TW584992B - Parallel interference cancellation receiver for multiuser detection of CDMA signals - Google Patents

Abstract

A parallel interference cancellation receiver that reduces impulse response interference using a model of the received signal similar to that used in block linear equalizers. Block linear equalizers comprise decorrelating receivers, zero-forcing receivers, minimum mean square error receivers and the like. The invention comprises an interference computation processor feedback loop for correcting the output of a direct interference canceller. The m iterative process removes interferers from the output symbols of a matched-filter. The PIC receiver uses received signal models of the various block linear equalizers that do not assume that each subchannel consists of several distinct paths. The receiver estimates the impulse response characteristic of each subchannel as a whole.

Description

584992 V. Description of the invention (1) The present invention relates to a multiple access digital communication system. More specifically, the present invention relates to a parallel interference cancellation receiver system and a method for receiving data from multiple users simultaneously. A multiple-access digital communication system enables multiple users to access the same communication medium to transmit or receive information. For example, the medium may include a network, a cable to a local area network (1 a n), a copper wire to a typical telephone system, or an air interface for wireless communications. A multi-access digital communication system with a known technique is shown in the first figure. «Haitong Media is deducting a communication channel. Communication technologies such as frequency-division multiple access (FDMA), time-division multiple access (TDMA), carrier-division multiple access (CSMA), password-division multiple access (CDMA), and others used to enable more than one user to store Take the same communication media. These techniques can be mixed together 'to produce a mixed diversity of multiple access systems. For example, the time-division duplex (TDD) of the third-generation W-CDMA standard is proposed, which is a coupling between Ding Long people and ⑼ ^. One example of a known CDMA communication system is shown in the second figure. CDM is a two-communication technology, in which data is transmitted by adjusting the data as a false interference signal and transmitted in a wide band (extended spectrum). The transmitted data signal may have a bandwidth of only a few kilohertz, and a force may be several million hertz. The communication channel is used simultaneously by the κ separate subchannel. Mother primary channel, all other secondary channels are considered interference. What is not shown in Fig. 1, a given channel of a given bandwidth is a unique one, ...; Eight repeats a predetermined pattern generated by a wide bandwidth, a false interference (PI) sequence. These unique user extensions ^

Page 5 584992 V. Description of the invention (2) The code is a typical false vertical to another close to zero. A data signal is a spread spectrum signal in bits. A carrier signal is adjusted, and dependent on the transmission, the digital spread spectrum signal pn sequence is extracted and then copied. The received signal may be related to a specific signal related to a specific extended password, and when the signal is not added to the specific extended password. The chip rate of each value of the extended cipher, which is equal to or faster than the data rate of the channel, is a value representing more than two binary bits in order to extend the data signal. Three versus four values. The concept of the symbol The bit content of each symbol specifies a number 'that a particular pulse or wave is converted to the symbol, which is demodulated by the adjustment point. The standard method for the extended secondary channel of a CDMA system and the desired secondary channels of all other secondary channels is interference, similar to interference, which shows' if the extended cipher cross-correlations the pn sequence is adjusted to produce a The digital number is then transmitted through the digital spread spectrum Ipux media. A receiver adjusts the number. The transmitted data is associated with a commensurate user signal when the extended password is perpendicular to another, and the user code 'if only the desired user's message is added to what other users know A chip, and has a crystal data rate. The wafer rate and the extension parameters. The range of values is possible. One symbol is used. The three- and four-symbol symbols each have a higher level of information, due to the specific pulse shape. The specific number of end-view symbols. The resources and transmissions from that source have been selected at the terminal code via the secondary channel to minimize interference between the desired. So, it is used to demodulate all other sub-channels used to process itself in the communication medium. Set

V. Description of the invention (3) The receiver used in this procedure is a single supplementary receiver # ^ RME receiver. For early users, a suitable filter will slightly interfere with each other because of different sub-channels. Another method is to resolve all sub-channels in the -receiver. The receiver can be counted in parallel to receive all uses of one transmission •. This thought

Lt =. Multi-user predicates can provide-important improvement of results is better than earlier user receivers.

Please refer to the third figure, which is the block diagram of Bai Zhi, Tian Xi ^ μ ^ ^ ^ V Baizhi uses a multi-user detection CDMA receiver. Those who are familiar with this technology include this function such as radio transmission frequency or rf ^ = = radio transmission frequency channel, analog-digital conversion, m's ^ signal, for analog or digital, including the signal processed by light man 2. It is said that the extension of all activated secondary channels in the brown mouth: ^ Xier detection performs multi-user detection and outputs Fei Zhao Tong: that corresponds to every-activated secondary channel. All or—this; the number may be processed. Second, the most frequent reason is the frequency of people. The detection of the user of the heart is to strengthen the calculation | Jin y: complex mathematical operations, and therefore it is difficult to economically = very ideal. Expenditure 2 eliminates or has developed the PIC receiving epigastric and Asian Union: close to the optimal performance of this tester. Linear car with the widest mean square difference ("tester"), forcing it to be a multi-stage iterative receiver, with soft decision or 584992

foundation. Block diagram of the system, such as when the CDMA communication is asynchronous or asynchronous (as shown in Figure 3 and Figure 4). The receiver evaluation from the communication media specific is transmitted to a channel evaluation data coupled to a channel evaluation This device uses the impulse response of each symbol in other sub-channels. The line is based on the solution, and the 2 :, f μ impulse response is evaluated with the extension of the sub-channel code, where ::: of the channel Data. $ Data is output to the data of individual users ^ Η I parallel detection of κ channel users in the physical management system, cry ^ The user detection method is implemented as a fixed array of thumb gates, microprocessing of aa bits into Processor (DSP) and the like. Fixed logic ,,,, and § microprocessor-driven systems allow greater programming speed when providing programming flexibility. Any tool for multi-user detection Perform a sequence of mathematical operations. In order to describe this function, the following variables typically define the structure and operation of a linear multi-user detector: K = total number of users / active transmitters in the system

Nc = —The number of wafers in the data block. The need for the number of chips is due to the change in the extension parameters, which is common to all users. For the example of synchronous CDMA, the symbol from the user with the largest extension parameter may consist of a data block. Therefore, Nc can be reduced to an extension parameter equal to ^. 'W = impulse response length of this communication channel in the chip. This is usually a predetermined parameter of the system. Q (k) = extension parameter of user K. This extension parameter is equal to the number of wafers 584992

This is a symbol used to extend user data. A system knows the extended parameters in advance, without the need to evaluate from the received data. ° force

