MIMO WIDEBAND RECEIVER AND TRANSMITTER, AND METHOD THEREOF

Abstract

In aspect, the disclosure includes a method of configuring a MIMO wideband receiver. The method would include estimating, on a SISO basis, a set of post-processing parameters for a plurality of receiver channels; receiving, by each of the plurality of receiver channels, a first test signal which is transmitted from a first transmitter channel on a MIMO basis; calculating a first set of crosstalk parameters in response to receiving the first test signal; receiving, by each of the plurality of receiver channels, a second test signal which is transmitted from a second transmitter channel on the MIMO basis; calculating a second set of crosstalk parameters in response to receiving second test signal; and calculating the set of post-processing parameters based on the first set of crosstalk parameters and the second set of crosstalk parameters by cancelling a crosstalk interference among plurality of receiver channels.

Claims

1. A method of configuring a multi-input multi-output (MIMO) wideband receiver comprising: estimating, on a single-input and single-out (SISO) basis, a set of post-processing parameters for a plurality of receiver channels; receiving, by each of the plurality of receiver channels, a first test signal which is transmitted from a first transmitter channel on a MIMO basis; calculating a first set of crosstalk parameters in response to receiving the first test signal; receiving, by each of the plurality of receiver channels, a second test signal which is transmitted from a second transmitter channel on the MIMO basis; calculating a second set of crosstalk parameters in response to receiving second test signal; and calculating the set of post-processing parameters based on the first set of crosstalk parameters and the second set of crosstalk parameters by cancelling a crosstalk interference among plurality of receiver channels.

2. The method of claim 1, wherein estimating, on the SISO basis, the set of post-processing parameters for the plurality of receiver channels comprising: estimating a second post-processing parameter and a third post-processing parameter only between the first transmitter channel and the first receiver channel; switching from between the first transmitter channel and the first receiver channel to between the second transmitter channel and the second receiver channel; and estimating a first post-processing parameter and a fourth post-processing parameter only between the first transmitter channel and the first receiver channel, wherein the set of post-processing parameters comprising the first post-processing parameter, the second post-processing parameter, the third post-processing parameter, and the fourth post-processing parameter.

3. The method of claim 1, wherein receiving, by each of the plurality of receiver channels, the first test signal which is transmitted from the first transmitter channel on the MIMO basis comprising: receiving, by a first receiver channel of the plurality of receiver channels, the first test signal which is transmitted from the first transmitter channel on the MIMO basis while not receiving from the second transmitter channel; grounding the second receiver channel; receiving, by a second receiver channel of the plurality of receiver channels, the first test signal which is transmitted from the first transmitter channel on the MIMO basis while not receiving from the first transmitter channel; and grounding the first receiver channel.

4. The method of claim 3, wherein calculating the first set of crosstalk parameters in response to receiving the first test signal comprising: obtaining a first crosstalk parameter and a second crosstalk parameter based on the first test signal received by the first receiver channel; and obtaining a third crosstalk parameter and a fourth crosstalk parameter based on the first test signal received by the second receiver channel, wherein the first set of crosstalk parameters comprising the first crosstalk parameter, the second crosstalk parameter, the third crosstalk parameter, and the fourth crosstalk parameter.

5. The method of claim 1, wherein receiving, by each of the plurality of receiver channels, the second test signal which is transmitted from the second transmitter channel on the MIMO basis comprising: receiving, by a first receiver channel of the plurality of receiver channels, the second test signal which is transmitted from the second transmitter channel on the MIMO basis while not receiving from the first transmitter channel; grounding the second receiver channel; receiving, by a second receiver channel of the plurality of receiver channels, the second test signal which is transmitted from the second transmitter channel on the MIMO basis while not receiving from the first transmitter channel; and grounding the first receiver channel.

6. The method of claim 5, wherein calculating the second set of crosstalk parameters in response to receiving the second test signal comprising: obtaining a fifth crosstalk parameter and a sixth crosstalk parameter based on the second test signal received by the first receiver channel; and obtaining a seventh crosstalk parameter and an eighth crosstalk parameter based on the second test signal received by the second receiver channel, wherein the second set of crosstalk parameters comprising the fifth crosstalk parameter, the sixth crosstalk parameter, the seventh crosstalk parameter, and the eighth crosstalk parameter.

7. The method of claim 5, wherein calculating the set of post-processing parameters based on the first set of crosstalk parameters further comprising: estimating the first crosstalk parameter and the second crosstalk parameter based on a least square technique.

8. The method of claim 6, wherein calculating the set of post-processing parameters based on the second set of crosstalk parameters further comprising: estimating the fifth crosstalk parameter and the sixth crosstalk parameter based on a least square technique.

9. The method of claim 1, further comprising: determining whether the set of post-processing parameters cancel out crosstalk among the plurality of receiver channels.

10. The method of claim 1, wherein the first test signal and the second test signal are different quadrature phase shift keying (QPSK) training sequences.

11. A method of configuring a multi-input multi-output (MIMO) wideband transmitter comprising: transmitting on a MIMO basis, through a first transmitter channel of a plurality of transmitting channels, a first test signal to be received by a first receiver channel; transmitting on the MIMO basis, through a second transmitter channel of the plurality of transmitting channels, a second test signal to be received by a second receiver channel; determining, a first received signal received by the first receiver channel and determining a second received signal received by the second receiver channel; estimating, a set of coupling parameters for the plurality of transmitter channels based on the first received signal and the second received signal; and calculating, based on the set of coupling parameters, a set of pre-processing compensation parameters by cancelling a crosstalk interference among the plurality of transmitter channels.

12. The method of claim 11, wherein transmitting by the first transmitter channel the first test signal to be received by the first receiver channel and transmitting by the second transmitter channel the second test signal to be received by the second receiver channel occur simultaneously.

13. The method of claim 11, wherein the first test signal and the second test signal are different quadrature phase shift keying (QPSK) training sequences.

14. The method of claim 11, wherein estimating the set of coupling parameters is based on a least square technique.

15. The method of claim 14, wherein estimating the set of coupling parameters comprising: determining the first received signal and the first received signal by setting the set of pre-processing compensation parameters to zero.

16. The method of claim 11, further comprising: determining whether the transmitter has cancelled the crosstalk interference among the plurality of transmitter channels by applying the pre-processing compensation parameters to a processor of the transmitter.

17. The method of claim 14, wherein estimating the set of coupling parameters further comprising: assuming the first receiver channel and the second receiver channel as an ideal receiver.

18. The method of claim 16, wherein the pre-processing compensation parameters are applied to the processor of the transmitter only once.

19. A multi-input multi-output (MIMO) wideband receiver comprising: a wireless receiver comprising a plurality of receiver channels comprising a first receiver channel and a second receiver channel; and a processor coupled to the wireless receiver and configured to: estimate, on a single-input and single-out (SISO) basis, a set of post-processing parameters for the plurality of receiver channels; receive, by each of the plurality of receiver channels, a first test signal which is transmitted from a first transmitter channel on a MIMO basis; calculate a first set of crosstalk parameters in response to receiving the first test signal; receive, by each of the plurality of receiver channels, a second test signal which is transmitted from a second transmitter channel on the MIMO basis; calculate a second set of crosstalk parameters in response to receiving second test signal; and calculate a set of post-processing parameters based on the first set of crosstalk parameters and the second set of crosstalk parameters by cancelling a crosstalk interference among the plurality of receiver channels.

20. A multi-input multi-output (MIMO) wideband transmitter comprising: a wireless transmitter comprising a plurality of transmitter channels comprising a first transmitter channel and a second transmitter channel; and a processor coupled to the wireless transmitter and configured to: transmit on the MIMO basis, through the first transmitter channel, a first test signal to be received by a first receiver channel and simultaneously transmitting, through the second transmitter channel, a second test signal to be received by a second receiver channel; determine, a first received signal received by the first receiver channel and determining a second received signal received by the second receiver channel; estimate, a set of coupling parameters for the plurality of transmitter channels based on the first received signal and the second received signal; and calculate, based on the set of coupling parameters, a set of pre-processing compensation parameters by cancelling a crosstalk interference among the plurality of transmitter channels.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The accompanying drawings are included to provide a further understanding of the disclosure, and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0014] FIG. 1 shows the above described concept in a flow chart according to one of the exemplary embodiments of the disclosure.

[0015] FIG. 2 shows the hardware diagram of a transmitter and a receiver according to one of the exemplary embodiments of the disclosure.

[0016] FIG. 3 shows a simplified conceptual diagram of an overall MIMO wideband transceiver system architecture according to one of the exemplary embodiments of the disclosure.

[0017] FIG. 4 illustrates a derivation of a conceptual model of a MIMO transmitter architecture according to one of the exemplary embodiments of the disclosure.

[0018] FIG. 5 illustrates a derivation of a conceptual model of a MIMO receiver architecture according to one of the exemplary embodiments of the disclosure.

[0019] FIG. 6 is a flow chart which shows steps of reducing crosstalk of a MIMO transmitter according to one of the exemplary embodiments of the disclosure.

[0020] FIG. 7 is a model block diagram of a MIMO transmitter having in-phase quadrature (IQ) imbalance (IQI) as well as coupling distortion according to one of the exemplary embodiments of the disclosure.

[0021] FIG. 8 shows a MIMO transmitter performing crosstalk pre-compensation according to one of the exemplary embodiments of the disclosure.

[0022] FIG. 9 illustrates using only q.sub.1 (n) and q.sub.2(n) for performing crosstalk pre-compensation according to one of the exemplary embodiments of the disclosure.

[0023] FIG. 10 illustrates a 22 MIMO transmitter architecture according to one of the exemplary embodiments of the disclosure.

[0024] FIG. 11 illustrates a block diagram for performing crosstalk calibration process for a transmitter according to one of the exemplary embodiments of the disclosure.

[0025] FIG. 12 illustrates a schematic diagram of a joint estimation process for performing the crosstalk adjustment of a MIMO transmitter according to one of the exemplary embodiments of the disclosure.

[0026] FIG. 13 is a flow chart which describes the steps of calculating the post-processing parameters for cancelling crosstalk according to one of the exemplary embodiments of the disclosure.

[0027] FIG. 14 is a system block diagram of a MIMO receiver having IQI and coupling distortion.

[0028] FIG. 15 shows a MIMO receiver performing crosstalk post-compensation according to one of the exemplary embodiments of the disclosure.

[0029] FIG. 16 is a conceptual diagram showing the relationship between crosstalk parameters and post-processing parameters according to one of the exemplary embodiments of the disclosure.

[0030] FIG. 17 is a block diagram which shows calculating post-processing parameters of a MIMO receiver according to one of the exemplary embodiments of the disclosure.

[0031] FIG. 18 is a conceptual diagram for testing a MIMO receiver according to one of the exemplary embodiments of the disclosure.

[0032] FIG. 19 is a flow chart which shows a procedure of reducing crosstalk of a MIMO receiver according to one of the exemplary embodiments of the disclosure.

[0033] FIG. 20 is a flow chart which shows steps of performing a crosstalk estimation and compensation procedure for a MIMO transceiver system according to one of the exemplary embodiments of the disclosure.

[0034] FIG. 21 is a system block diagram of a MIMO transceiver system according to one of the exemplary embodiments of the disclosure.

[0035] FIG. 22 shows a system architecture of a MIMO transceiver system which utilizes the disclosed method according to one of the exemplary embodiments of the disclosure.

[0036] FIG. 23 shows a block diagram of a process of reducing crosstalk at the receiving end of a MIMO transceiver system according to one of the exemplary embodiments of the disclosure.

[0037] FIG. 24 is a block of the MIMO transceiver system after processing through the receiving end according to one of the exemplary embodiments of the disclosure.

[0038] FIG. 25 is a flow chart showing steps of crosstalk reducing procedures at the transmitting end and the receiving end after having estimated processing parameters for the transceiver system according to one of the exemplary embodiments of the disclosure.

[0039] FIG. 26 is a block diagram which shows using information from the receiving end to perform crosstalk reducing procedures at the transmitting end and the receiving end according to one of the exemplary embodiments of the disclosure.

DETAILED DESCRIPTION OF DISCLOSED EMBODIMENTS

[0040] Reference will now be made in detail to the present exemplary embodiments of the disclosure, examples of which are illustrated in the accompanying drawings. Wherever possible, the same reference numbers are used in the drawings and the description to refer to the same or like parts.

[0041] As described previously, the current multi-antenna technology has to be able to provide more than 80 MHz of bandwidth which would result in continuous miniaturization and integration of RF components. As a MIMO system transmits and receives multiple RF signals within a small-area of a circuit board or an integrated circuit (IC) chip, crosstalk between RF signals may cause unintended signal mixing, signal distortion, and a reduction of the quality of the signal.

