the extended bit stream, a characteristic vertical cross code system is allocated to the signal, and a direct sequential vertical cross code is used to generate Produced by Device24. An MxN block interleaving operation is performed by the block interleaver 18 on the output spread spectrum coded signal, so that the repeated bits will not be transmitted contiguously in time, but a block size of N bits Separated. The block interleaver 18 does not add or delete bits, but will change their transmission sequence. For example, by transforming a matrix of N×M bits, the other way is that the block interleaver 18 is a spiral , Diagonal, or block diagonal interleaver, rather than a pure block interleaver. The spread spectrum coded block signal is then applied to a radio carrier frequency by means of a modulator 33.

The transmitter 16 of FIG. 5 is formed by adding an interleaver 18 in a CDMA transmitter according to the present invention, which has accurate parameters (M, N) suitable for the spread spectrum code generated by the generator 24 .

FIG. 6 is another transmitter 35 according to the present invention, which includes an information source 28, a digital source encoder 30, and an error correction encoder 32.

The specific embodiment shown in FIGS. 5 and 6 includes an interleaving process that goes beyond the interleaving performed by the interleaver 18. The purpose of the further interleaving is to avoid the same continuous occurrence in the error correction decoder of the receiver 14. Sampling block error. Any such additional interleaving processing within the block is considered to be part of the error correction coding processing.

The output from the error correction encoder 32 is connected to a block repeater unit 36, which saves N consecutive bits of a block, and then repeats the block M times. A block symbol generator 37 selectively provides the symbol of each repeated block. Therefore, the block symbol generator 37 only needs to generate perpendicular codes at the block rate, instead of generating signal samples or a "chip rate" rate. The symbol from the block symbol generator 37 is a combined signal sample, such as the bit b3 from the block repeat unit 36, which uses a mutually exclusive OR gate or modulo-2 adder unit 38. Another way is to use a mod-2 adder. A chip rate hash code is generated from an access code generator 40 to randomly output the bit stream from the block symbol adder 38. The code generated by the access code generator 40 must be that all signals are equally vertical, such as signals in the same unit of a cellular telephone system.

The access code generator 40 can operate in many different embodiments. In the first specific embodiment, the use of the access code generator 40 is optional and can be omitted in some systems. The signals that cross each other will then be transmitted in the same unit. If redundant crossover codes that are not configured in a cell are available, they are helpful by using them in adjacent cells, so that the proportion of adjacent cell interference will be removed. When the transmission of a unit cannot be synchronized with the transmission of an adjacent unit, the CDMA system of the prior art cannot use, for example, vertical crossover between units. However, when the present invention is implemented, inaccurate synchronization is not a vertical handicap between the units. However, if the entire set of mutually perpendicular codes is used in the first unit, then an adjacent unit uses the second set of codes, which are perpendicular to each other, but the codes of the first unit are non-vertical. Such an additional set of codes ideally has a non-vertical intersection controlled by any other set of codes, and can be obtained by using the technique of the above-referenced US Patent No. 5,353,352, which is included in the block symbol generator 37 .

In the second embodiment, the access code generator generates a chip rate code with a length equal to the block length, and repeats the code of the repeated block. The code of the next set of repeated blocks will then change and so on. The attribute provided by the second technique is that the multipath signals delayed by some chips are produced by the access code generator 52 and the block symbol generator 54 in the receiver of FIG. 7 The symbol format cannot be extended. In this way, the multipath transmission is caused by the averager 58 to cause additional inter-symbol interference between the unexpanded symbols, which is determined by the maximum possible equalizer 60 example. When a different access code uses signals in different units, the access code is ideally similar to all signals in the same unit. The access code is ideally selected according to the technology disclosed in US Patent No. 5,353,352 to achieve non-vertical traffic control between units.

In the third embodiment, the access code generator 40 is selected to provide a non-delayed multi-path delayed signal crossover. This is achieved by applying similar read signal changes to any pair of adjacent chips in the half-block repetition, unlike the sign change in the other half-block repetition. This has a delay effect of +/- one chip relative to the nominal propagation delay caused in multipath signals, and these multipath signals are perpendicular to the nominal propagation path. The multipath signal then does not cross, but is quite similar to the code of another signal. This option will be used when only half of the available codes are used to identify between the signals in the unit, and the other half of the perpendicular codes appear on the multipath of +/- one delay chip to distinguish the multipath.