Ns (k) = number of symbols sent by user K is (k) = Nc / q⑴

X

NsT = g Ns (k) = total number of sent symbols. Data (information) submitted by the user. The data is in the form of a vector. One of the vectors is the moment of one of the data by the index of the subscript variable: as a result of the operation vector and the matrix operation, all vectors are defined as the row direction and the nth 70 prime is The nth symbol transmitted by the kth user. ^ The impulse response of the secondary channel represented by vector by user k. In order to be evaluated in this receiver. Receiving of subchannel impulse response *. Vector, the element of which is a typical complex number, which represents can be: from ', the incoming amplitude and phase difference of the channel. The extended password using the needle by -person V (:) is represented by a vector. For multi-testing purposes, it is useful to include extended symbols: specific symbols. The fenfenbo 1 is only π π, with extension ~ has been used to extend the kth use S ^ ^ ^ (k'n) is defined as the extended password, which is defined as: νχ vt 之 t ”of ― Symbol. Mathematically, it is applied to all .1 company i in (n-DQ⑴ + 1 S i "Q⑴ and 0, 2 :: -2 '/ wide system vector element subscript. It extends, and the reason is When the data arrives through the extended MMA group data, # > 2 4 kh (k) transmission. The vector r (k) represents when ~ prime can be determined: subscription channel observation data The vector r⑴ 之 ～

Page 9 584992 5. Description of the invention (6) The signal received by the equation 1 at the receiver includes all user signals r⑴ plus noise. Therefore, we can define the received data vector 1 as follows: r = Σ r (k) + η · Equation 2 k = 7 λ In Equation (2), this vector η represents the noise coming in through the communication channel. Fifth, it is a system and method that is known as a linear multi-user detector. The sub-channel impulse response vector h⑴ and the extended code are used to generate a series of transmission response matrices for the mother-user. A matrix is a subscript by a subscript variable. The matrix system is arranged in a rectangular rectangle. The subscript variable is the column index and the second subscript variable is the row index & h. The system round response matrix is typically represented by Λ Α (th column, row _ Yuan & i also replaces the horse Aα) Ding Fansu represents Ai, n⑴ and is defined as:

Equation 3 The phase a_ of each row of the matrix A (k) is further specified by -f. The image transmitted by the user k during the attention period, the received data, and the appropriate filtered response. 1 Qing refers back to the fifth and sub-channel pulses (pain coupling to the extended password of all users). A⑴'s line is as follows; ^ ::. So, A⑴ contains the filtered response

Page 10

584992 V. Description of Invention (7)

A

0 0 b 〇 Equation 4 〇 Equation 5 Equation 6. A vector, Vk) is where each vector bn0o has a factor of Q⑴ + W -1 and offsets Q (k) (η -1) from the top of the matrix An (k) by the following equation. Acyclic symbol several times; b / 10 off bj⑴ and i off Jiuyuan's prime branch can be non-zero value refers to the support of this vector. Therefore, the transmission matrix of a system is generated, and the traditional transmission matrix of the overall system is as follows: The system connecting all users A = [A (1), ..., A (k), ..., A⑴], According to the adjustment technique of the previous technique, the non-zero element A of h (! 〇_ Equation 7 can be a complex number. The prime can be a complex number. Then, according to the equation 4, 5, 6, and 7, the combination matrix A in the previous technique Multi-user two-, two-to-two total system transmission response

584992 V. Description of the invention (8)

A (1) 1,1 (1) 1,2 (1) 1,3 (1) 1,4 (1) 1,50 ο ο ο ο ο ο 7 ^ ο ο ο))) υι 'οο ^ ου5 Cc C (> 8 (> 8 (LO (—.8 ίΓ8 rn ^ έέV) 1 \ — /)-3 V) 4 \ — / 5 ο ο ο πν ^ / t π ^ / * πν ^ Λ · ο ο 7 ^ 7 ^ 7 ^ 7 ^) 1) 2) 3) 4) 5 ο ο ο ο ο ο ο ο 006 · 06 · 06 > 06 »06» 0 ο ο 7 ^ ο ο ο ο ο ο ο ο ω Μ ω ^ ω Μ ω ^ ο ο ο ο ο ο ο ο ύ ύ ύ 7 ^ \ ϊκ 1 VI / 9 «VJ 3 \ 1/4 VI / 5 ο. ο ο ο ο ο ο 'U 4' u 4, (14, f οοοοοοοο έ & 7 ^ ο ο ο ο (1) 3.1 '(ι) 3,2 (1) 3,3 (ι) 3,4 (1) 3,5οοοοοοοο 7 ^ ύ ύ 7 ^ ο ο (1) 2,1 (1) 2,2 (1) 2,3 FREE2,4 (1) 2,5οοοο οοοοοοοο ο ο ο ο οοοοοοοοο (21,1 (21,2 (21 , 3 (21,4 (21,5 (21.6 (21,70 ο ο 7 ^ 7 ^ ο ο ο 0 (2) 2,1 (2) 2,2 (2) 2,3 (2) 2,4 (2) 2,5 (2) 2,6 (2) 2,7 ο ο ο οοοοοοοο 000000000000 (2) 4,1 (2) 4,2 (2) 4.3 (2) 4,4 (2) 4, 5 (2) 4.6 (2) 4,7 2) 1,2) 22) " 2) 42) 52) · 62) (3 (3 (3 (3 (3 (3 (3 (ο ο ο ο ο ο Equation 8 /; For 2 users (k), a data set (Nc = 16) has 16 chips with channel pulses Should have a length of 4 (W = 4) and a first user-extended parameter 2 (Q (2) = 4). In the overall system transmission response matrix A, bn i (k) represents the ith element of the coupled system and The channel response of the nth symbol of the kth user. The processing of the received data r uses the overall system transmission response matrix A, which represents the storage of matching filter responses to generate a matching filter output.