[0042] Based on the above, this disclosure provides a method of reducing crosstalk of a MIMO transceiver system by calibrating the MIMO transceiver of a multi-antenna wireless communication system. The disclosure uses the digital signal processing to estimate parameters of a wideband crosstalk response and compensate for the wideband crosstalk distortion. A pre-compensation procedure could be performed at the transmitter end, and a post-compensation procedure could be provided to the receiver. The disclosure includes various exemplary embodiments for performing the method of reducing crosstalk of a MIMO transceiver system. The exemplary embodiments include performing the above described method according to the crosstalk information at the transmitting end only, at the receiving end only, at both the transmitting end and the receiving end, and other variations of such. Experiments have been performed to verify the effects of the disclosure and experimental results are included toward the end of the disclosure.

[0043] According to the exemplary embodiment of performing the above described method to reduce crosstalk at the transmitting end only, a mathematical model of the transmitting end is provided as well as the procedures for tuning the transmitter to order to estimate the coupling parameters of the transmitting end through a least square (LS) method. During the performance of the LS method and after the matrix has been arranged, pre-compensation parameters of the transmitting end could be obtained. According to the exemplary embodiment of performing the above described method to reduce crosstalk at the receiving end only, a mathematical model of the receiving end is provided. The process of tuning the receiver would first include estimating the crosstalk parameters of the receiving end according to various conditions. After an inverse matrix operation is performed, post-processing parameters could be obtained. The signal at the receiving end could then be post-processed to compensate for the crosstalk and to detect the received value. According to the exemplary embodiment of performing the above described method to reduce crosstalk at both the transmitting end and the receiving end, a mathematical model of the corresponding transceiver architecture is provided. The procedure would include estimating the calibration process and eliminating the respective crosstalk signals in the transceiver. Overall, for each of the exemplary embodiments, the above described method would involve generating or assuming a mathematical model based on relevant components of a transceiver system, estimating the crosstalk factor based on the mathematical model, and performing the compensation based on the estimated crosstalk factor.

[0044] FIG. 1 shows steps of the method of configuring a MIMO wideband receiver and steps of the method of configuring a MIMO wideband transmitter according to one of the exemplary embodiments of the disclosure. The steps performed by a receiver would include not limited to steps S101S106, and steps performed by a transmitter would include not limited to S111S115. Referring to FIG. 1, in step S101, the receiver would estimate, on a SISO basis, a set of post-processing parameters (e.g. P1, P2, P3 P4) for a plurality of receiver channels (e.g. RX1 RX2). In step S102, the receiver would receive, by each of the plurality of receiver channels, a first test signal (e.g. U.sub.1(n)) which is transmitted from a first transmitter channel (e.g. TX1) on a MIMO basis. In step S103, the receiver would calculate a first set of crosstalk parameters (e.g. e.sub.11 e.sub.12 e.sub.21 e.sub.22) in response to receiving the first test signal. In step S104, the receiver would receive, by each of the plurality of receiver channels, a second test signal (e.g. U.sub.2 (n)) which is transmitted from a second transmitter channel (e.g. TX2) on the MIMO basis. In step S105, the receiver would calculate a second set of crosstalk parameters (e.g. f.sub.11 f.sub.12 f.sub.21 f.sub.22) in response to receiving second test signal. In step S106, the receiver would calculate the set of post-processing parameters based on the first set of crosstalk parameters and the second set of crosstalk parameters by cancelling a crosstalk interference among plurality of receiver channels.

[0045] According one of the exemplary embodiments, estimating, on the SISO basis, the set of post-processing parameters for the plurality of receiver channels may involve estimating a second post-processing parameter (e.g P2) and a third post-processing parameter (e.g. P3) only between the first transmitter channel and the first receiver channel, switching from between the first transmitter channel and the first receiver channel (e.g. RX1) to between the second transmitter channel and the second receiver channel (e.g. RX2), and estimating a first post-processing parameter (e.g. P1) and a fourth post-processing parameter (e.g. P4) only between the first transmitter channel and the first receiver channel. The set of post-processing parameters may include the first post-processing parameter (e.g. P1), the second post-processing parameter (e.g. P2), the third post-processing parameter (e.g. P3), and the fourth post-processing parameter (e.g. P4).

[0046] According one of the exemplary embodiments, receiving, by each of the plurality of receiver channels, the first test signal which is transmitted from the first transmitter channel on the MIMO basis may involve receiving, by a first receiver channel of the plurality of receiver channels, a first test signal which is transmitted from a first transmitter channel on a MIMO basis while not receiving from the second transmitter channel and grounding the second receiver channel; receiving, by a second receiver channel of the plurality of receiver channels, the first test signal which is transmitted from a first transmitter channel on the MIMO basis while not receiving from the first transmitter channel and grounding the first receiver channel.

[0047] According one of the exemplary embodiments, calculating the first set of crosstalk parameters in response to receiving the first test signal may involve obtaining a first crosstalk parameter (e.g. e.sub.11) and a second crosstalk parameter (e.g. e.sub.12) based on the first test signal received by the first receiver channel and obtaining a third crosstalk parameter (e.g. e.sub.21) and a fourth crosstalk parameter (e.g. e.sub.22) based on the first test signal received by the second receiver channel. The first set of crosstalk parameters may include the first crosstalk parameter, the second crosstalk parameter, the third crosstalk parameter, and the fourth crosstalk parameter.

[0048] According one of the exemplary embodiments, receiving, by each of the plurality of receiver channels, the second test signal which is transmitted from the second transmitter channel on the MIMO basis may involve receiving, by a first receiver channel of the plurality of receiver channels, a second test signal which is transmitted from a second transmitter channel on a MIMO basis while not receiving from the first transmitter channel, grounding the second receiver channel; receiving, by a second receiver channel of the plurality of receiver channels, the second test signal which is transmitted from a second transmitter channel on the MIMO basis while not receiving from the first transmitter channel, and grounding the first receiver channel. The above described first test signal and the second test signal could be different quadrature phase shift keying (QPSK) training sequences.

[0049] According one of the exemplary embodiments, calculating the second set of crosstalk parameters in response to receiving the second test signal may involve obtaining a fifth crosstalk parameter (e.g. f.sub.11) and a sixth crosstalk (e.g. f.sub.12) parameter based on the second test signal received by the first receiver channel, and obtaining a seventh crosstalk parameter (e.g. f.sub.21) and an eighth crosstalk parameter (e.g. f.sub.22) based on the second test signal received by the second receiver channel. The second set of crosstalk parameters comprising a fifth crosstalk parameter, a sixth crosstalk parameter, a seventh crosstalk parameter, and an eighth crosstalk parameter.

[0050] According one of the exemplary embodiments, calculating the set of post-processing parameters based on the first set of crosstalk parameters may further involve estimating the first crosstalk parameter (e.g. e.sub.11) and the second crosstalk parameter (e.g. e.sub.12) based on a least square technique, and calculating the set of post-processing parameters based on the second set of crosstalk parameters may further involve estimating the fifth crosstalk parameter and the sixth crosstalk parameter based on a least square technique.

[0051] According one of the exemplary embodiments, the method may further include determining whether the set of post-processing parameters cancel out crosstalk among the plurality of receiver channels.

[0052] As for the transmitter, in step S111, the transmitter would transmit on a MIMO basis, through a first transmitter channel of a plurality of transmitting channels, a first test signal to be received by a first receiver channel. In step, the transmitter would transmit on the MIMO basis, through a second transmitter channel of the plurality of transmitting channels, a second test signal to be received by a second receiver channel. In step S113, the transmitter would determine, a first received signal received by the first receiver channel and determine a second received signal received by the second receiver channel. In step S114, the transmitter would estimate, a set of coupling parameters (e.g., c.sub.11, c.sub.12, c.sub.21, c.sub.22) for the plurality of transmitter channels based on the first received signal and the second received signal. In step S115, the transmitter would calculate, based on the set of coupling parameters, a set of pre-processing compensation parameters (e.g. q.sub.1 q.sub.2 q.sub.3 q.sub.4) by cancelling a crosstalk interference among plurality of transmitter channels.

[0053] According to one of the exemplary embodiments, transmitting by the first transmitter channel the first test signal to be received by the first receiver channel and transmitting by the second transmitter channel the second test signal to be received by the second receiver channel may occur simultaneously. The above described first test signal and the second test signal could be different QPSK training sequences. The above described estimating the set of coupling parameters could be performed based on a least square technique. The above described estimating the set of coupling parameters may involve determining the first received signal and the second received signal by setting the set of pre-processing compensation parameters to zero.

[0054] According to one of the exemplary embodiments, the method may further include determining whether the transmitter has cancelled the crosstalk interference among plurality of transmitter channels by applying the pre-processing compensation parameters to a processor of the transmitter. Estimating the set of coupling parameters may further involve assuming the first receiver channel and the second receiver channel as an ideal receiver. The pre-processing compensation parameters could be applied to the processor of the transmitter only once.

[0055] FIG. 2 shows the hardware diagram of a transmitter and a receiver according to one of the exemplary embodiments of the disclosure. It should be noted that the transmitter end which comprising a processor 201, an analog transmitting circuit 202, a first transmitter channel 203 and a second transmitter channel 204 and the receiver end, which comprising a processor 211, an analog receiving circuit 212, a first receiver channel 213 and a second receiver channel 214 could be independently integrated as two separate chips or integrated as a single chip, could be located on the same circuit board or located on two separate circuit boards that are electrically disconnected. The processor 201 of the transmitter could be one or more ICs having processing capabilities and would control the analog transmitting circuit 202 to implement functions of the above describe method of configuring a MIMO wideband transmitter and its embodiments. The processor 201 may implement functions of TX digital as show in the drawings and described in the corresponding written descriptions, and the analog transmitting circuit 202 may implement functions of TX analog as show in the drawings and described in the corresponding written descriptions. The processor 201 may output digital signals to be digitized by a digital (D/A) converter into an analog baseband signal which is then upconverted into RF frequency and transmitted through a MIMO antenna array of the transmitting circuit 202. The analog transmitting circuit 202 and its MIMO antenna array may have multiple channels including a first transmitter channel 203 and a second transmitter channel 204.

[0056] The processor 211 of the receiver could be one or more ICs having processing capabilities and would control the analog receiving circuit 212 to implement functions of the above describe method of configuring a MIMO wideband receiver and its embodiments. The processor 211 may implement functions of RX digital as show in the drawings and described in the corresponding written descriptions, and the analog receiving circuit 212 may implement functions of RX analog as show in the drawings and described in the corresponding written descriptions. The processor 211 may receive digital signals which were digitized by an analog-digital digital (A/D) converter from an analog baseband signal which has been down-converted from RF frequency and received through a MIMO antenna array of the analog receiving circuit 212. The analog receiving circuit 212 and its MIMO antenna array may have multiple channels including a first receiver channel 213 and a second receiver channel 214.

[0057] FIG. 3 shows a simplified conceptual diagram of an overall MIMO wideband transceiver system architecture according to one of the exemplary embodiments of the disclosure. In transmitter block 301, the transmitter would obtain a digital baseband transmitting signal by using a processor (e.g. 201) for estimating the crosstalk factor based on a mathematical model. Next, the transmitter block 301 would perform a pre-processing procedure on the digital baseband transmitting signal and subsequently perform a digital to analog (D/A) conversion on the pre-processed digital baseband transmitting signal to generate a pre-processed analog baseband transmitting signal which contains a crosstalk factor. The transmitter block 302 would up-convert the pre-processed analog baseband transmitting signal into a pre-processed analog RF transmitting signal to be transmitted by using a MIMO antenna array. The pre-processed analog RF transmitting signal is to be received by the MIMO receiver antenna array of the receiver block 303 as an analog RF receiving signal which is assumed to contain the crosstalk factor. The analog RF receiving signal would then be down-converted into an analog baseband receiving signal.

[0058] The receiver block 304 may perform an analog-to-digital (A/D) conversion on the analog baseband receiving signal to generate a digital baseband receiving signal. Subsequently, the receiver block 304 would perform a post-processing procedure by using a processor (e.g. 211) on the digital baseband receiving signal to estimate the original digital baseband transmitting signal based on the crosstalk factor.