In the fourth embodiment, the access code generator 40 is an arbitrary code generator or a device other than the above. The multipath signal is then neither perpendicular to nor encoded the same as the undelayed signal. If it is intended to demodulate the path signal, a RAKE type equalizer will be used, in which the receiver uses the different time-shifted output of the access code generator 52 without spreading the received signal, and uses an averager 58 For multiple examples of, each different average is performed to generate a multiple average of each signal light with respect to different transmission delays. The different rays then combine the roughly quantized coefficients in a RAKE equalizer such as a RAKE receiver, which is described in the above-referenced US Patent No. 5,305,349. The fourth specific embodiment is not recommended for use, in which the non-generation of the vertical cross code will be affected by the relative transmission delay or synchronization error.

It is helpful that the array of signals that are not perpendicular, such as the signals in different units of a cellular wireless telephone system, are provided using different codes.

The receiver 14 in FIGS. 2, 3, and 4 is ideally constructed according to the present invention as shown in FIG. Some signals such as intended signals, interference signals, noise, and multi-path loss signals are received from the antenna 44 and provided to the input 45 of a down converter 46. The down converter 46 down-converts the radio frequency signal into a signal for proper processing, which is ideally a complex baseband signal. The complex fundamental frequency signal can be in Cartesian (X, Y) form, which has a Y or "I", a real number element, and a Y, or "Q" image element, or a polar form (R, TBETA), or The Logpolar format (log(R), THETA) is like the title "Logpolar Signal Processing" of US Patent No. 5,048,059 issued on September 10, 1991, which is only listed here for reference. The down-converted samples from the output 47 of the down converter 46 are then applied to a sign changer 48, which is connected to an access code generator 52. The down-converted samples 47 are then changed in sign by the adder 48 according to the access code symbol format provided to the access code generator 52 to remove the sign provided by an opposite transmitter code generator An access code, such as the access code generator 40 in FIG. 6. When different codes are provided for the I and Q samples at the transmitter 16 in FIG. 5, the opposite code systems are used for the I and Q samples at the receiver 14 in FIG. 7, respectively.

The real I and image Q elements from the sign changer 48 are deinterleaved by a deinterleaver 56, which works by simultaneously blocking all chips that are repetitive with respect to the same signal sampling information bit. The individual repeated symbols are also achieved by using symbol change in the symbol changer 50, which is achieved based on a set of vertical symbol grids provided by a block symbol generator 54. Another way is that the block is not chaotically processed by using the access code generator 52 to perform. Attributable to the fact that the two sign changes in the sign changers 48 and 50, respectively, are equivalent to a single sign change determined by the product of the divided signs. Therefore, it does not matter whether the pure symbol change is provided before or after deinterleaving, as long as the access code generator 52, or the block symbol generator 54, or a combination can generate the correct symbol sequence.

After the repetitions are blocked at the same time and the symbols of all the repetitions are equalized, the repetitions will be combined by an averager 58, which ideally averages or adds all the repetitions in the M-bit window , Where M is the number of repetitions. Another way is that the averager 58 is a low-pass bandwidth filter, which is similar to the filter of a block moving averager. The output of the averager 58 is then down-converted from M samples per bit to one sample per bit to generate the bit sequence b₁, B₂, B₃.... These samples include Inter-Symbol Interference (ISI). Due to the multipath transmission, they will be fed back to a maximum possible equalizer 60 next. The output value from the equalizer 60 is ideally in a "soft" form, and its 1's and 0's are represented by the indication value of the degree of "coincidence" or "zero" instead of being determined by the technical 1/0. U.S. Patent No. 5,099,499 issued by Hamar is a description derived from soft decisions, and the disclosure herein is incorporated by reference. The use of soft decisions is to improve the efficiency of an error correction decoder 64, and the error correction decoder will receive the equalized signal and generate non-technical decisions from the source decoder 66 and a "bad information frame" indicator. The source decoder 66 converts the output bit stream into, for example, a speech signal, and uses the bad information frame indicator from the error decoder 64 to mask error events and prevent noise from bursts that cannot be recognized for speech quality. In addition, if a relative interleaver is used in the transmitter 16, a deinterleaver 62 is used between the equalizer 60 and the error correction decoder 64. The de-interlacing process by the de-interlacing gas 62 is not related to the use of the de-interlacing device 56 in order to improve the vertical crossing in the case of a time error or a path.