Pg. 12 584992 V. Description of the invention (9) “Inward” is represented by y. The matching y = AHr t is determined as the matrix M represents the Her equation of the matrix a. 9 The Hermitian shift term is defined as y = A, and the secondary product in i =) shift term. Conjugate operations on numbers. The matching cry 1 lose, ^ bar represents a multiple processing. It is derived from the system that the second table is a linearly received secondary forced zero group line equalizer (ZF-BLE) receiving 5 | series ^, and its matrix with-is designated as 0 = A „AH series— Linear Receiver J :: Receiver (MMSE-BLE). The receiver is a linear receiver Guang Qian Z 軚 The matrix is designated as 〇 = aha + om, where 〇2 memory / array. One S-the variation of noise, and 1 is the known homo-moment system and the 1 ° pair of r-matrix squares, and its main diagonal and other anisotropic algebras, the size of the two Λ-matrix selection systems, in order to advance according to the line Addition operation is omitted. Si: De-correlator (de-correlation receiver) 'matrix A is simplified by ignoring (interference ^ f to simplify). Only the extended password and its intersection are considered as de-unification. Correlation matrix, usually refers to R is often used as a receiver of type 15 of the channel. This matrix can be assumed by the above A (that is, each time the six-channel response is a pulse) assuming w =; [and dagger = 1 And construct it. In 0. ~ a parent ^ association matrix 1 ^ is the receiver structure defined by the receiver ^ ^ 1 ^ The sub-processing decorrelator is often used as a more complex multi-user Measured in the receiver of the target matrix is generated once, the multiuser detector will invert 584992

The inverse of the target matrix is then multiplied by the matched filter output vector y to produce a prediction of the data vector d, where the predicted d = 0_ly. The inverse of the target matrix is a complex and large number of operations. The number of operations required increases, such as a CDMA receiver with an unsynchronized ^ square number of 0 size. The size of 0 is very large, which makes the process of inversion impossible. Utilize the technology of direct modern numbers to reduce the complexity of inverting the target matrix. However, for some applications, these technologies may not be able to implement 0, V. Invention Description (ίο) The matrix defined as 0-1. The difference from a linear receiver is that the pIC receiver does not invert the receiver:,. Therefore, the P receiver provides another option, its complexity is lower than the line. Evening ^ users / f tester. The sixth figure is a typical conventional technique of p I C receiving / 0 receiving negative input vector r is input to a plurality of channel evaluators, which = ί estimates the secondary channel impulse response of each user. The secondary channel pulse response 2 is output to a data evaluation and interference cancellation processor, which can evaluate the received data of the three households in parallel. The evaluated received data is output-human frequency, shell material processing group for further processing. As shown in the seventh figure, the conventional techniques used in the PIC receiver evaluate the interference cancellation process. The receiver assumes that each time the channel composition is L clearly crying, the path t is from a given transmitter to a receiving two via the transmission medium. Each path [, the relative delay, the amplitude and phase are located in the connection. cry. ί 4 estimates j by the channel evaluation process shown in Figure 6 ;. For each L of each user K in the system, a clear path. So 'a total u disperser is assigned to the disperser library. Every

V. Description of the invention (11) The disperser is used from its own use to evaluate the same user's time = to generate an evaluation of the received data. The different paths of the L—complete data channel are coupled to generate data for users who are transmitting. As shown in the seventh figure, the large ratio facets (mrc) in the conventional technique are combined. Other estimates are rounded to a sign _ = to be available. The coupled data is evaluated by its turn: the = = device: it generates the estimated symbol information, delay, delay code, and the relative delay between the KL paths, and the interference cancellation processor is known. This information is used to generate the :: receive path (that is, the evaluation of the interference of r, 2,3 ... 1 ^, the L signal path of the technical / other user, and the L-1 path of the same user The received signal. The interference evaluation is subtracted from the disperser output, and the processor passes the coupled processor to generate a received data evaluation. The connection and data evaluation is again used to generate a modified interference evaluation, which is used Evaluate with the modified data of the other group. Theoretically, the procedure can be repeated indefinitely. However, the procedure is terminated after repeated two or three times. The difference between SD-PIC and HD-PIC is the symbol generation process. For SD ~ PIC, the symbol generation process generates trust information about the reception, and for HD-PIC, the symbol generation circuit does not generate trust information for the symbols that violate the reception. The difference is only about the reception. Internal processing of the data evaluation unit of the industry. Both PIC receivers can generate soft ° and Shi Gengbei $ determining symbol evaluation 'for further processing by a dedicated sub-channel data processor as shown in Figure 6. As shown in Figure 7, by placing two final losers Data symbol generator to generate the final receiver output, and is different from this page 15 584992 V. Description of the invention (12) " " ----------- 'Internal data symbol generation circuit. * The conventional art P 1C receiver existing in the received signal mode has someone] has the problem of X. The conventional art p 1C receiver assumes that each channel of the group two: ,,, L / knife path, its transmission The signal is in the transmission medium. The separation (distribution (by the coupling processor)) calculation of the channel coordination is the result of this hypothesis. However, the receiver constructed by this assumption can only be modified due to the extended sequence The non-rectangular shape is more commonly known as Multiple Access Interference (MAI). It cannot correct the interference during the transmission of the communication channel because the time extension of these symbols is between the different symbols of the user. This signal corrupts The formation is better known as Intersymbol Interference (ISI). ISI causes the phenomenon of "fat finger effect". This fat finger effect occurs when two paths from the same user have such a small relative time delay that The delay can not be distinguished into two different paths by the receiving cry. The receiver cannot be routed from any path, which affects all users and causes poor receiver operation. ^ Because of all known techniques, the PIC receiver uses the L path to simplify To separate the channels of the dissolution and coupling operations, so the two

The correctness of the receiver using a linear multi-user detector: PIC. Noise receiving signal mode-parallel interference cancellation The receiver system and method uses a receiving signal mode similar to a linear equalizer to reduce the pulse variation, and j Order should interfere. The linear equalization includes a demodulation receiver, a forced-zero receiver, and a square-difference receiver and its analogs. The invention package L of this case minimizes the averager feedback loop to modify a direct interference canceller wheel ^ T Kaizhuang, κ. The m times

584992

The process removes interference from an output symbol that matches the process. The pc receiver utilizes the received signal patterns of different sets of linear equalizers, and it does not assume that each channel of several different paths is combined. The receiver evaluates the pulse response characteristics of each channel as a whole. Therefore, the object of the present invention is to provide a system and method for receiving and coding a plurality of signals on a CDMA interface. -Another object of the present invention is to provide a PIC receiver system and method, which has higher accuracy and requires less operations. The other purposes and advantages of the system and method can be easily understood by those skilled in the art after reading the best practices of this case. Brief description of the diagram The diagram is a simple block diagram of a multi-access communication system of the conventional art. The second figure is a simple block diagram of a conventional CDMA communication system. The third picture is the simplified CDMA receiver with multiple users.

r ^ l 〇 V The fourth diagram is a simple block diagram of a multi-user debt detector of the conventional art. The fifth figure is a block diagram of a conventional multi-user detector. The sixth diagram is a system block diagram of the conventional art PI receiver. β The seventh diagram is a block diagram of the system of the PI C data evaluation and interference cancellation technique. The eighth figure is a system block diagram of a PIC receiver in this case. The ninth picture is the system block of the linear soft decision PI c receiver in this case The tenth picture is the system block of the linear hard decision PI c receiver in this case