[0059] MIMO wideband transceiver system could be demarcated into a transmitting end (i.e. MIMO transmitter (e.g. 201 202 203 204)) and a receiving end (i.e. MIMO transmitter (e.g. 211 212 213 214)). To further describe the method of configuring the wideband MIMO transmitter and the structure of the wideband MIMO transmitter, the disclosure provides several exemplary embodiments as shown in FIG. 4FIG. 12. FIG. 4 is an architecture of a transmitting end according to one of the exemplary embodiments of the disclosure. For the architecture of FIG. 4 which shows the transmitting end, the pre-processing procedure would include a pre-compensation procedure. For the ease of elucidation, a 22 MIMO transmitter and a 22 MIMO receiver is assumed. In FIG. 4, a first transmitter channel is assumed to transmit a first transmitting signal (U.sub.1(n)), and a second transmitter channel is assumed to transmit a second transmitting signal (U.sub.2(n)). U.sub.1(n) would experience an interference signal based on a signal from U.sub.2(n) and vice versa. The interference signal would mix with U.sub.1(n) to cause the first output r.sub.1(n) to be distorted. Similarly, the second output r.sub.2(n) would also be distorted due to the interference from U.sub.1(n). However, by using the algorithms to be provided in latter parts of the disclosure, the crosstalk factor c.sub.11(n), c.sub.21(n), c.sub.12(n), and c.sub.22(n) at the transmitting end could be estimated so as to subsequently derive the pre-compensation matrix q.sub.1, q.sub.2, q.sub.3, q.sub.4 accordingly. Next, the pre-compensation matrix could be placed as a part of the transmitter block (e.g. 201) so as to pre-compensate for the crosstalk to be received by the receiving end in order to maintain overall performance of the transceiver system.

[0060] FIG. 5 extends upon the concepts of FIG. 4 and includes a wideband MIMO receiver (i.e. receiving end). As shown in FIG. 5, the crosstalk problem also exists in the receiving end since V.sub.p,1(n) receives not only intended signal from a first channel but also unintended signal, destined toward V.sub.p,2(n), from a second channel. Therefore, a pre-processing procedure would be performed to cancel out the crosstalk shared among receiving channels. In particular, the receiver processing parameters P.sub.1, P.sub.2, P.sub.3, P.sub.4 would be configured to resolve the receiver crosstalk factor d.sub.11, d.sub.12, d.sub.13, d.sub.14. As described previously, the crosstalk may occur as the result of signal mixing between U1(n) and U2(n) at the transmitting end to cause distortion at r.sub.1(n) and r.sub.2(n). However, the crosstalk parameters could be obtained by an algorithm to be described in further detail as the post-processing correlation matrix is derived, and then the post-processing parameters P.sub.1, P.sub.2, P.sub.3, P.sub.4 could be obtained for performing the post-processing procedure by the receiving end. Consequently, the crosstalk factor could be suppressed accordingly.

[0061] To describe the estimation and pre-compensation for the crosstalk at the transmitting end of a wideband communication system, the disclosure provides further details as shown in FIG. 6FIG. 9 and their corresponding descriptions. FIG. 6 is a flow chart which shows steps of reducing crosstalk of a MIMO transmitter according to one of the exemplary embodiments of the disclosure. In step S601, the transmitter would estimate MIMO transmitter coupling parameters for at least two transmitting channels and at least two receiving channels. In step S602, the transmitter would estimate transmitter pre-processing parameters (e.g. q.sub.1, q.sub.2, q.sub.3, q.sub.4). In step S603, the transmitter would transmit a MIMO single carrier test signal or a MIMO multi-carrier test signal. In step S604, the transmitter would compensation for the transmitter pre-processing (or interference) parameters (e.g. c.sub.11, c.sub.12, c.sub.21, c.sub.22).

[0062] To further explain the above steps, FIG. 7 shows a model block diagram of a MIMO transmitter having in-phase quadrature (IQ) imbalance (IQI) as well as coupling distortion according to one of the exemplary embodiments of the disclosure. The crosstalk factor at the transmitting end refers to the scenario where a cross-frequency interference signal is generated among multiple channels of a wideband RF circuit after baseband signals have been upconverted into RF frequency signals. There could be multiple crosstalk factor signals generated on the chip as the crosstalk phenomenon may occur among multiple RF transmitters. This crosstalk factor may affect any one of the multiple channels causing distortions and affecting the performance of the transceiver.

[0063] When a signal is transmitted through a wideband transmitter having multiple inputs, the signal is bound to be accompanied by the IQ Imbalance (IQI) of the broadband radio frequency, and then the crosstalk response (coupling/crosstalk) is generated through the crosstalk scene of the transmitter as shown in FIG. 7 which could be used as a model to represent a main signal and a coupled signal due to the cross talk phenomenon. The received signal r.sub.1(n) could be represented by equation 1.

[00001]$\begin{array}{cc}{r}_l\left(n\right)=u_l\left(n\right)+\underset{{}_{m=1,ml}}{\overset{M}{.Math.}}.Math.c_{m.Math.l}\left(n\right).Math.u_m\left(n\right)+v_l\left(n\right)& \mathrm{equation}.Math..Math.1\end{array}$

[0064] In equation 1, .Math. stands for convolution. u.sub.m (n): stands the I/Q modulation signal (with broadband IQ imbalance factor) for the m.sup.th antenna. c.sub.ml(n): stands for the filtered response value (L_cm length) of the m.sup.th antenna to the crosstalk of the l.sup.th antenna transmitter, where c.sub.ml(n)=[c.sub.ml(n), c.sub.ml(n1), . . . , c.sub.ml(nL.sub.cm+1)].sup.T. v.sub.l(n): indicates the noise of the l.sup.th antenna.

[0065] FIG. 8 shows a MIMO transmitter crosstalk pre-compensation architecture according to one of the exemplary embodiments of the disclosure. In this disclosure, it is assumed that the broadband RF imperfection factor has been calibrated, and for the multiple inputs transmitter, the crosstalk factor response and its corresponding crosstalk adjustment technique is provided for a 22 MIMO broadband system. The same technique could be extended to a NN MIMO broadband system where N is greater than 2 by using the same or a similar principle.

[0066] Referring to FIG. 8, at the digital transmitting end (Tx digital), U.sub.1(n) and U.sub.2(n) are the original transmission signal without crosstalk, and U.sub.1(n) and U.sub.2(n) are input into a crosstalk pre-compensation filter represented by q.sub.1(n), q.sub.2(n), q.sub.3(n), q.sub.4(n) for pre-processing, and the pre-processed signals u.sub.p,1(n) and u.sub.p,2(n) are obtained. When the MIMO RF transmitter crosstalk occurs at the analog end (Tx analog), the first RF transmission signal r.sub.1(n) and the second RF transmission signal r.sub.2(n) would both be affected. As long as the pre-compensation parameters q.sub.1(n), q.sub.2(n), q.sub.3(n), q.sub.4(n) can be accurately estimated, it would help r.sub.1(n) and r.sub.2(n) to avoid crosstalk and to maintain the original signal integrity.

[0067] In order to estimate the crosstalk factor of the transmitter in a wideband MIMO system, the Least Square (LS) technique could be used to estimate the broadband crosstalk factor at the transmitting end. Such technique may enhance the interference effect on the unknown signal and avoid high computational complexity. Next, and then estimate the transmitter pre-compensation vector of the transmitter could be estimated based on the algorithms to be provided in order to solve the crosstalk factor among different channels of the MIMO transmitter so as to achieve high-quality communication requirements of the broadband MIMO system. The technique is provided as follows.

[0068] First, there is no pre-compensation action before estimating the crosstalk factors c.sub.11(n), c.sub.21(n), c.sub.12(n), c.sub.22(n), and thus q.sub.1(n)=q.sub.2(n)=q.sub.3(n)=q.sub.4(n)=0. Therefore, for the 1=1 and m=2 scenarios, m=2 is the crosstalk signal of the second transmitter channel (TX2), so the signal to be received by the first receiver channel (RX1), r.sub.1(n), could be expressed by equation 2.

r.sub.1(n)=c.sub.11(n).Math.u.sub.1(n)+c.sub.21(n)u.sub.2(n)equation 2

[0069] for the 1=2 and m=1 scenarios, m=1 is the crosstalk signal of the first transmitter channel (TX1), so the signal to be received by the second receiver channel (RX2), r.sub.2(n), could be expressed as equation 3.

r.sub.2(n)=c.sub.22(n).Math.u.sub.2(n)+c.sub.12(n).Math.u.sub.1(n)equation 3

[0070] Equation 2 could be expressed in the matrix form which is shown as equation 4.

[00002]$\begin{array}{cc}{r}_1=U_1.Math.c_{1.Math.1}+U_2.Math.c_{2.Math.1}.Math.r_1=\left[U_1.Math..Math.U_2\right]\left[\begin{array}{c}{c}_{11}\\ c_{21}\end{array}\right]& \mathrm{equation}.Math..Math.4\end{array}$

[0071] Equation 3 could be expressed in the matrix form which is shown as equation 5.

[00003]$\begin{array}{cc}{r}_2=U_1.Math.c_{12}+U_2.Math.c_{22}.Math.r_2=\left[U_1.Math..Math.U_2\right]\left[\begin{array}{c}{c}_{12}\\ c_{22}\end{array}\right]& \mathrm{equation}.Math..Math.5\end{array}$

[0072] In equation 4 and 5, r.sub.1 and r.sub.2 are the vector representations of r.sub.1(n) and r.sub.2(n), U.sub.1 and U.sub.2 are convolution matrix representations of u.sub.1(n) and u.sub.2(n), and u=[u1 u2].

[0073] However, when estimating the crosstalk factor at the transmitting end, two sets of QPSK modulation signals could be used as the known training codes for u.sub.1(n) and u.sub.2(n), so equation 4 could be used with the least squares technique so as to allow the signal transmitted from TX1 be known based on the training code to in order to obtain the received signal from RX1 by using equation 6.

[00004]$\begin{array}{cc}\left[\begin{array}{c}{\hat{c}}_{1.Math.1}\\ {\hat{c}}_{2.Math.1}\end{array}\right]=U^{+}.Math.r_1& \mathrm{equation}.Math..Math.6\end{array}$

[0074] Similarly, equation 5 could be used with the least squares technique so as to allow the signal transmitted from TX2 be known based on the training code in order to obtain the received signal from RX2 by using equation 7.

[00005]$\begin{array}{cc}\left[\begin{array}{c}{\hat{c}}_{1.Math.2}\\ {\hat{c}}_{2.Math.2}\end{array}\right]=U^{+}.Math.r_2& \mathrm{equation}.Math..Math.7\end{array}$

[0075] In equation 7, U.sup.+=(U.sup.HU).sup.1U.sup.H.

[0076] Based on equation 6 and equation 7 as shown above, the unknown parameters c.sub.11, c.sub.21, c.sub.22, c.sub.12 could be solved, and then base on the algorithm to be presented, the pre-compensation parameters q.sub.1, q.sub.2, q.sub.3, q.sub.4 of the transmitting end could be derived.

[0077] FIG. 9 shows using only q.sub.1(n) and q.sub.2(n) for performing the pre-compensation procedure according to one of the exemplary embodiments of the disclosure. Assuming that the signal of TX1 is u.sub.p,1(n), then u.sub.p,1(n) could be represented as equation 8.

TX.sub.p,1: u.sub.p,1(n)=u.sub.1(n)+q.sub.2(n).Math.u.sub.2(n)equation 8

[0078] Assuming that the signal of TX2 is u.sub.p,2(n), then u.sub.p,2(n) could be represented as equation 9.

TX.sub.p,2: u.sub.p,2(n)=u.sub.2(n)+q.sub.1(n).Math.u.sub.1(n)equation 9

[0079] If u.sub.p,1(n) from equation 8 is replaced by u.sub.1(n) of equation 2, then it can represent the to be received signal r.sub.1(n) after the TX1 signal is pre-compensated only by the crosstalk factor q.sub.1(n), q.sub.2(n) which are used to compensate for the received signal r.sub.1(n) as shown in equation 10.

[00006]$\begin{array}{cc}\begin{array}{c}{r}_1\left(n\right)=.Math.c_{1.Math.1}\left(n\right).Math.u_{p,1}\left(n\right)+c_{2.Math.1}\left(n\right).Math.u_{p,2}\left(n\right)\\ =.Math.c_{1.Math.1}\left(n\right).Math.\left\{u_1\left(n\right)+q_2\left(n\right).Math.u_2\left(n\right)\right\}+\\ .Math.c_{2.Math.1}\left(n\right).Math.\left\{u_2\left(n\right)+q_1\left(n\right).Math.u_1\left(n\right)\right\}\end{array}& \mathrm{equation}.Math..Math.10\end{array}$

[0080] The equation 10 could be further expanded to express r.sub.1(n) as equation 11.

[00007]$\begin{array}{cc}{r}_1\left(n\right)=\left\{c_{1.Math.1}\left(n\right)+c_{2.Math.1}\left(n\right).Math.q_1\left(n\right)\right\}.Math.u_1\left(n\right)+\left\{c_{1.Math.1}\left(n\right).Math.q_2\left(n\right)+c_{2.Math.1}\left(n\right)\right\}.Math.u_2\left(n\right).Math.0& \mathrm{equation}.Math..Math.11\end{array}$

[0081] If u.sub.p,2(n) from equation 8 is replaced by u.sub.2(n) of equation 3, then it can represent the to be received signal r.sub.2(n) after the TX2 signal is compensated by the pre-compensation parameter which are used for eliminating the crosstalk factor as shown in equation 12.