Universally assigned US Patent No. 08/305,727 filed on September 14, 1994, titled "Simultaneous Demodulation and Decoding Device", discloses a demodulation technology that can perform all of the equalizer 60 The function, deinterleaver 62, and error correction decoder 64 can replace the use of these individual units. It is disclosed here for reference.

When the transmitter is not exactly synchronized, it violates the actual vertical crossing and retains a small part of certain transmitted symbols, such as described by equation (1). For example, the joint demodulation method of two signals can be processed according to the following items: If the described signal b_NAnd a₂Is a vector belonging to the current symbol to be demodulated<u style="single">V</u>(i) where<img file="TW387175B_D0002.tif" />and<img file="TW387175B_D0003.tif" />and<u style="single">V</u>(i-1) is similar to the combination of N transmission symbols from the previous block b_N"And a₂", and then first, combine the repetition with the'b' symbol and then with the symbol format of the'a' symbol, we can get the S as shown below_aAnd S_bSum of: S_b=4b_N-a₂+a₂'S_a=4a₂-b_N+b_N"or<maths><img file="TW387175B_D0004.tif" /></maths>

When all the signals are to be demodulated on the cellular base station or artificial satellite ground station, so other non-vertical crossovers can be completely demodulated, decided by feedback, or combined with the above-mentioned reference of US Patent No. 5,151,919 It is compensated by another demodulation method.

Therefore, the sum vector S_b, S_a4 of<u style="single">V</u>i is composed of a small amount of previous vector<u style="single">V</u>(i-1) and the next vector<u style="single">V</u>(i+1) destroyed, the quantity described by the "inter-vector interference" (IVI) coefficient is the matrix M in the following equation₀, M₁, And M₂：<u style="single">S</u>=MO.<u style="single">V</u>(i-1)+M1.<u style="single">V</u>i+M2.<u style="single">V</u>(i+1)............(5) The middle item is mixed by multiplying the above formula (4) by the inverse matrix of matrix M1<maths><img file="TW387175B_D0005.tif" /></maths>To get<maths><img file="TW387175B_D0006.tif" /></maths>It is equal to equation (4) multiplied by M1^-1。

Previous vector<u style="single">V</u>(i-1) and the next vector<u style="single">V</u>The effect of (i+1) can be approximately calculated by using equation (6)<u style="single">S'</u>(i-1) and<u style="single">S'</u>(i+1) and remove them and substitute them into equation (6) to obtain<u style="single">V</u>(i) One-by-one improved estimate<u style="single">S'</u>(i). This process is repeated to expand the needs to obtain the correct needs.

However, it is more common that the IVI expressed by equation (5) is not confused by using the matrix conversion equalizer described by the following equation<img file="TW387175B_D0007.tif" />, Where L is the selective size, and the equivalent matrix H(j) is selected to obtain the desired equivalent accuracy.

When only a few N symbols in each block will be affected, especially when the symbols are further processed by an error correction decoder, it is unnecessary to over-complicate the remaining non-vertical processing errors. It may fully comply with the symbols that will be affected by the remaining non-vertical intersections. Before applying them to the error correction decoder, a soft value indicates the uncertainty of the larger symbol.

The present invention can repeat operations with any number of blocks, not just the power of 2 from the Walsh-Hadamard symbol format of the vertical cross group. The general ability of the present invention is to determine that a radio signal can be changed in phase by any desired amount, and not only to invert it by 180 degrees. For example, a general phase shift of 120 degrees can be expressed by the multiplication of the following complex factors: S=EXP(j2π/3).

Assuming that a block of symbols is to be transmitted in three repetitions according to the present invention, the first transmitter will transmit its block of symbols with continuous phase shifts of 0, 120, and 240 degrees, and these different angles are used Repeat in three blocks. With these symbols, where S₀, S₁, S₂The lines represent 0, 120, and 240, respectively.