584992 V. Description of the invention (14) * Figure. Block diagram tenth-The diagram shows the system of the non-linear soft-determined PIC receiver in this case. Please refer to the diagram for the component symbols described in the embodiment of this case. Please refer to the eighth figure. In the parallel interference cancellation method of the present invention, in a CDMA channel, after receiving, a plurality of users pass round to receive the call and receive it: a self-use input pin #, transmitted at "time"; Form, dagger 3—the input vector r contains the coupling data from each user ’s secondary channel, a channel evaluation processor 21 to derive each impulse response evaluation h⑴, and the combined response matrix eight total system uses 5 2 = estimation and interference 1 ^ 23 is eliminated to generate interference-free user data d (k) and an output 25, and each user data d⑴ 'is output in the form of an output vector from the received channel data r. The parallel interference cancellation receiver ⑽ includes a plurality of processors, which have an attached memory, a quantity and a matrix. Another embodiment can implement the functions of the gate array and DSPs of the present invention to perform various processing n. Using the total number of needles, each user ’s extended parameter Q⑴ (k = 1, 2, 3 ... κ) is the priori, transmitted by teaching or by being preloaded in the pic receiver 17 0 After demodulation, the received signal 3 ^ is input 丨 9 to the channel evaluation processor 21, where the impulse response evaluation of individual k channels is simulated as "such as the vector h⑴, to correct the inter-symbol interference (ISI), which is The MAI caused by the secondary channel's own symbol and the secondary channel symbol of all other users of the received data signal. The individual k-channel impulse response evaluation dagger

JO inch V. Description of the invention (15) is input to the first memory 29, (Equation 3) is coupled to generate an extended cryptographic matrix Aη (· 〇.) Which is used for the same user. = One system, system transmission response evaluation matrix 31, where-the total system transmission absolute η㉟ output to-the second memory input response matrix A including all system matrices A are combined. All sub-channels in the total system transmission are ; [:; Car = including lianlian zi, is information about Liyou channel, and it may contain any possible interference. The signal received in the poor material signal r received by the office is the total system transmission. The response matrix .A is canceled by 23, which measures and interferes with fr based on the received data-# ^ f. 4 ^ * d, > ^ ί £ ^: 3: 3: 3; 3K such as the inserter, vi The terb i decoder and its analogs. 2 3 — The canceler 23 is shown in the ninth figure. The data is referred to as = ί matching filter 35, whose 忒 input of the matching filter is 1 9 data vector r to generate phase The matching filter rounds out the vector of y. An adder 37 is used to remove the feedback% disturbance c from the output y of the matching filter 35. The direct interference canceller 38 is used to obtain the user data ^ d (k > 1 evaluation, a repetition counter / converter 41, a feedback interference processor 43, two and one pay generation | § 4 5 for self evaluation User data d⑴ combined symbols. In order to obtain user data d (k) of a specific user from the coupled user data r, the user data r must use a matching adapter

Page 19 584992

35 or similar. The device 35 of this technique needs a response feature, so that if it matches the filter's secondary rhyme, it is the first-order pepper plastic mold, the extension pulse shape and the use of the province-of-people-frequency pulse response coupled to the human peak. 1 右 冲 # π Complex conjugate number to give 51% before transmission. No. r input 19 to the filter 35 does not match the given response characteristics to produce a lower output. Swallow-matched matching transition device p ^ C; Shiri V, Shiren 丄 M 々 粁 Suzuki, Jing please, room tube: 35 is similar to the payment performed by the linear multi-user receiver. ， 滤 运 #The matching filter 3 5 operations such as square 々: q: Description ′ This extension is not the same as the conventional art of parallel interference cancellation receiver ^. The input user data r conforms to the extended password v (k) and the channel impulse response hoo of each channel k. Each element of the output vector y of the matching filter 35 is a data vector d of the pass One of the corresponding symbols in the first rough evaluation. ~, The general system transmission response matrix A provides the response characteristics of the matching clear number 3 $. Each row of the total system transmission response matrix A is A vector representing the response characteristics of a particular symbol. The received data vector: is input to the phase 2 & filter 35, where it conforms to each response characteristic of the response matrix A transmitted from the total system to produce The matching transitioner output vector y. Each element of the output vector y corresponds to a preliminary evaluation of a specific symbol transmitted by a given user. The matching filter M output vector y is input to the Direct interference canceller 39. The direct interference canceller 3 9 performs part of the interference cancellation operation on the matching filter 35 out of the vector y. This operation may be a carry operation or a more complicated trick. A a part The interference cancellation vector y is output as the data symbol evaluation d, and is really input to ^ via a repeat / counter conversion 41a.

584992 V. Description of the invention (17); ------ feedback interference processor 43. The feedback interference processor 43 uses the direct interference canceller 39 to output the evaluation,... 乂. The interference output is reached as a vector c, which is not deleted by the direct interference canceller = first. The interference evaluation C is subtracted from the matched filter 35 = outward y. The result Z is the output direction y of the matched filter 35 minus the interference evaluation c. This repeated interference reduction process can be repeated m times' depending on the desired signal correction level. After m iterations, the interference is removed from the matched filter 35, and the repetition / counter conversion 41 is rotated 41b to output d at the final output sign to produce 45. The calculation of the negative feedback control of the direct interference canceller 39 / feedback interference processor 43 is as shown in the ninth figure, which represents a m-times repeated receiver. For example, 'if m = 2, the pic receiver 17 has repeated the cancellation procedure twice. The interference vector output c (m) by the feedback interference processor 43 and the symbol evaluation vector output d (m) by the direct interference canceller 39 are repeated for mth, and the equation 1 0 c (m) = Td (ml) Equation 11 ^ the direct interference canceller 39 executes the matched filter output σ multiplied by the matrix S ′ and the feedback interference processor 4 3 executes the symbol