[00008]$\begin{array}{cc}\begin{array}{c}{r}_2\left(n\right)=.Math.c_{2.Math.2}\left(n\right).Math.u_{p,2}\left(n\right)+c_{1.Math.2}\left(n\right).Math.u_{p,1}\left(n\right)\\ =.Math.c_{2.Math.2}\left(n\right).Math.\{u_2\left(n\right)+q_1\left(n\right).Math.u_1\left(n\right).Math.\}+\\ .Math.c_{1.Math.2}\left(n\right).Math.\left\{u_1\left(n\right)+q_2\left(n\right).Math.u_2\left(n\right)\right\}\end{array}& \mathrm{equation}.Math..Math.12\end{array}$

[0082] The equation 12 could be further expanded to express r.sub.1(n) as equation 13.

[00009]$\begin{array}{cc}{r}_2\left(n\right)=\left\{c_{2.Math.2}\left(n\right).Math.q_1\left(n\right)+c_{1.Math.2}\left(n\right)\right\}.Math.u_1\left(n\right).Math..Math.0+.Math.\left\{c_{2.Math.2}\left(n\right)+c_{1.Math.2}\left(n\right).Math.q_2\left(n\right)\right\}.Math.u_2\left(n\right)& \mathrm{equation}.Math..Math.13\end{array}$

[0083] Further, in equation 11, in order to eliminate the crosstalk signal in u.sub.2(n) from TX2 so as to make the crosstalk signal in RX1 be zero as the zero crosstalk of r.sub.1(n)=r.sub.2(n) is satisfied, the equation could be re-organized as equation 14.

c.sub.11(n).Math.q.sub.2(n)+c.sub.21(n)=0.Math.c.sub.21+c.sub.11q.sub.2=0equation 14

[0084] In equation 13, in order to eliminate the crosstalk signal in u1(n) from TX1 so as to make the crosstalk signal in RX2 be zero as the zero crosstalk of r.sub.2(n)=r.sub.2(n) is satisfied, the equation could be re-organized as equation 15.

c.sub.12(n)+c.sub.22(n).Math.q.sub.1(n)=0.Math.c.sub.12+c.sub.22q.sub.1=0equation 15

[0085] In equation 14 and 15, c.sub.11 and c.sub.22 are the convolution matrix of c.sub.11(n) c.sub.22(n), c.sub.21 c.sub.12 are the crosstalk response vector of c.sub.21(n) c.sub.12(n), and q.sub.1 q.sub.2 are the only crosstalk canceling factor of q.sub.1(n) q.sub.2(n) pre-compensation vector.

[0086] However, in order to obtain the pre-compensation vector of the pre-compensation parameters of the transmitting end, the crosstalk response parameter of the transmitting end of the matrix C could be estimated by the least square technique as previously described, and thus the matrix C could be derived. After performing an inverse matrix operation on equation 14 and an inverse matrix operation on equation 15, q.sub.1 and q.sub.2 could be derived as equation 16 and equation 17.

q.sub.2=(C.sub.11.sup.HC.sub.11).sup.1C.sub.11.sup.Hc.sub.21equation 16

q.sub.1=(C.sub.22.sup.HC.sub.22).sup.1C.sub.22.sup.Hc.sub.12equation 17

[0087] In equations 16, q.sub.2 is a pre-compensation parameter for cancelling m=2 crosstalk signal within 1=1, and q.sub.1 is a pre-compensation parameter for cancelling m=1 crosstalk signal within 1=2.

[0088] However, since the above-described suppression of the crosstalk factor is only performed by using the pre-compensation vector q.sub.1(n) q.sub.2(n) for eliminating the crosstalk factor, the original main signal strength has been weakened so that additional pre-compensation processing is required for maintaining the main signal strength in order for the pre-compensation vector for the crosstalk of the transmitter be fully estimated. Therefore, based on the architecture of FIG. 9, the crosstalk problem at the transmitting end should be solved and at the same time the main signal strength could be maintained. Before the transmitting end would experience the crosstalk, the compensation parameter is added in advance in order to eliminate the upcoming crosstalk response of the transmitting end. By maintaining the original signal strength, the pre-compensated transmitting signal of the TX1 original signal can be expressed by equation 18.

TX.sub.p,1: u.sub.p,1(n)=q.sub.3(n).Math.u.sub.1(n)+q.sub.2(n).Math.u.sub.2(n)equation 18

[0089] The pre-compensated transmitting signal of the TX2 original signal could be expressed as equation 19.

TX.sub.p,2: u.sub.p,2(n)=q.sub.4(n).Math.u.sub.2(n)+q.sub.1(n).Math.u.sub.1(n)equation 19

[0090] By replacing u.sub.1(n) of equation 18 with u.sub.p,1(n), it represents the to be received signal r.sub.1(n) after the signal in Tx1 has been compensated by the pre-compensation parameter as shown in m equation 20.

[00010]$\begin{array}{cc}\begin{array}{c}{r}_1\left(n\right)=.Math.c_{1.Math.1}\left(n\right).Math.u_{p,1}\left(n\right)+c_{2.Math.1}\left(n\right).Math.u_{p,2}\left(n\right)\\ =.Math.c_{1.Math.1}\left(n\right).Math.\left\{q_3\left(n\right).Math.u_1\left(n\right)+q_2\left(n\right).Math.u_2\left(n\right)\right\}+\\ .Math.c_{2.Math.1}\left(n\right).Math.\left\{q_4\left(n\right).Math.u_2\left(n\right)+q_1\left(n\right).Math.u_1\left(n\right)\right\}\end{array}& \mathrm{equation}.Math..Math.20\end{array}$

[0091] Equation 20 could be expanded to derived equation 21.

[00011]$\begin{array}{cc}{r}_1\left(n\right)=\left\{c_{1.Math.1}\left(n\right).Math.q_3\left(n\right)+c_{2.Math.1}\left(n\right).Math.q_1\left(n\right)\right\}.Math.u_1\left(n\right).Math..Math.\left(n\right)+.Math.\left\{c_{1.Math.1}\left(n\right).Math.q_2\left(n\right)+c_{2.Math.1}\left(n\right).Math.q_4\left(n\right)\right\}.Math.u_2\left(n\right).Math..Math.0& \mathrm{equation}.Math..Math.21\end{array}$

[0092] By replacing u.sub.2(n) of equation 3 with u.sub.p,2(n) of equation 19, it represents i the to be received signal r.sub.2(n) after the signal in Tx2 has been compensated by the pre-compensation parameter as shown in equation 22.

[00012]$\begin{array}{cc}\begin{array}{c}{r}_2\left(n\right)=.Math.c_{2.Math.2}\left(n\right).Math.u_{p.Math.2}\left(n\right)+c_{1.Math.2}\left(n\right).Math.u_{p.Math.1}\left(n\right)\\ =.Math.c_{22}\left(n\right).Math.\left\{q_4\left(n\right).Math.u_2\left(n\right)+q_1\left(n\right).Math.u_1\left(n\right)\right\}+\\ .Math.c_{1.Math.2}\left(n\right).Math.\left\{q_3\left(n\right).Math.u_1\left(n\right)+q_2\left(n\right).Math.u_2\left(n\right)\right\}\end{array}& \mathrm{equation}.Math..Math.22\end{array}$

[0093] Equation 22 could be expanded to derive equation 23.

[00013]$\begin{array}{cc}{r}_2\left(n\right)=\left\{c_{2.Math.2}\left(n\right).Math.q_1\left(n\right)+c_{1.Math.2}\left(n\right).Math.q_3\left(n\right)\right\}.Math.u_1\left(n\right).Math..Math.0+.Math.\left\{c_{2.Math.2}.Math.\left(n\right).Math.q_4\left(n\right)+c_{1.Math.2}\left(n\right).Math.q_2\left(n\right)\right\}.Math.u_2\left(n\right).Math..Math.\left(n\right)& \mathrm{equation}.Math..Math.23\end{array}$

[0094] For equation 21, in order for RX1 to receive the signal only from TX1 and set it to 1, and eliminate the crosstalk signal from TX2 in RX1 and make it 0 thus satisfying the zero crosstalk purpose of r2(n)u2(n), the above equation can be re-organized as equation 24.

[00014]$\begin{array}{cc}\{\begin{array}{c}{c}_{11}\left(n\right).Math.q_3\left(n\right)+c_{21}\left(n\right).Math.q_1\left(n\right)=\left(n\right)\\ c_{11}\left(n\right).Math.q_2\left(n\right)+c_{21}\left(n\right).Math.q_4\left(n\right)=0\end{array}.Math.\{\begin{array}{c}{C}_{11}.Math.{\underset{\_}{q}}_3+C_{21}.Math.{\underset{\_}{q}}_1=\underset{\_}{e}\\ C_{11}.Math.{\underset{\_}{q}}_2+C_{21}.Math.{\underset{\_}{q}}_4=0\end{array}& \mathrm{equation}.Math..Math.24\end{array}$

[0095] For equation 23, in order for RX2 to receive the signal only from TX2 and set it to 1, and eliminate the crosstalk signal from TX1 in RX2 and make it 0 thus satisfying the zero crosstalk purpose of r1(n)u1(n), the above equation can be re-organized as equation 25.

[00015]$\begin{array}{cc}\{\begin{array}{c}{c}_{12}\left(n\right).Math.q_3\left(n\right)+c_{22}\left(n\right).Math.q_1\left(n\right)=0\\ c_{12}\left(n\right).Math.q_2\left(n\right)+c_{22}\left(n\right).Math.q_4\left(n\right)=\left(n\right)\end{array}.Math.\{\begin{array}{c}{C}_{12}.Math.{\underset{\_}{q}}_3+C_{22}.Math.{\underset{\_}{q}}_1=0\\ C_{12}.Math.{\underset{\_}{q}}_2+C_{22}.Math.{\underset{\_}{q}}_4=\underset{\_}{e}\end{array}& \mathrm{equation}.Math..Math.25\end{array}$

[0096] In equations 24 and 25, c.sub.11 c.sub.21 c.sub.12 c.sub.22 are the convolution matrix of c.sub.11(n) c.sub.21(n) c.sub.12(n) c.sub.22(n), q.sub.1 q.sub.2 q.sub.3 q.sub.4 is the response vector of q.sub.1(n) q.sub.2(n) q.sub.3(n) q.sub.4(n), and e=[1 0.sup.T].sup.T is a vector with the first element being 1 and the other elements being 0. After re-arranging equations 24 and 25, equations 26 and 27 could be respectively derived.

[00016]$\begin{array}{cc}\left[\begin{array}{cc}{C}_{11}& C_{2.Math.1}\\ C_{1.Math.2}& C_{2.Math.2}\end{array}\right]\left[\begin{array}{c}{\underset{\_}{q}}_3\\ {\underset{\_}{q}}_1\end{array}\right]=\left[\begin{array}{c}\underset{\_}{e}\\ \underset{\_}{0}\end{array}\right]& \mathrm{equation}.Math..Math.26\\ \left[\begin{array}{cc}{C}_{11}& C_{2.Math.1}\\ C_{1.Math.2}& C_{2.Math.2}\end{array}\right]\left[\begin{array}{c}{\underset{\_}{q}}_2\\ {\underset{\_}{q}}_1\end{array}\right]=\left[\begin{array}{c}\underset{\_}{0}\\ \underset{\_}{e}\end{array}\right].Math..Math.\mathrm{where}.Math..Math.C=\left[\begin{array}{cc}{C}_{11}& C_{2.Math.1}\\ C_{1.Math.2}& C_{2.Math.2}\end{array}\right]& \mathrm{equation}.Math..Math.27\end{array}$

[0097] In order to obtain the response vector of the pre-compensation parameters of the transmitting end, the above described LS technique could be used to estimate the crosstalk response parameters of the matrix C. Since matrix C is already a known parameter, after performing an inverse matrix operation of equation 26 and an inverse matrix operation of equation 27, equations 28 and 29 could be respectively derived.

[00017]$\begin{array}{cc}\left[\begin{array}{c}{\underset{\_}{q}}_3\\ {\underset{\_}{q}}_1\end{array}\right]={\left(C^H.Math.C\right)}^{-1}.Math.C^H\left[\begin{array}{c}\underset{\_}{e}\\ \underset{\_}{0}\end{array}\right]& \mathrm{equation}.Math..Math.28\\ \left[\begin{array}{c}{\underset{\_}{q}}_2\\ {\underset{\_}{q}}_4\end{array}\right]={\left(C^H.Math.C\right)}^{-1}.Math.C^H\left[\begin{array}{c}\underset{\_}{0}\\ \underset{\_}{e}\end{array}\right]& \mathrm{equation}.Math..Math.29\end{array}$

[0098] Accordingly, the transmitter pre-compensation vector of the transmitting end could be obtained through equations 28 and 29 so as to complete the pre-compensation procedure for eliminating the crosstalk response in each channels of the transmitter.