S₀=1, S₁=EXP(j2π/3), and S₂=P(j4π/3)=EXP(-j2π/3), the first transmitter will transmit S₀.(b₁, B₂, B₃....B_N);S₁(b₁, B₂, B₃....B_N);S₂(b₁, B₂, B₃....B_N); where (b₁, B₂, B₃....B_N) Is a symbol block of modulation rather than a phase shift. The second transmitter will transmit S0. (a₁, A₂....A_N);S₂.(a₁, A₂....A_N);S₁.(a₁, A₂....A_N); where (a₁, A₂....A_N) Represents the symbol block of its modulation, and the third transmitter will transmit S₀.(c₁, C₂....C_N);S₀.(c₁, C₂....C_N);S0.(c₁, C₂....C_N), where (c₁, C₂....C_N) Is the modulation symbol block of the third transmitter.

These three transmissions are vertical, because the following sequence S₀, S₀, S₀, S₀, S₀, S₀.....; S₀, S₁, S₂, S₀, S₁, S₂....; and S₀, S₂, S₁, S₀, S₂, S₁...; is to cross each other, even in the case of time displacement. Such a complex number of intersecting sequences is called Fourier series, and by forming them into continuous powers of EXP(j2π/L), it can be any repetition length L of the symbol. The simpler thing is that when the number of repetitions L is a power of 2, the actual Walsh-Hadamard code will be used.

According to an aspect of the present invention, other vertical interleaved sequences can also be constructed, for example, by allowing a group of continuous multipliers for continuous repetition, it is neither complicated nor limited to a binary value of +/-1. In particular, when the multipliers are selected as 1 or 0, the following vertical sequence 100000010000000100000...010000001000000010000...001000000010000000100...000100000001000000010...000010000000100000001...000001000000010000000...000000100000001000000...000000010000000100000...will appear, which in fact describes 8 time slots In the TDMA system, each signal is transmitted in the time slot when a '1' occurs, rather than when a '0' occurs. Therefore, a TDMA system is regenerated as a special case of the vertical cross-code multiple access system of the present invention that does not affect the delay. Similarly, when the complex weighting system is selected from the vertical Fu Liye series, when such as (b₁, B₂, B₃....B_NThe symbol block of) represents when the same symbol'b' is repeated N times, and when the output signal of each transmitter is output by using a filter to output a smooth signal, the present invention in this special case provides FDMA Signals, in which different transmissions are intertwined with each other irrespective of related delays, or are unsuitable due to different occupancy and no related channels.

According to another aspect of the present invention, the TDMA and FDMA systems can be regenerated and treated as special systems, and the vertical cross CDMA mode, which is not affected by delay, can be added to the FDMA or TDMA system by modifying its encoding method. Please refer to Figure 8(a), a prior art GSM TDMA signal burst and message frame format is composed of eight time slots, each time slot includes a signal burst, which has a synchronization block surrounded by data bits (syncword) element. In standard GSM, the data bits in each of the eight time slots belong to different communication links or telephone calls. The evolution of GSM allows a connection to use most time slots and provides a higher bit rate. The situation is that the data bits in consecutive time slots can come from the same communication connection or call.

Another way is that FIG. 8(b) shows how the same data bit of FIG. 8(a) is continuously repeated with or without using a phase inversion or phase change according to the present invention, so as to form an undesirable Vertical cross CDMA signal affected by delay. In Figure 8(b), the positioning of each repetition is ideally across two signal bursts. The advantage is that it can avoid the guard time from occurring between time slots and prevent a block from being affected by a sync block. distribute. This has a positive effect on how the vertical cross is protected when the time is not right, and it also avoids the need to provide the vertical cross phase shift sequence to the block synchronization S. When a block spans two time slots, the block will be allocated by the guard time, in which zero energy will be transmitted instead of being allocated by the sync block. When the energy symbol overlaps the data symbol, and the zero-energy symbol in the guard time produces interference that is lower than the entire energy symbol of the block synchronization, this will not reduce the improper timing.

Other repetition configurations in the burst can of course be used, and there is no need to have eight repetitions. For example, using a Fourier series, seven repetitions can be used with the eighth time slot for reception on the mobile side to avoid a multiplex filter connecting the transmitter and receiver to the same antenna at the same time.

Those skilled in the art can now have the advantages disclosed by the present invention, and it is obvious that the present invention can accept many forms and specific embodiments. Certain specific embodiments have been proposed so as to provide an understanding of the invention. The purpose is to illustrate these specific embodiments, not to limit the scope of the present invention. To be more precise, it is to make all modifications, similarities and changes covered by the present invention not violate the spirit and scope of the present invention defined in the appendix.