This vector d (m) is repeated at the last m times. As is known in the art, the output of the direct interference canceller 39 is processed by the final output; the interference cancellation receiver, the information of this output is evaluated in the output symbol: at: 45 'its f hard or soft determines the end view In accordance with the present invention, the needs of the system and the system are required. The number of repetitive m from the dried fish ρ ,, ^ γ, and the evaluation of this data is divided by ~ 卞 镘 4 i§d (m) can be written as w d⑽ = OST + 〇 + (一 z) -1 Equation i 2 where the steady-state response is ^ ss ~ (ST + I) Sy Equation 1 3 and the temporary response is 孑 ⑽ = (-1) 157 ^ 0 ^ + /)-1 Equation 14 When the number of repetitions increases, If the temporary response approaches zero, the pic receiver 17 converges to the steady state response. When this happens, the receiver converges to a stable state as described in Equation 13, which is explained in the technical paper of Α · Reznik, titled the new type of PIC multi-user receiver and its

Page 22 584992 V. Description of the invention (19) The relationship between ZF-BLE and MMSE-BLE multi-user receiver, this reference is the internal report of October 19, 1999 IterDigita1 Communication Corporat i on. Known techniques The steady state response of linear receivers such as ZF-BLE, MMSE-BLE and other de-correlator is defined as' d = 0 ^ ly Equation 1 5 where 0 is the matrix of the standard. Please refer back to equations 1 3 and 15; if the matrices S and T are selected, then (ST + I)! S-, and if the receiver I defined in the equation and η converges, it will converge to the receiver structure. The defined linear receiver. The linear algebra must be such that (ST + 〇-is =: 〇-1, matrices s, T, and 0 must satisfy the following definitions:

T S-1 Equation 1 6 Equation private 16 divides the receiver structure 0 to two separate matrices τ and Y! Is not to transform the receiver structure. The matrix τ defines the feedback interference processor 43. The matrix S (the inversion of the matrix S 1) defines the direct interference canceller 39. According to the present invention, in the feedback loop, the inversion of the matrix 0 is reset by the inversion of another matrix and a series of matrix multiplications. One advantage of the invention 17 is that the transformation complexity of the matrix S-1 is lower than the original receiver structure. For example, the matrix can be a diagonal matrix (only in the main

Page 23 J54992 V. Invention ^ Γ (20) Involving the matrix of main = zero input) ... The inversion of the diagonal matrix only involves the inversion of each individual element in the diagonal. The line should improve the performance of the PIC17, the main diagonal of the matrix T is zero. As shown in the reference article of A. Reznik and the receiver = the light-combination matrix and the two formulas of 3, the general PIC connection linear connection form is generated. ―One of the receiver structures Q mentioned r receiver, matrix S is defined as Equation 1 7 = diag (O)

Define a matrix where the input of the main diagonal is equal to X, the line turns in 'and all other elements of the matrix are equal to zero. Use the solution matrix of Equation 16 to generate τ = 0-S, 1 Equation 1 8 Since the direct interference canceller 39 performs z (m) multiplied by the matrix § (which is the inversion of diag (O)), the canceller 39 performs the The carry of each individual element of the vector z (m). The matrix of d (m) is multiplied by the matrix τ and the feedback interference processor 43 'calculates the interference element. A receiver including this structure refers to a parallel interference cancellation receiver with elementary interference cancellation in the feedback or a p 丨 C-f I receiver. μ For a system requiring a forced zeroing mechanism, the receiver must

Page 24 584992 V. Description of the invention (21) Convergence to the ZF-BLE linear receiver. The ZF-PIC-fl of the present invention 17 uses a ZF-BLE receiver structure. 0 = AHA. Therefore, the matrices S and T are defined as S'1 = diag (AHA) Equation 19 and T 2 AHA-diag (AHA) Equation 20 For a system that requires a minimum mean squared difference reception mechanism, the receiver must converge To this MMSE-BLE linear receiver. The MMSE-PIC-fl receiver of the present invention 17 uses the MMSE-BLE receiver structure 0 = ahA + σ2I. Therefore, the matrices s and τ are defined as S1 = diag (AHA) + Equation 21 and T = A «A-diag (A« A) Equation 22 For a system that requires a disassociated receiver, the total system response matrix A It is combined in the channel evaluator processor 21 and is combined into a cross-correlation matrix that ignores the beta channel effect. The structure of the receiver is the same as that of the ZF-Pic-f I structure described in the month 'J, but using a modified version of the a matrix. Those skilled in the art can understand the construction of other PIC receivers with complete interference cancellation in the feedback loop (PIC-f I receiver). This system and the method of the invention 17 can be used to lightly integrate all existing Linear, receiver mode. The two linear receiver modes, ZF — BLE and 〇3 £, have been exemplified in the example. Using the method of the present invention, a linear receiver is preferentially selected to determine convergence. Another embodiment of the invention 1 7 uses the same system structure 2 3

Page 25 584992, Description of Invention (22) The elimination of internal symbol interference (I SI) is entrusted to the direct interference canceller 39. The feedback interference processor 43 is used to eliminate multiple access interference (MAI). This embodiment refers to one of the parallel interference cancellers with direct ISI cancellation (pic-dlSI). As described in A. Reznik's reference article, this method is more complex than pic-II, but it can provide improved performance. For a system that requires a forced zeroing mechanism, the receiver must converge to a ZF-BLE linear receiver. Utilizing the system and method of the present invention, the receiver is a ZF-PIC-dlSI receiver whose S and T matrices are defined as 0 Equation 2 3 and ^ (Κ) T = AHA -s-1 for A minimum mean square difference receiver is required to converge to the MMSE-βΐ ^ Ε: system and method. The receiver refers to a system with S and τ matrices defined as one of the receiving mechanisms of Equation 2 4. receiver. Using the MMSE-Pic-dISI receiver of one of the seventeenth aspects of the present invention,

584992 V. Description of the invention (23) 0 s (k) H λ (k) Ο (Κ) Η The construction of equation 25 and equation 26 uses the receiver mode. In both embodiments, the PIC-dISI ′ -linear receiver is

T = ΑΗΑ -S'1 + σ2 I In general, the total system response ^ 2 1 is combined. The structure of the receiver 'but the two A matrixes need a decorrelator receiver — the response matrix A is combined in the channel evaluator and processed in a cross correlation matrix that ignores the channel effect is the same The described 2 ′-? 1 (:-( 1131 structure modified version. Other PIC receivers with direct ISI cancellation Invention 17 System and method Lightweight existing linear linear receiver modes ZF-BLE and MMSE The actual receiver structure is shown. As the PIC-f I receiver is preferred to determine convergence. The receiver described above can converge well when the level of interference is low. Many systems, such as accurate power control UMTS 3rd Generation Broadband CDMA, "" Knife with Duplex (FDD) "shows low interference level. As described before