[0099] Based on the disclosure above, a crosstalk estimation system is proposed for transmitter-side crosstalk calibration. The system block diagram could be represented as FIG. 10 illustrates a 22 MIMO transmitter architecture according to one of the exemplary embodiments of the disclosure. The system includes a TX digital block which performs the above described crosstalk pre-processing, a TX analog block containing crosstalk parameters, and an RX analog block which is a receiver assumed to be in an ideal state.

[0100] The system of FIG. 10 is further expanded upon as shown in FIG. 11. First, the (LS) method is used by combining the transmitting end and the receiving end to simultaneously transmit and receive the estimation by using the same frequency. Next, a known QPSK training code could be used as the reference signal U.sub.1(n), U.sub.2(n). The above described inverse matrix and the subsequent convolution could be performed with the receiving signal R.sub.1(n)R.sub.2(n) to re-arrange the matrix so as to estimate crosstalk response c.sub.11 c.sub.21 c.sub.12 c.sub.22 of the transmitting. Also, as previously described, the estimated crosstalk response c.sub.11 c.sub.21 c.sub.12 c.sub.22 of the transmitter could be arranged into C through a matrix and then converted by an inverse matrix operation to estimate the pre-compensation parameter q.sub.1 q.sub.2 q.sub.3 q.sub.4. The purpose of such is to make RX1 only receive the signal from TX1, but not the coupling interference signal from TX2. At the same time, RX2 would receive the signal from TX2 without including the coupling interference signal from TX1, which satisfies equations 24 and 25.

[0101] After the cross-talk response c.sub.11 c.sub.21 c.sub.12 c.sub.22 and the pre-compensation vector q.sub.1 q.sub.2 q.sub.3 q.sub.4 are estimated, single carrier or multi-carrier signal to be transmitted could be added to the pre-compensation vector so that RX1 only receives the signal from TX1, while RX2 only receives the signal from TX2. The system is capable of obtaining the crosstalk response and the pre-compensation vector through only one estimation which may occur when the power is turned on, and then the estimated parameters could be used continuously to complete the pre-compensation transmission and reception for the signal to be tested. The overall process has been described in FIG. 6.

[0102] FIG. 12 illustrates a schematic diagram of a joint estimation process for performing the crosstalk adjustment of a MIMO transmitter according to one of the exemplary embodiments of the disclosure. Steps S1201, S1202, and S1203 are performed based on a joint estimation method of TX1, TX2 and RX1, RX2 while assuming the above described 22 MIMO system. In step S1201, both TX1 and TX2 would each transmit different known QPSK training codes. In step S1202, c.sub.11, c.sub.21, c.sub.12, and c.sub.22 are estimated. In step S1203, RX1 would receive the QPSK training code from TX1 and RX2 would receive the QPSK training code from TX2. In step S1204, the crosstalk factor of the transmitting end would be estimated. In step S1205, the compensation parameters q.sub.1, q.sub.2, q.sub.3, q.sub.4 would be estimated.

[0103] Next, in order for the disclosure to further describe the method of configuring the wideband MIMO receiver and the structure of the wideband MIMO receiver, the disclosure provides several exemplary embodiments as shown in FIG. 13FIG. 19. FIG. 13 is a flow chart which describes the steps of calculating the post-processing parameters for cancelling crosstalk according to one of the exemplary embodiments of the disclosure. In step S1301, a SISO based measurement would be performed to estimate the post-processing parameters P.sub.1 P.sub.2 P.sub.3 P.sub.4 by switching among channels of the transmitter such as by switching between TX1 and TX2 as well as by switching between RX1 and RX2. For example, TX1 and RX1 could be connected while other paths are isolated from the connection between TX1 and RX1. Next, TX1 and RX2 could be connected while other paths are isolated. Next, TX2 and RX1 could be connected while other paths are isolated. Next, TX2 and RX2 could be connected while other paths are isolated, and so forth. In step S1302, post-processing parameters P.sub.1 P.sub.2 P.sub.3 P.sub.4 would be obtained based on the measurement of step S1301. In step S1303, a MIMO based measurement would be performed in response to transmitting MIMO single carrier or multi-carrier test signal to obtain sets of crosstalk parameters. In step S1304, based on the estimated the post-processing parameters P.sub.1 P.sub.2 P.sub.3 P.sub.4 and sets of crosstalk parameters, the post-processing parameters P.sub.1 P.sub.2 P.sub.3 P.sub.4 could be derived.

[0104] FIG. 14 is a system block diagram of a MIMO receiver having IQI and coupling distortion. The disclosure will provide a mechanism to estimate and compensate for the wideband crosstalk factor of the MIMO receivers and the post-processing parameters of the MIMO receiver in this section. The wideband crosstalk factor at the receiving end refers to a scenario in which a crosstalk factor is stored when the multi-channel radio frequency circuit is fabricated at the receiving end before the frequency-transmitted signal is received by the radio frequency. Such phenomenon could be more pronounced on a PC board on a RF chip. The channel receiving signals may interfere with each other, causing crosstalk between multiple receiving ends as well signal distortions to affect the performance of multiple receiving signals. After the wideband crosstalk of the receiving end is resolved, and the crosstalk factor of the receiving end could be estimated and subsequently compensated with a post-processing procedure.

[0105] However, when the signal is transmitted through the multi-input and wideband system having crosstalk, a signal could be received at the receiving end and be corrupted because of cross channel coupling or crosstalk effect before the signal receives RF down-conversion, and then the down-converted received signal could be carried along with the receiver's broadband RF imperfect factor (IQ Imbalance, IQI) of the receiver. Such phenomenon is shown in the block diagram of FIG. 14.

[0106] However, the above describe problem could be resolved. FIG. 15 shows a MIMO receiver having crosstalk post-compensation according to one of the exemplary embodiments of the disclosure. The disclosure would provide an equivalent model which indicates the main signal as 1 and the coupled signal as m, and the signal transmitted by the first channel to the receiving end having crosstalk is shown in equation 101.

[00018]$\begin{array}{cc}{t}_l\left(n\right)=y_l\left(n\right)+\underset{m=1,ml}{\overset{M}{.Math.}}.Math.d_{m.Math.l}\left(n\right).Math.y_m\left(n\right)& \mathrm{equation}.Math..Math.101\end{array}$

[0107] In addition, the signal t.sub.1 (n) which is distorted by crosstalk of the receiving end is down-converted and thus received by the wide-band IQI factor of the receiving end. The received signal z.sub.1 (n) could be obtained as shown in equation 101.

z.sub.1(n)=f.sub.1l(n).Math.t.sub.l(n)+f.sub.2l(n).Math.t*.sub.l(n)+w.sub.l(n)equation 101

[0108] Wherein, in the equation 101, d.sub.ml(n) represents the filter response value of the m.sup.th antenna to the crosstalk of the l.sup.th antenna receiving end, and in the equation 102, w.sub.l(n) represents the noise of the l.sup.th antenna. However, the disclosure may assume that the broadband RF imperfection factor has been adjusted, and then the multi-input wideband system receiver broadband crosstalk factor response and its post-processing crosstalk adjustment method would be performed as provided. For the simplicity of disclosure, a 22 MIMO system is to be assumed.

[0109] In the transmitting end, U.sub.1(n) U.sub.2(n) are assumed to be the original transmission signal without crosstalk. As such signal enters the TX analog section, the crosstalk of the transmitting end could be obtained from the multipath of r.sub.1(n) and r.sub.2(n). When entering the RX analog section of the receiver, crosstalk at the receiving end would occurs. V.sub.p,1, (n) and V.sub.P,2(n) respectively would represent the receive signals having crosstalk, and Z1(n) and Z2(n) would represent the signals output from the RX digital section and having been compensated by the post-process compensation parameter P.sub.1(n) P.sub.2(n) P.sub.3(n) P.sub.4(n). If the post-processing compensation parameter P.sub.1(n) P.sub.2(n) P.sub.3(n) P.sub.4(n) could be be accurately estimated, the Z1(n) and Z2(n) would be able to output signals having to have no crosstalk out of the receiving end.

[0110] Thus, a mathematical modelling method for estimating the crosstalk response at the receiving end of this 22 MIMO wideband receiving end system is to be provided. The receiving end crosstalk response e.sub.11 e.sub.12 f.sub.21 f.sub.22 could be derived from the mathematical model of the receiving end post processing parameter P.sub.1 P.sub.2 P.sub.3 P.sub.4.

[0111] Since the transmitting end and the receiving end both contain a crosstalk factor on the transceiver of the MIMO transceiver system, in order to estimate the coupling amount of the receiving end and subsequently eliminate the crosstalk, it could be helpful to isolate and simplify the remaining signals through several conditions. First, the signal is to be transmitted twice, first from TX1 and second from TX2 signal. A switch is utilized before the receiving end to performing switching between a connection state and a grounding state of the transmitted signal so as to interface with RX1 and RX2 of the receiver. The permutation of the 22 MIMO transceiver is shown in Table 1 below.

TABLE-US-00001 TABLE 1 2 2 MIMO with 4 TX1 TX2 sets of conditions for (1 represents on and 0 (1 represents on and 0 estimating crosstalk represents off) represents off) RX1 TX1 = 1 {grave over ()} TX2 = 0 TX1 = 0 {grave over ()} TX2 = 1 (1 represents on and 0 RX1 = 1 {grave over ()} RX2 = 0 RX1 = 1 {grave over ()} RX2 = 0 represents off) First set of conditions Third set of conditions RX2 TX1 = 1 {grave over ()} TX2 = 0 TX1 = 0 {grave over ()} TX2 = 1 (1 represents on and 0 RX1 = 0 {grave over ()} RX2 = 1 RX1 = 0 {grave over ()} RX2 = 1 represents off) Second set of conditions Forth set of conditions

[0112] In order to estimate the crosstalk factor at the receiving end of the broadband MIMO system, the QPSK signal is to be used as the training code. The LS method could be used to estimate the broadband crosstalk factor at the receiving end. The disclosure would also provide a procedure to estimate the post-processing vector at the receiving end, to solve the crosstalk factor at the receiving end of the MIMO transceiver system, and to achieve the high-quality communication requirements of the wide-band MIMO system in the following section.

[0113] First, when estimating the crosstalk factor d.sub.11(n), d.sub.21(n), d.sub.12(n), d.sub.22(n) at the receiving end, there is no pre-compensation and post-processing for the crosstalk factor between the transmitting end and the receiving end before and after the transmitting end, and thus q.sub.1(n)=q.sub.2(n)=q.sub.3(n)=q.sub.4(n) and p.sub.1(n)=p.sub.2(n)=p.sub.3(n)=p.sub.4(n)=0. Therefore, in the first set of conditions, only the TX1 transmit signal with crosstalk through the transmitting end, and only RX1 receives the received signal before being interfered by the crosstalk of the receive end (TX1=QPSK, TX2=0, RX1=1, RX2=0). Thus, in the scenario where TX1 receives the main signal and TX2 receives the crosstalk, the received signal in RX1 after transmission of TX1 could be expressed as by equation 103.

z.sub.1(n)=u.sub.1(n).Math.c.sub.11(n).Math.d.sub.11(n)equation 103

[0114] Based on equation 103, the convolution of crosstalk c.sub.11(n) and d.sub.11(n) received at the receiving end could be represented as a new crosstalk variable a shown in equation 104.

z.sub.1(n)=u.sub.1(n).Math.e.sub.11(n).Math.z.sub.1=U.sub.1e.sub.11equation 104

[0115] However, for the first set of conditions, in the scenario where TX2 transmits the main signal and TX1 transmits the crosstalk signal end, the receiving signal at RX2 after the TX2 transmission could be expressed as equation 105.

z.sub.2(n)=u.sub.1(n).Math.c.sub.11(n).Math.d.sub.12(n)equation 105

[0116] According to equation 105, the crosstalk c.sub.11(n) and d.sub.12(n) received at the receiving end can be convolved and renamed to a new crosstalk variable ei2(n), as shown in the equation 106.

z.sub.2(n)=u.sub.1(n).Math.e.sub.12(n).Math.z.sub.2=U.sub.1e.sub.12equation 106

[0117] Next, by inverting the matrix of equation 104 and equation 106, the new crosstalk parameters e.sub.11 and d.sub.12 could be obtained from the first set of conditions, as expressed by the following equation (4.7).

[00019]$\begin{array}{cc}\{\begin{array}{c}{\underset{\_}{e}}_{11}=U_1^{+}.Math.{\underset{\_}{z}}_1\\ {\underset{\_}{e}}_{12}=U_1^{+}.Math.{\underset{\_}{z}}_2\end{array},\mathrm{obtain}.Math..Math.{\underset{\_}{e}}_{11}.Math.`.Math..Math.{\underset{\_}{e}}_{12}& \mathrm{equation}.Math..Math.107\end{array}$

[0118] Next, in the second set of conditions, only the TX1 would transmit signal with crosstalk through the transmit end, and only RX2 would receive crosstalk signal before the receiving end (TX1=QPSK, TX2=0, RX1=0, RX2=1). At this time, in the scenario where TX1 transmits the main signal and TX2 transmits the crosstalk signal end, the RX1 would receive signal after the signal transmission from TX1 transmission which could be expressed as equation 108.

z.sub.1(n)=u.sub.1(n).Math.c.sub.12(n).Math.d.sub.21(n)equation 108

[0119] Among them, according to the equation 108, the crosstalk c.sub.12 and d.sub.12 received at the receiving end can be convolved and renamed as a new crosstalk variable, as shown in equation 109.

z.sub.1(n)=u.sub.1(n).Math.e.sub.21(n).Math.z.sub.1=U.sub.1e.sub.21equation 109

[0120] However, for the second group of conditions, in the scenario where TX2 transmits the main signal and TX1 transmits the crosstalk signal end, the receiving signal transmitted by TX2 and received by RX2 could be expressed as equation 110.

z.sub.2(n)=u.sub.1(n).Math.c.sub.12(n).Math.d.sub.22(n)equation 110

[0121] According to equation 110, the crosstalk c.sub.12(n) and d.sub.22(n) received at the receiving end can be convolved and renamed to a new crosstalk variable, as shown in the following equation 111.

z.sub.2(n)=u.sub.1(n).Math.e.sub.22(n).Math.z.sub.2=U.sub.1e.sub.22equation 111

[0122] Subsequently, the equations 109 and 111 could be inverted, and the new crosstalk parameter could be obtained from the second set of conditions, as expressed by the following equation 112.

[00020]$\begin{array}{cc}\{\begin{array}{c}{\underset{\_}{e}}_{21}=U_1^{+}.Math.{\underset{\_}{z}}_1\\ {\underset{\_}{e}}_{22}=U_1^{+}.Math.{\underset{\_}{z}}_2\end{array},\mathrm{obtain}.Math..Math.{\underset{\_}{e}}_{21}.Math.`.Math..Math.{\underset{\_}{e}}_{22}& \mathrm{equation}.Math..Math.112\end{array}$

[0123] Then, in the third set of conditions, only the TX2 transmit signal with crosstalk through the transmit end and only RX1 would receive the signal before the receive end with crosstalk (TX1=0, TX2=QPSK, RX1=1, RX2=0). In the scenario where the main signal is transmitted from TX1 and the crosstalk is transmitted from TX2, the received signal z.sub.1(n) from RX1 after being transmitted by the TX1 can be expressed as equation 113.

z.sub.1(n)=u.sub.2(n).Math.c.sub.21(n).Math.d.sub.11(n)equation 113

[0124] According to equation 113, The convolution of the crosstalk c.sub.21(n) and d.sub.11(n) that can be received at the receiving end and can be renamed to a new crosstalk variable according to equation 114.

z.sub.1(n)=u.sub.2(n).Math.f.sub.11(n).Math.z.sub.1=U.sub.2f.sub.11equation 114

[0125] However, for the third group of conditions, in the scenario where TX2 transmits the main signal and TX1 transmits the crosstalk signal, the receiving signal of RX2 transmitted by TX2 can be expressed as equation 115.

z.sub.2(n)=u.sub.2(n).Math.c.sub.21(n).Math.d.sub.12(n)equation 115

[0126] Then, according to the above equation 115, the crosstalk c.sub.21(n) and d.sub.12(n) received at the receiving end can be convolved and renamed as a new crosstalk variable, as shown in the following equation 116.

z.sub.2(n)=u.sub.2(n).Math.f.sub.12(n).Math.z.sub.2=U.sub.2f.sub.12equation 116

[0127] Subsequently, the equations 114 and 116 could be inverted, and the new crosstalk parameters could be obtained from the third set of conditions, as expressed by the following equation 117.

[00021]$\begin{array}{cc}\{\begin{array}{c}{\underset{\_}{f}}_{21}=U_2^{+}.Math.{\underset{\_}{z}}_1\\ {\underset{\_}{f}}_{22}=U_2^{+}.Math.{\underset{\_}{z}}_2\end{array},\mathrm{obtain}.Math..Math.{\underset{\_}{f}}_{11}.Math.`.Math..Math.{\underset{\_}{f}}_{12}& \mathrm{equation}.Math..Math.117\end{array}$

[0128] Finally, in the fourth set of conditions, only the TX2 transmit signal with crosstalk through the transmit end and only RX2 would receive signal before the crosstalk of the receive end (TX1=0, TX2=QPSK, RX1=0, RX2=1). In the scenario where TX1 transmits the main signal and the TX2 transmits the crosstalk signal, the signal received by RX1 and transmitted by the TX1 could be expressed as equation 118.

z.sub.1(n)=u.sub.2(n).Math.c.sub.22(n).Math.d.sub.21(n)equation 118

[0129] Then, according to the above equation 118, the convolution of the crosstalk c.sub.22(n) and d.sub.21(n) received at the receiving end can be renamed to a new crosstalk variable f.sub.21 as equation 119.

z.sub.1(n)=u.sub.2(n).Math.f.sub.21(n).Math.z.sub.1=U.sub.2f.sub.21equation 119

[0130] However, for the fourth set of conditions, in the scenario where TX2 transmits the main signal and TX1 transmits the crosstalk signal end, the signal z.sub.2(n) received by RX2 receiving signal after being transmitted by TX2 can be expressed as equation 120.

z.sub.2(n)=u.sub.2(n).Math.c.sub.22(n).Math.d.sub.22(n)equation 120

[0131] Then, according to the above equation 120, the crosstalk c.sub.22(n) and d.sub.22(n) received at the receiving end can be convolved and renamed to a new crosstalk variable f.sub.22(n), as shown in the following equation 121.

z.sub.2(n)=u.sub.2(n).Math.f.sub.22(n).Math.z.sub.2=U.sub.2f.sub.22equation 121

[0132] By performing a reverse matrix operation of equation 119 and 121, based on the fourth set of conditions, a new crosstalk variable f.sub.21 and f.sub.22 could be obtained as shown in equation 122.

[00022]$\begin{array}{cc}\{\begin{array}{c}{\underset{\_}{f}}_{21}=U_2^{+}.Math.{\underset{\_}{z}}_1\\ {\underset{\_}{f}}_{22}=U_2^{+}.Math.{\underset{\_}{z}}_2\end{array},\mathrm{obtain}.Math..Math.{\underset{\_}{f}}_{21}.Math.\text{'}.Math..Math.{\underset{\_}{f}}_{22}& \mathrm{equation}.Math..Math.122\end{array}$

[0133] Among them, in the above four sets of conditions in equations 107, 112, 117 and 122, both z.sub.1 and z.sub.2 are vector representations of z.sub.1 and z.sub.2, and U.sub.1 and U.sub.2 are convolution matrices of u.sub.1(n) and u.sub.2(n).

[0134] When estimating the crosstalk response at the receiving end, the QPSK modulation signal could be used as the known training code of the transmitting end u.sub.1(n) or u.sub.2(n). By using a switch, the crosstalk or signal entering the receiving end could be controlled and thus forming a new crosstalk response at the receiving end and its post-processing compensation architecture. FIG. 16 is a conceptual diagram showing the relationship between crosstalk parameters and post-processing parameters according to one of the exemplary embodiments of the disclosure. However, according to the second set of conditions and the third set of conditions, after signal from TX1/TX2 having crosstalk is received by RX2/RX1, the signal will be coupled to both the transmitting end and the receiving end. Therefore, when the first set of conditions and the fourth set of conditions are satisfied, according to the above equations 107 and 122, the crosstalk response parameter {right arrow over (e)}.sub.11 e.sub.12 f.sub.21 f.sub.22 could be estimated by the LS method. The post-processing compensation parameter p.sub.1 p.sub.2 p.sub.3 p.sub.4 of the receiving end by using algorithms provided in the next section of the disclosure. In order to solve the problem of multipath crosstalk at the receiving end, the post-processing vectors p.sub.1(n) p.sub.2(n) p.sub.3(n) p.sub.4(n) could be used as shown in FIG. 15 to complete the crosstalk response suppression interference at the receiving end. However, estimating the compensation vectors p.sub.1(n) p.sub.2(n) p.sub.3(n) p.sub.4(n) at the receiving end, since both the transmitting end and the receiving end contain a crosstalk factor on the transceiver of the MIMO system, there is no pre-compensation for the transmitting end before the transmitting end. Thus q.sub.1(n)=q.sub.2(n)=q.sub.3(n)=q.sub.4(n)=0. Therefore, the TX1 RF signal transmitted after the crosstalk response of the transmitting end can be expressed as equation 123.

TX.sub.1:r.sub.1(n)=c.sub.11(n).Math.u.sub.1(n)+c.sub.21(n).Math.u.sub.2(n)equation 123

[0135] The TX2 RF signal r2(n) after the transmitter crosstalk response is transmitted can be expressed as equation 124.

TX.sub.2: r.sub.2(n)=c.sub.12(n).Math.u.sub.1(n)+c.sub.22(n).Math.u.sub.2(n)equation 124

[0136] The received signal v.sub.p,1(n) after r.sub.1(n) receives crosstalk, the response of the receiving end is expressed as equation 125.

v.sub.p,1(n)=r.sub.1(n).Math.d.sub.11(n)+r.sub.2(n).Math.d.sub.21(n)equation 125

[0137] The received signal v.sub.p,2(n) after r.sub.2(n) receives crosstalk, the response of the receiving end is expressed as equation 126.

v.sub.p,2(n)=r.sub.1(n).Math.d.sub.12(n)+r.sub.2(n).Math.d.sub.22(n)equation 126

[0138] As seen from the above figure that when the analog signal receives crosstalk by the receiving end and enters the digital end, and the analog signal is processed by the receiving end to obtain the receiving signal in RX1. The equation can be expressed as equation 127.

z.sub.1(n)=p.sub.3(n).Math.v.sub.p,1(n)+p.sub.2(n).Math.v.sub.p,2(n)equation 127

[0139] At the same time, when the analog signal vp,2(n) after receiving crosstalk of the receiving end enters the digital domain and performs the post-processing compensation of the receiving end to obtain the receiving signal z.sub.2(n) through RX2, the equation can be expressed as equation 128.

z.sub.2(n)=p.sub.1(n).Math.v.sub.p,1(n)+p.sub.4(n).Math.v.sub.p,2(n)equation 128

[0140] However, according to the above description, in order to eliminate the crosstalk at the receiving end, it could be helpful to isolate and simplify the remaining signals, thereby forming the above four sets of conditions. In the first set of conditions, only the TX1 transmit signal with crosstalk through the transmitting end, and only RX1 would receive signal before receiving crosstalk at the receiving end (TX1=QPSK, TX2=0, RX1=1, RX2=0). At this time, since the RF signal and the RF signal transmitted by the crosstalk response of the TX1 and TX2 transmitters respectively have only the signal from U.sub.1(n) at the TX1, the part of the signal can be obtained from the equation 123 and 124 and expressed as equations 129 and 130 below.

TX.sub.1:r.sub.1(n)=c.sub.11(n).Math.u.sub.1(n)equation 129

TX.sub.2: r.sub.2(n)=c.sub.12(n).Math.u.sub.1(n)equation 130

[0141] Then, the crosstalk response is input to the receiving end, and the equations 129 and 130 are substituted into the equation 125 to obtain the signal v.sub.p,1(n). Next, the convolution of c.sub.11(n) and d.sub.11(n) is renamed to the new crosstalk variable e.sub.11(n), and the convolution is performed between c.sub.11(n) and d.sub.11(n). The new crosstalk variable e.sub.21(n) is as shown in equation 131.

[00023]$\begin{array}{cc}\begin{array}{cc}{v}_{p,1}\left(n\right).Math.=& .Math.\left\{u_1\left(n\right).Math.c_{11}\left(n\right)\right\}.Math.d_{11}\left(n\right)+\\ & .Math.\left\{u_1\left(n\right).Math.c_{12}\left(n\right)\right\}.Math.d_{21}\left(n\right)\\ =& .Math.u_1\left(n\right).Math.e_{11}\left(n\right)+u_1\left(n\right).Math.e_{21}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.131\end{array}$

[0142] However, in the first set of conditions, only the signal r.sub.1(n) is input through the switch before receiving the crosstalk at the receiving end, so that the v.sub.p,1(n) signal of the RX1 only contains the r.sub.1(n) RF signal, such as equation 132.