584992 V. Description of the invention (24) The invention 17 is not limited to the receiver described. Any choice of matrix 8 provides a viable receiver structure. As a receiver structure, 0 'Equation 16 defines any number of receiver structures, which converges to a linear receiver defined by the receiver structure 0. Different choices for s Different choices for the complexity and performance of a desired receiver. When should the performance of this canceller 39 be developed? 1 (> 01151 receiver, you can get 1 more performance than J performance. However, the direct interference canceller 39 is better than the complex user inversion of more complex matrix to increase the receiver mode, it must Reverse the structure of the receiver ... Hard to determine and determine the description of the parallel interference cancellation receiver. Heart to talk about the interference cancellation receiver 0, to form the feedback path sum of the hard decision 49, and add a pay number generator. As shown in Fig. 11, the nonlinearity, s, and s are evaluated as d (m) by using a non-linear soft-determined symbol. The interference cancellation receiver 51 can be replaced by the device shown in Fig. 11. The symbol generator 49, although the present invention has been covered by i ::::: mentioned in the scope of the present invention in the more favorable scope, but it can be understood as the applicant. / Cross, for those who are familiar with this technology, page 28 584992 Brief description of the diagram 17 receiver 19 input 21 channel evaluation processor 2 3 interference canceller 2 5 output 29 first memory 31 second memory 3 3 !, 332, 333… 33κ: times 35 matching filtering 32: 3S 37 addition Generator 38 direct interference canceller 41 repeat counter / converter 41a: repeat / counter conversion 43

Page 29

Claims (1)