[00024]$\begin{array}{cc}\begin{array}{c}{v}_{p,1}\left(n\right)=\left\{u_1\left(n\right).Math.c_{11}\left(n\right)\right\}.Math.d_{11}\left(n\right)\\ =u_1\left(n\right).Math.e_{11}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.132\end{array}$

[0143] At the same time, after entering the crosstalk response of the receiving end, the equations 129 and 130 are substituted into the equation 126 to obtain the v.sub.p,2(n) signal. The convolution of c.sub.11 (n) and d.sub.12 (n) is renamed to the new crosstalk variable e.sub.12(n), and c.sub.12 is obtained. The convolution with d.sub.22(n) is renamed to the new crosstalk variable e.sub.22(n) as in equation 133.

[00025]$\begin{array}{cc}\begin{array}{cc}{v}_{p,2}\left(n\right).Math.=& .Math.\left\{u_1\left(n\right).Math.c_{11}\left(n\right)\right\}.Math.d_{12}\left(n\right)+\\ & .Math.\left\{u_1\left(n\right).Math.c_{12}\left(n\right)\right\}.Math.d_{22}\left(n\right)\\ =& .Math.u_1\left(n\right).Math.e_{12}\left(n\right)+u_1\left(n\right).Math.e_{22}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.133\end{array}$

[0144] According to the first set of conditions, only the signal r.sub.1(n) is input through the switch before the crosstalk at the receiving end, and the v.sub.p,2(n) signal of the RX2 only contains the crosstalk RF signal of r.sub.1(n), as expressed in equation 134.

[00026]$\begin{array}{cc}\begin{array}{c}{v}_{p,2}\left(n\right)=\left\{u_1\left(n\right).Math.c_{11}\left(n\right)\right\}.Math.d_{12}\left(n\right)\\ =u_1\left(n\right).Math.e_{12}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.134\end{array}$

[0145] Subsequently, after entering the digital end processing, it is assumed that the post-processing parameter p.sub.1(n) p.sub.2(n) p.sub.3(n) p.sub.4(n) can counteract the signal of the crosstalk response of RX1 and the signal v.sub.p,1(n) of the crosstalk response of RX2, so the equations 132 and 134 are substituted into the equation 127. RX1 receives the signal z.sub.1(n) as shown in equation 135.

[00027]$\begin{array}{cc}{z}_1\left(n\right)=p_3\left(n\right).Math.\left\{u_1\left(n\right).Math.e_{11}\left(n\right)\right\}+p_2\left(n\right).Math.\left\{u_1\left(n\right).Math.e_{12}\left(n\right)\right\}& \mathrm{equation}.Math..Math.135\end{array}$

[0146] The equation 135 could be rearranged to the equation that the TX1 transmits the signal u.sub.1(n) in the RX1 reception signal z.sub.1(n), and then the processing vector suppresses the received crosstalk response, as shown in the following equation 136.

[00028]$\begin{array}{cc}{z}_1\left(n\right)=u_1\left(n\right).Math.\left\{p_3\left(n\right).Math.e_{11}\left(n\right)\right\}+u_1\left(n\right).Math.\left\{p_2\left(n\right).Math.e_{12}\left(n\right)\right\}& \mathrm{equation}.Math..Math.136\end{array}$

[0147] After expanding the equations 132, 134 and substitute them into equation 128, Z.sub.2(n) could be obtained at RX2 as expressed in equation 137.

[00029]$\begin{array}{cc}{z}_2\left(n\right)=p_1\left(n\right).Math.\left\{u_1\left(n\right).Math.e_{11}\left(n\right)\right\}+p_4\left(n\right).Math.\left\{u_1\left(n\right).Math.e_{12}\left(n\right)\right\}& \mathrm{equation}.Math..Math.137\end{array}$

[0148] After rearranging equation 137 as transmitting signal u.sub.1(n) for TX1 in RX2 receive signal z.sub.2(n), the subsequent processing vector suppresses the equation for receiving the crosstalk response as shown in equation 138.

[00030]$\begin{array}{cc}{z}_2\left(n\right)=u_1\left(n\right).Math.\left\{p_1\left(n\right).Math.e_{11}\left(n\right)\right\}+u_1\left(n\right).Math.\left\{p_4\left(n\right).Math.e_{12}\left(n\right)\right\}& \mathrm{equation}.Math..Math.138\end{array}$

[0149] In the first set of conditions, the TX1 transmission signal U.sub.1(n) is the main signal. According to the above equations 136 and 138, it can be known that the equation 136 RX1 receiving signal z.sub.1(n) maintains the original signal reception (equation And=1), at the same time, the RX2 receiving signal z.sub.2(n) in the equation 138 formula is suppressed (the equation and =0). Therefore, the effective set equation of the first set of conditions can be unified, as shown in the following equation 139.

[00031]$\begin{array}{cc}\{\begin{array}{c}{p}_3\left(n\right).Math.e_{11}\left(n\right)+p_2\left(n\right).Math.e_{12}\left(n\right)=\left(n\right)\\ p_1\left(n\right).Math.e_{11}\left(n\right)+p_4\left(n\right).Math.e_{12}\left(n\right)=0.Math.\end{array}.& \mathrm{equation}.Math..Math.139\end{array}$

[0150] Equation 139 can be expressed as a matrix form as equation 140.

[00032]$\begin{array}{cc}\{\begin{array}{c}{\underset{\_}{E}}_{11}.Math.{\underset{\_}{p}}_3+{\underset{\_}{E}}_{12}.Math.{\underset{\_}{p}}_2=\underset{\_}{e}\\ {\underset{\_}{E}}_{11}.Math.{\underset{\_}{p}}_1+{\underset{\_}{E}}_{12}.Math.{\underset{\_}{p}}_4=\underset{\_}{0}\end{array}& \mathrm{equation}.Math..Math.140\end{array}$

[0151] In the second set of conditions, only the TX1 transmit signal with crosstalk through the transmit end, and only RX2 receives signal before receiving crosstalk of the receive end (TX1=QPSK, TX2=0, RX1=0, RX2=1). It can be found that since the second set of conditions is consistent with the conditions of the first set of conditions, only the signal u.sub.1(n) from TX1 exists, so the RF signal r.sub.1(n) and r.sub.2(n) transmitted after the analog end crosstalk response is transmitted through TX1 and TX2 respectively. And the RF signal can be sequentially expressed as shown in equations 129 and 130.

[0152] Then, after entering the analog crosstalk receiving end, the equation 129 and the equation type are substituted into the equation type to obtain the v.sub.p,2(n) signal of the second set of conditions, which only contains the u.sub.1(n) signal of the TX1, so according to the equation 131 above, new crosstalk parameters e.sub.11(n) and e.sub.21(n) could be obtained as equation 141.

[00033]$\begin{array}{cc}\begin{array}{cc}{v}_{p,1}\left(n\right).Math.=& .Math.\left\{u_1\left(n\right).Math.c_{11}\left(n\right)\right\}.Math.d_{11}\left(n\right)+\\ & .Math.\left\{u_1\left(n\right).Math.c_{12}\left(n\right)\right\}.Math.d_{21}\left(n\right)\\ =& .Math.u_1\left(n\right).Math.e_{11}\left(n\right)+u_1\left(n\right).Math.e_{21}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.141\end{array}$

[0153] However, in the second set of conditions, only the signal r.sub.2(n) is input through the switch before receiving the crosstalk at the receiving end, so that the v.sub.p,1(n) signal of the RX1 only contains the r.sub.2(n) RF signal, such as equation 142.

[00034]$\begin{array}{cc}\begin{array}{c}{v}_{p,1}\left(n\right)=\left\{u_1\left(n\right).Math.c_{11}\left(n\right)\right\}.Math.d_{21}\left(n\right)\\ =u_1\left(n\right).Math.e_{21}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.142\end{array}$

[0154] At the same time, after entering the crosstalk response of the receiving end, the equations 129 and 130 formulas are substituted into the equation 126 to obtain the signal V.sub.p,2(n), and since it only contains the signal U.sub.1(n) of TX1, according to the above equation 133, new crosstalk parameters e.sub.12(n) and e.sub.22(n) could be obtained and, as shown in equation 143.

[00035]$\begin{array}{cc}\begin{array}{cc}{v}_{p,2}\left(n\right).Math.=& .Math.\left\{u_1\left(n\right).Math.c_{11}\left(n\right)\right\}.Math.d_{12}\left(n\right)+\\ & .Math.\left\{u_1\left(n\right).Math.c_{12}\left(n\right)\right\}.Math.d_{22}\left(n\right)\\ =& .Math.u_1\left(n\right).Math.e_{12}\left(n\right)+u_1\left(n\right).Math.e_{22}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.143\end{array}$

[0155] According to the second set of conditions, only the input signal r.sub.2(n) is transmitted through the switch before the crosstalk is introduced at the receiving end, and the RX2 would only contain signal V.sub.p,2(n) which contains the crosstalk RF signal of R.sub.2(n), such as shown in equation 144.

[00036]$\begin{array}{cc}\begin{array}{c}{v}_{p,2}\left(n\right)=\left\{u_1\left(n\right).Math.c_{11}\left(n\right)\right\}.Math.d_{22}\left(n\right)\\ =u_1\left(n\right).Math.e_{22}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.144\end{array}$

[0156] Subsequently, after entering the digital terminal, it is assumed that the post-processing parameters P.sub.1(n) p.sub.2(n) P.sub.3(n) P.sub.4(n) can counter the signal V.sub.p,1(n) of the crosstalk response of RX1 and V.sub.P,2(n) of the crosstalk response of RX2, so the equations of 142 and 144 could be substituted into equation 127, and thus the receiving signal Z.sub.1(n) at RX1 could be obtained and expressed as equation 145.

[00037]$\begin{array}{cc}{z}_1\left(n\right)=p_3\left(n\right).Math.\left\{u_1\left(n\right).Math.e_{21}\left(n\right)\right\}+p_2\left(n\right).Math.\left\{u_1\left(n\right).Math.e_{22}\left(n\right)\right\}& \mathrm{equation}.Math..Math.145\end{array}$

[0157] The equation (4.45) can be rearranged into the equation for z.sub.1(n) of RX1 corresponding to U.sub.1(n) of the TX1 transmit signal, and then the processing vector suppresses the received crosstalk response, as shown in the following equation 146.

[00038]$\begin{array}{cc}{z}_1\left(n\right)=u_1\left(n\right).Math.\left\{p_3\left(n\right).Math.e_{21}\left(n\right)\right\}+u_1\left(n\right).Math.\left\{p_2\left(n\right).Math.e_{22}\left(n\right)\right\}& \mathrm{equation}.Math..Math.146\end{array}$

[0158] Substituting equations 142 and 144 into 148 would derive Z.sub.2(n) at RX2 such as equation 147.

[00039]$\begin{array}{cc}{z}_2\left(n\right)=p_1\left(n\right).Math.\left\{u_1\left(n\right).Math.e_{21}\left(n\right)\right\}+p_4\left(n\right).Math.\left\{u_1\left(n\right).Math.e_{22}\left(n\right)\right\}& \mathrm{equation}.Math..Math.147\end{array}$

[0159] The equation 147 could be rearranged into Z.sub.2(n) of RX2 corresponding to U.sub.1(n) in TX1, and then the processing vector suppresses the received crosstalk response, as shown in the following equation 148.

[00040]$\begin{array}{cc}{z}_2\left(n\right)=u_1\left(n\right).Math.\left\{p_1\left(n\right).Math.e_{21}\left(n\right)\right\}+u_1\left(n\right).Math.\left\{p_4\left(n\right).Math.e_{22}\left(n\right)\right\}& \mathrm{equation}.Math..Math.148\end{array}$

[0160] Finally, in the second set of conditions, the u.sub.1(n) of the TX1 transmit signal is the main signal. According to the above equations 136 and 148, it can be known that the RX1 receive signal z.sub.1(n) of the equation 146 would need to maintain the original signal reception (equal and =1). At the same time, the RX2 receiving signal z.sub.2(n) in equation 148 must be suppressed (the equation and =0). Therefore, the effective set equation of the first set of conditions can be unified, as shown in the following equation 149.