584992 The μ cancel ... f interference canceller (17) in a receiver, which separates the received communication signal (r) from a CDMA bismuth μ multiple transmitters, into a plural one slave signal (d (k )), The interference canceller (17) is further characterized by: a two-frequency evaluation processor (21) for receiving the communication signal (r), which: other impulse response evaluations (A) are in a plurality of locations The desired signal (d⑴), = impulse response evaluation is coupled to a data evaluation and interference canceller (2 3); ▲ f material sweat estimation and interference canceller (2 3) receives the communication signal (r) and the pulse response f Evaluate (A) and output the plurality of desired signals (d⑴)). • If the interference canceller (丨 7) of the first scope of the patent application, the channel evaluation processor (2 1) is further characterized by: / a frequency evaluation unit (2 7) is coupled to the communication signal (r) The output channel impulse response is evaluated (h⑴) on the plurality of desired signals ㈠⑴), which channel is estimated? § (2 7) is coupled to a system response matrix combiner (2 9); the system response matrix combiner (2 9) outputs the system response matrix (A⑻⑻) to the plurality of desired signals (d⑴); and the The system response matrix (A⑻⑻) is combined in a total system response matrix (A) output. 3 · If the interference canceller (1 7) of the scope of the patent application, the data evaluation and interference canceller (23) further features: a matching filter, the filter (3 5) is lightly connected to the communication The signal (Γ) and the channel evaluation processor (21) output (A) to output the evaluation of the desired signal with a selective signal (y), to the first (+) of a totalizer (3 7) John input; the totalizer 37 has an output (z (m)) coupled to a direct interference cancellation Page 30 584992 Case No. 90121091 ΠΠ.-X month month day amendment VI. Patent application range device (39); The direct interference canceller (3 9) is a selective signal (d) of a plurality of desired signals (d⑴) (m)) output the desired signal in a scaled evaluation to an input of a repeating counter (41); the repeating counter (41) has a first output (41a) coupled to a feedback interference processor ( 43); the feedback interference processor (4 3) outputs the interference evaluation (c (m)) of the selective signal to the totalizer (37)-the second input (-); and the interference evaluation (c ( m)) is obtained by subtracting m repetitions from the output (y) of the matching filter (3 5) by the repetition counter (4 1), and the repetition counter (4 1) passes through a second output ( 4 1 b), output the evaluation of the desired signal (d (m)) as a plurality of desired signals (d ⑴) which are non-selective signals. 4. If the interference canceller of item 3 of the patent application scope ( 17), further characterized by the direct interference canceller (3 9) operation defined by an S matrix, and borrowing A defined matrix T of the feedback interference processor (43) operation, which is the relationship between two 0 T + S-1 0-based matrix wherein a target matrix that defines a receiver structure. 5. If the interference canceller (1 7) of item 4 of the scope of patent application is further characterized by the definition of the matrix S as S 2 (diag (O) wide 1 and the definition of the matrix T as T 2 0- diag ( O) 〇 Page 31, 584992, case number 90121091, year-by-month repair, JL ·% amendment VI, patent application scope 6 · If the interference canceller (17) of the patent application scope item 5 is further characterized by the structure of the receiver 0 A forced equalization zero linear equalizer. 7 · If the interference canceller (17) of the scope of the patent application is applied, it is further characterized that the receiver structure 0 is a minimum average square difference group linear equalizer. 8 · If the interference canceller (1 7) of the scope of the patent application is applied, it is further characterized in that the total system response matrix (A) is a decorrelator, and the receiver structure 0 is a forced equalization zero group linear equalization Device. 9 · If the interference canceller (1 7) in the fourth item of the patent application scope is further characterized by the definition of the matrix S, the direct interference canceller (3 9) completes the internal symbol interference (ISI) of each selective signal ) And the definition of the matrix T, the feedback interference processor (4 3) calculates the multiple access interference (MAI) that becomes a factor of each selective signal by the desired signal (d (k)). ) 〇1 0 · If the interference canceller (1 7) of item 9 of the patent application scope, it is further characterized in that the definition of the matrix S is 0 A (l) f / A (1) ^ (k) H ^ (k ) 0 A (K) H ^ (K) Page 32 584992 Amendment No. 90121091 6. The scope of the patent application and the definition of the matrix T are T = 0-S · 1 〇1 1 · If the interference canceller (1 7) of the 10th scope of the patent application, it is further The feature is that the receiver structure 0 is a forced zero-zero linear equalizer. 1 2 · If the interference canceller (17) in the scope of the patent application is No. 10, it is further characterized in that the total system response matrix (A) is a decorrelator, and the receiver structure 0 is a forced zeroing group Linear equalizer. 1 3 · If the interference canceller (1 7) of item 9 of the patent application scope, it is further characterized in that the matrix S is defined as Ο 5 + σ2 / Ο Page 33 584992 Case No. 90121091 3 倏 Year of the year 'J 5 supplementary, 32. 11. Amendment of the year 6, the scope of the patent application is characterized by a hard decision symbol generator (4 9) coupled between the repetition counter (41) The generation of the first output (41a) and the input of the feedback interference processor (43) is hardly determined between the desired signal evaluation (d (m)). 1 5 · If the interference canceller (1 7) of the scope of application for the patent, it is further characterized that the definition of the matrix S is S = (diag (O))-1 and the definition of the matrix T is T = 0 -diag (0) ° 1 6 · If the interference canceller (17) of the 15th item of the patent application scope, it is further characterized in that the receiver structure 0 is a forced zero return group linear equalizer. 1 7. The interference canceller (17) according to item 15 of the scope of patent application, further characterized in that the receiver structure 0 is a minimum average square difference group linear equalizer. 1 8. The interference canceller (17) according to item 15 of the scope of patent application, further characterized in that the total system response matrix (A) is a decorrelator, and the receiver structure 0 is a minimum mean squared difference Set of linear equalizers. 19. If the interference canceller (17) of item 14 of the scope of patent application is further characterized by the definition of the matrix S, the direct interference canceller (39) completes the internal symbol interference of each selective signal The elimination of (ISI) and the definition of the matrix T make the feedback interference processor (4 3) calculate the multiple access interference that becomes a factor of each selective signal by the desired signal (d (k)) (MAI 20. If the interference canceller (17) of item 19 of the scope of patent application, it is further characterized by the definition of the matrix S as Page 34 584992 Case No. 90121091 VI. Scope of Patent Application Amend s ^ (k) H ^ (k) 0 0 ^ {Κ) Η ^ {Κ) and the definition of the matrix τ is T = ΑΗΑ-S-1. 2 1 · If the interference canceller (17) in the scope of the patent application is No. 20, it is further characterized in that the receiver structure 0 is a forced zero return group linear equalizer. 2 2 · If the interference canceller (1 7) of the 20th patent application scope, it is further characterized in that the total system response matrix (A) is a decorrelator, and the receiver structure 0 is a minimum mean square difference Set of linear equalizers. 2 3 · If the interference canceller (1 7) of item 19 of the scope of patent application, it is further characterized by the definition of the matrix S as 0 S + σ2 / Ο i (K) H λ (κ) Page 35 584992 Amendment No. 90121091 VI. Scope of patent application The definition of the matrix T is T = AHA-SM + σ2Ι; and the definition of the receiver structure 0 is a minimum average square difference group linear equalizer. 24. The interference canceller (17) according to item 4 of the scope of patent application, further characterized by a non-linear soft decision symbol generator (53) coupling the first output (41a) between the repeat counter (41) Interference with the input of the feedback processor (43) is hardly determined between the desired signal evaluation (d (m)). 2 Pan. If the interference canceller (1?) Of the 24th scope of the patent application, it is further characterized by the definition of the matrix S as S = (diag (O))-1 and the definition of the matrix T as T 2 0 -diag (0 2 26. If the interference canceller (1 7) of the 25th item of the patent application scope, it is further characterized that the receiver structure 0 is a forced zero return group linear equalizer. 2 7. The interference canceller (17) of the 25th item of the scope is further characterized in that the receiver structure 0 is a minimum average square difference group linear equalizer. 2 8. The interference canceller of the 25th item of the patent application scope ( 17), which is further characterized in that the total system response matrix (A) is a decorrelator, and the receiver structure 0 is a forced zero-zero linear equalizer. 2 9. As described in item 24 of the scope of patent application The interference canceller (1 7) is further characterized by the definition of the matrix S such that the direct interference canceller (3 9) completes each Page 36 584992 92. 11 · -3 Amendment 584992 92. 11 · -3 Amendment Amendment No. 90121091 Sixth, the patent application scope-the elimination of internal symbol interference (ISI) of a selective signal, and the definition of the matrix T makes The feedback interference processor (4 3) uses the desired signal (d (k)) to calculate the multiple access interference (MAI) which becomes a factor of each selective signal. The interference canceller of item 9 (1 7) is further characterized in that the definition of the matrix S is 0 〇 and the definition of the matrix τ is Τ ΑΗΑ-S · 31. For example, the interference canceler of the 30th item in the scope of patent application ( 17), which is further characterized in that the receiver structure 0 is a forced equalization-zero group linear equalizer. 3 2. If the interference canceller (1 7) of the 30th scope of the patent application, it is further characterized in that the total system response matrix (A) is a decorrelator, and the receiver structure 0 is a minimum mean square difference Set of linear equalizers. 3 3 · If the interference canceller (17) in the scope of patent application No. 29, it is further characterized by the definition of the matrix S as Page 37 584992 Case No. 90121091 Amendment VI. Patent application scope 0 S + σ * 7 0 ^ (Κ) Η ^ (Κ) \ \ The definition of the matrix T is T = ΑΗΑ-S _! + Σ2Ι; and the receiving The definition of the equalizer structure 0 is a minimum average square difference group linear equalizer. 34. A method (17) for separating a received communication signal (r) from a plurality of transmitters into a plurality of desired signals (d (k)) on a CDMA interface, characterized by · a) A total system response matrix (A) is generated from the impulse response evaluation (h (k)) of the received signal (r); b) filtering the received signal (r) is based on the total system response matrix (A), which Selective signal (y) is used to generate the desired signal evaluation; c) Select a receiver structure (0) for convergence; d) Obtain an S matrix from the receiver structure; e) Obtain from the receiver structure A T matrix; f) by multiplying the matrix S, arranging the filter output (y) in proportion to the desired signal evaluation (d (m)); g) calculating the interference evaluation (c (m)), By multiplying the scaled output (d (m)) and the matrix T; Page 38 584992 Case No. 90121091 92. η.-Ϋ Month and month:; ,, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year, year; (C (m)); i) by multiplying by the matrix S, the filter output is proportionally reduced minus the interference evaluation (z (m)) as the desired signal evaluation (d (m)); j) repeat Steps (g) to (i) are repeated until m times; and k) outputs the desired signal evaluation (d (m)) as a plurality of desired signals (d (k)) without the selective signal. 35. The method (1 7) according to item 34 of the scope of patent application, wherein step d) is further characterized by the definition of the matrix S as S = (diag (O))-1 and step e) is further characterized by The definition of the matrix T is T = 0- diag (O) 〇 step c) into the linear mean step c) into the squared difference step a), and a forced reduction step f) and g) ratio The arrangement is eliminated by 36. The method (1 7) as described in item 35 of the scope of patent application, wherein the step-by-step feature is to select the receiver structure (0) as a forcible return-to-zero scale. 37. The method (17) according to item 35 of the scope of patent application, wherein the step-by-step feature is to select the receiver structure (0) as a minimum average linear equalizer. 38. If the method (1 7) of item 35 of the scope of patent application, wherein the step-by-step feature is to generate the total system response matrix (A) as a solution step c), the further feature is to select the receiver structure (0 ) Is a zero-group linear equalizer. 3 9 · If the method (1 7) of item 34 of the scope of patent application, further features of which are defining the matrix S, so that the picture completes the internal symbol interference (I S I) of each selective signal. Page 39 584992 Case No. 90121091 Supplement 1 from year 11 Amendment VI. Patent application scope and definition of the matrix T, so that the calculation of interference assessment is calculated by the desired signal (d (k)) into each selectivity Multiple access interference (MAI) of signal factors 〇 4 0 · As for the method (17) of the 39th scope of the patent application, where step f) is further characterized by the definition of the matrix S as 0 A ⑴ η A ( 1) The definition of the matrix T in SA {k) H a (k) 0 A {K) H ^ {K) and step g) is T = 0-S · 1. 4 1 · The method (17) of item 40 in the scope of patent application, wherein step c) is further characterized in that the receiver structure (0) is selected as a forced zeroing linear equalizer. 4 2 · If the method (17) of item 40 of the scope of patent application, the further feature of step c) is selecting the total system response matrix (A) as a decorrelator, and the further feature of step c) is selecting The receiver structure (0) is a forced zero-zero linear equalizer. 4 3 · The method (17) of item 39 in the scope of patent application, wherein the matrix S is further characterized in step f) as Page 40 584992 Case No. 90121091 vllt Amendment 6. The scope of patent application is strict A (l) 0 5 VV ^ (Κ) Η ^ (Κ) νσΊ The further feature of this matrix T in step g) is T = AHA-SM + σ2I; and the receiver structure 0 is a minimum average square difference group linear equalizer. 4 4 · The method (17) of item 34 in the scope of patent application, wherein step g) is further characterized by generating a hard signal decision (d (m)) which is determined by the desired signal. 45. The method (17) of item 44 of the scope of patent application, wherein the further feature of step d) is that the matrix S is defined as S two (diag (O))-1 and the further feature of step e) is the matrix T The definition is T = 0-diag (O) 〇 4 6 · If the method (17) of the 45th scope of the patent application, the further feature of step c) is to select the receiver structure (0) as a forced return Zero-group linear equalizer. 47. The method (1 7) according to item 45 of the scope of patent application, wherein step c) is further characterized by selecting the receiver structure (0) as a minimum average square difference group linear equalizer. 4 8 · If the method (1 7) of item 45 in the scope of patent application, wherein a further feature of step a) is to generate the total system response matrix (A) as a decorrelator, and Page 41 584992 Amendment No. 90121091 6. Scope of Patent Application Step c) is further characterized in that the receiver structure (0) is selected as a forced zero-zero linear equalizer. 49. The method (17) of item 44 of the scope of patent application, wherein steps f) and g) are further characterized by defining the matrix S so that the proportional arrangement completes the internal symbol interference of each selective signal ( ISI), and define the matrix T, so that the calculated interference evaluation calculates the multiple access interference (MAI) that becomes a factor of each selective signal by the desired signal (d (k)) ° 5 0 · The method (17) of item 49 in the scope of patent application, wherein the further feature of the matrix S in step f) is ⑴ 0 ^ (K) HA (K) 0 and the further progress of the matrix T in step g) Features are T = 0-S-1. 5 1 · The method (17) of item 50 of the scope of patent application, wherein step c) is further characterized in that the receiver structure (0) is selected as a forced-zeroing group linear equalizer. 5 2 · The method (1 7) of item 40 in the scope of patent application, wherein the step c) is further characterized by generating the total system response matrix (A) as a decorrelator, and Page 42 584992 Case No. 90121091 Fanjiang-3 Amendment HI Supplement Amendment 6. The scope of the patent application is further characterized by selecting the receiver structure (0) as a forced-zero group equalizer. 5 3 · The method (1 7) according to item 4 of the scope of patent application, wherein the further feature of the matrix S in step f) is 0 A ⑴η A (1) S ^ (k) αΊ ^ (K) H ^ {K) 0 The matrix T is further characterized in step g) as T = AHA-SM + σ2 1; and the receiver structure 0 is a minimum average square difference group linear equalizer. 5 4 · The method (17) of item 34 of the scope of patent application, wherein step g) is further characterized by generating a non-linear soft decision based on the desired signal evaluation (d (π〇). 5 5 · As applied The method (17) of item 54 of the patent, wherein step d) is further characterized by the definition of the matrix S as S = (diag (O))-1 and step e) is further characterized by the definition of the matrix T for Page 43 584992 Case No. 90121091 Amendment VI. Patent application scope T 0-diag (O) 5 6 · If the method (17) of the patent application scope item 5 5 is adopted, the further feature of step c) is selecting the receiving The equalizer structure (0) is a forced zero-zero linear equalizer. 57. The method (17) of item 55 in the scope of patent application, wherein step c) is further characterized by selecting the receiver structure (0) as a minimum average square difference group linear equalizer. 5 8 · If the method (17) of the 55th item of the patent application scope, wherein the step c) is further characterized by generating the total system response matrix (A) as a decorrelator, and the step c) is further characterized by selection The receiver structure (0) is a forced zero-zero linear equalizer. 59. The method (17) of item 54 of the scope of patent application, wherein steps f) and g) are further characterized by defining the matrix S so that the proportional arrangement completes the internal symbol interference of each selective signal ( ISI), and define the matrix T, so that the calculated interference evaluation calculates the multiple access interference (MAI) that becomes a factor of each selective signal by the desired signal (d (k)). 0 6 0 · The method (17) of item 59 in the scope of patent application, wherein the matrix S is further characterized by step A) as A (1) ^ / / 4 (D 0 0 Page 44 584992 Amendment No. 90121091 6. The scope of patent application and the further feature of the matrix T in step g) is T-S -1 6 1 · Method (1 7) of item 60 of the patent scope, where Step c) is further characterized in that the receiver structure (0) is selected as a forced zero-zero linear equalizer. 6 2 · The method (1 7) of item 60 in the scope of patent application, wherein the further feature of step c) is selecting the total system response matrix (A) as a decorrelator, and the further feature of step c) is selecting The receiver structure (0) is a forced zero-zero linear equalizer. 6 3 · The method (17) according to item 5 of the scope of patent application, wherein the further feature of the matrix S in step f) is 0 Α (Κ) Η ^ (Κ) A⑴⑴ S ^ {k) H ^ { k) Ο Page 45 Page 46