[00041]$\begin{array}{cc}\{\begin{array}{c}{p}_3\left(n\right).Math.e_{21}\left(n\right)+p_2\left(n\right).Math.e_{22}\left(n\right)=\left(n\right)\\ p_1\left(n\right).Math.e_{21}\left(n\right)+p_4\left(n\right).Math.e_{22}\left(n\right)=0.Math.\end{array}& \mathrm{equation}.Math..Math.149\end{array}$

[0161] Then, the equation 149 could be expressed as a matrix form as equation 150.

[00042]$\begin{array}{cc}\{\begin{array}{c}{\underset{\_}{E}}_{21}.Math.{\underset{\_}{p}}_3+{\underset{\_}{E}}_{22}.Math.{\underset{\_}{p}}_2=\underset{\_}{e}\\ {\underset{\_}{E}}_{21}.Math.{\underset{\_}{p}}_1+{\underset{\_}{E}}_{22}.Math.{\underset{\_}{p}}_4=\underset{\_}{0}\end{array}& \mathrm{equation}.Math..Math.150\end{array}$

[0162] In the third set of conditions, only the TX2 transmit signal with crosstalk through the transmitting end and only RX1 would receive signal before receiving the crosstalk of the receive end (TX1=0, TX2=QPSK, RX1=1, RX2=0). After the crosstalk response of the TX1 and TX2 transmitters, the RF signal r.sub.1(n) and the RF signal r.sub.2(n) are transmitted only to have the signal U.sub.2(n) from TX2. Therefore, the part of the signal U.sub.2(n) could be obtained from the equations 123 and 124, which could be expressed as equations 151 and 152.

TX.sub.1: r.sub.1(n)=c.sub.21(n).Math.u.sub.2(n)equation 151

TX.sub.2: r.sub.2(n)=c.sub.22(n).Math.u.sub.2(n)equation 152

[0163] Then, when entering the crosstalk response at the receiving end, and after equations 151 and 152 are substituted into equation 125, the signal v.sub.p,1(n) could obtained; then the convolution between c.sub.21(n) and d.sub.11(n) is renamed as the new crosstalk variable, and the convolution is performed. The signal v.sub.p,1(n) could as expressed as equation 153.

[00043]$\begin{array}{cc}\begin{array}{cc}{v}_{p,1}\left(n\right).Math.=& .Math.\left\{u_2\left(n\right).Math.c_{21}\left(n\right)\right\}.Math.d_{11}\left(n\right)+\\ & .Math.\left\{u_2\left(n\right).Math.c_{22}\left(n\right)\right\}.Math.d_{21}\left(n\right)\\ =& .Math.u_2\left(n\right).Math.f_{11}\left(n\right)+u_2\left(n\right).Math.f_{21}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.153\end{array}$

[0164] However, in the third set of conditions, only the signal r.sub.1(n) is input through the switch before the crosstalk at the receiving end, so that the signal V.sub.p,1(n) of the RX1 only contains the RF signal r.sub.1(n), as expressed in equation 154.

[00044]$\begin{array}{cc}\begin{array}{c}{v}_{p,1}\left(n\right)=\left\{u_2\left(n\right).Math.c_{21}\left(n\right)\right\}.Math.d_{11}\left(n\right)\\ =u_2\left(n\right).Math.f_{11}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.154\end{array}$

[0165] At the same time, after entering the crosstalk response of the receiving end, the equations 151 is substituted into equation 126 to obtain the signal V.sub.p,2(n). The convolution between c.sub.21(n) and d.sub.21(n) is renamed to the new crosstalk variable f.sub.12(n), and convolution between c.sub.22(n) and d.sub.22(n) is renamed to the new crosstalk variable f.sub.22(n). The convolution is renamed to a new crosstalk variable as equation 155.

[00045]$\begin{array}{cc}\begin{array}{cc}{v}_{p,2}\left(n\right).Math.=& .Math.\left\{u_2\left(n\right).Math.c_{21}\left(n\right)\right\}.Math.d_{12}\left(n\right)+\\ & .Math.\left\{u_2\left(n\right).Math.c_{22}\left(n\right)\right\}.Math.d_{22}\left(n\right)\\ =& .Math.u_2\left(n\right).Math.f_{12}\left(n\right)+u_2\left(n\right).Math.f_{22}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.155\end{array}$

[0166] According to the third set of conditions, only the signal is r.sub.1(n) input through the switch before the crosstalk at the receiving end, and the V.sub.p,2(n) of the RX2 signal only contains the crosstalk RF signal r.sub.1(n), such as equation 156.

[00046]$\begin{array}{cc}\begin{array}{c}{v}_{p,2}\left(n\right)=\left\{u_2\left(n\right).Math.c_{21}\left(n\right)\right\}.Math.d_{12}\left(n\right)\\ =u_2\left(n\right).Math.f_{12}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.156\end{array}$

[0167] Subsequently, after entering the digital terminal, it is assumed that the post-processing parameters P.sub.1(n) P.sub.2(n) P.sub.3(n) P.sub.4(n) can counter the signal V.sub.p,1(n) of the crosstalk response of RX1 and the V.sub.p,2(n) of crosstalk response of RX2, so the equations 154 and 156 are substituted into 127 to obtain received signal z.sub.1(n) at RX1 as shown in equation 157.

[00047]$\begin{array}{cc}{z}_1\left(n\right)=p_3\left(n\right).Math.\left\{u_2\left(n\right).Math.f_{11}\left(n\right)\right\}+p_2\left(n\right).Math.\left\{u_2\left(n\right).Math.f_{12}\left(n\right)\right\}& \mathrm{equation}.Math..Math.157\end{array}$

[0168] It is also possible to rearrange equation 157 for z.sub.1(n) of the RX1 corresponding to U.sub.2(n) of the transmit signal at TX2, and then the processing vector suppresses the receive crosstalk response, as shown in equation 158.

[00048]$\begin{array}{cc}{z}_1\left(n\right)=u_2\left(n\right).Math.\left\{p_3\left(n\right).Math.f_{11}\left(n\right)\right\}+u_2\left(n\right).Math.\left\{p_2\left(n\right).Math.f_{12}\left(n\right)\right\}& \mathrm{equation}.Math..Math.158\end{array}$

[0169] Substituting equations 154 and 156 into equation 128, Z.sub.2(n) at RX2 could be obtained and expressed as equation 159.

[00049]$\begin{array}{cc}{z}_2\left(n\right)=p_1\left(n\right).Math.\left\{u_2\left(n\right).Math.f_{11}\left(n\right)\right\}+p_4\left(n\right).Math.\left\{u_2\left(n\right).Math.f_{12}\left(n\right)\right\}& \mathrm{equation}.Math..Math.159\end{array}$

[0170] The equation 159 can be rearranged to Z.sub.2(n) of RX2 corresponding to u.sub.2(n) of the TX2 transmit signal, and then the processing vector suppresses the received crosstalk response equation, as shown in the equation 160.

[00050]$\begin{array}{cc}{z}_2\left(n\right)=u_2\left(n\right).Math.\left\{p_1\left(n\right).Math.f_{11}\left(n\right)\right\}+u_2\left(n\right).Math.\left\{p_4\left(n\right).Math.f_{12}\left(n\right)\right\}& \mathrm{equation}.Math..Math.160\end{array}$

[0171] Finally, in the third set of conditions, the TX2 transmits u.sub.2(n) signal which is the main signal. According to the above equations 158) and 160, it can be seen that the RX1 receive signal z.sub.1(n) of the equation 158 is suppressed and eliminated (equation and =0). At the same time, the RX2 receiving signal z.sub.2(n) in (4.60) must be maintained to receive the original signal (equal and =1). Therefore, the effective set equation of the first set of conditions can be unified as shown in equation 161.

[00051]$\begin{array}{cc}\{\begin{array}{c}{p}_3\left(n\right).Math.f_{11}\left(n\right)+p_2\left(n\right).Math.f_{12}\left(n\right)=0.Math.\\ p_1\left(n\right).Math.f_{11}\left(n\right)+p_4\left(n\right).Math.f_{12}\left(n\right)=\left(n\right)\end{array}& \mathrm{equation}.Math..Math.161\end{array}$

[0172] Then, the equation 161 could expressed in a matrix form as equation 162.

[00052]$\begin{array}{cc}\{\begin{array}{c}{\underset{\_}{F}}_{11}.Math.{\underset{\_}{p}}_3+{\underset{\_}{F}}_{12}.Math.{\underset{\_}{p}}_2=\underset{\_}{0}\\ {\underset{\_}{F}}_{11}.Math.{\underset{\_}{p}}_1+{\underset{\_}{F}}_{12}.Math.{\underset{\_}{p}}_4=\underset{\_}{e}\end{array}& \mathrm{equation}.Math..Math.162\end{array}$

[0173] In the fourth set of conditions, only the TX2 transmit signal with crosstalk through the transmit end and only RX2 receives signal before the crosstalk of the receive end (TX1=0, TX2=QPSK, RX1=0, RX2=1). Since the third set of conditions is consistent with the conditions of the fourth set of conditions, only the signal u.sub.2(n) from TX2 exists, so the RF signal r.sub.1(n) and the RF signal r.sub.2(n) transmitted by the analog end crosstalk response of TX1 and TX2 respectively can be expressed as equations 151 and 152.

[0174] Then, after entering the analog crosstalk receiving end, the 151 and 152 equations are subdivided into equation 125 and thus based on the fourth set of conditions the v.sub.p,1(n) signal could be obtained but only contain the signal u.sub.2(n) of TX2, so according to the above equation 153, the new crosstalk parameters f.sub.11(n) and f.sub.21(n) could be obtained and as shown in equation 163.

[00053]$\begin{array}{cc}\begin{array}{cc}{v}_{p,1}\left(n\right).Math.=& .Math.\left\{u_2\left(n\right).Math.c_{21}\left(n\right)\right\}.Math.d_{11}\left(n\right)+\\ & .Math.\left\{u_2\left(n\right).Math.c_{22}\left(n\right)\right\}.Math.d_{21}\left(n\right)\\ =& .Math.u_2\left(n\right).Math.f_{11}\left(n\right)+u_2\left(n\right).Math.f_{21}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.163\end{array}$

[0175] However, in the fourth set of conditions, only the signal r.sub.2(n) is input through the switch before the crosstalk at the receiving end, so that the signal r.sub.2(n) of the V.sub.p,1(n) signal at RX1 only contains the RF signal, such as shown in equation 164.

[00054]$\begin{array}{cc}\begin{array}{c}{v}_{p,1}\left(n\right)=\left\{u_2\left(n\right).Math.c_{22}\left(n\right)\right\}.Math.d_{21}\left(n\right)\\ =u_2\left(n\right).Math.f_{21}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.164\end{array}$

[0176] At the same time, after entering the crosstalk response of the receiving end, the equations 151 and 152 are substituted into the equation 126 type to obtain the v.sub.p,2(n) signal, and since it only contains the signal u.sub.2(n) of TX2, according to the above equation 155, the new crosstalk parameters could be obtained and as shown in equation 165.

[00055]$\begin{array}{cc}\begin{array}{cc}{v}_{p,2}\left(n\right).Math.=& .Math.\left\{u_2\left(n\right).Math.c_{21}\left(n\right)\right\}.Math.d_{12}\left(n\right)+\\ & .Math.\left\{u_2\left(n\right).Math.c_{22}\left(n\right)\right\}.Math.d_{22}\left(n\right)\\ =& .Math.u_2\left(n\right).Math.f_{12}\left(n\right)+u_2\left(n\right).Math.f_{22}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.165\end{array}$

[0177] According to the fourth set of conditions, only the input signal r.sub.2(n) is transmitted through the switch before the crosstalk at the receiving end, and the V.sub.p,2(n) signal at RX2 would only contains the crosstalk RF signal r.sub.2(n), as shown in equation 166.

[00056]$\begin{array}{cc}\begin{array}{c}{v}_{p,2}\left(n\right)=\left\{u_2\left(n\right).Math.c_{22}\left(n\right)\right\}.Math.d_{22}\left(n\right)\\ =u_2\left(n\right).Math.f_{22}\left(n\right)\end{array}& \mathrm{equation}.Math..Math.166\end{array}$

[0178] Subsequently, after entering the digital terminal, it is assumed that the post-processing parameters P.sub.1(n) P.sub.2(n) P.sub.3(n) P.sub.4(n) can counter the v.sub.p,1(n) signal of the crosstalk response of RX1 and v.sub.p,2(n) of the crosstalk response of RX2, so the equations 164 and 166 are substituted into equation 127, and z.sub.1(n) at RX1 could be obtained and expressed equation 167.

[00057]$\begin{array}{cc}{z}_1\left(n\right)=p_3\left(n\right).Math.\left\{u_2\left(n\right).Math.f_{21}\left(n\right)\right\}+p_2\left(n\right).Math.\left\{u_2\left(n\right).Math.f_{22}\left(n\right)\right\}& \mathrm{equation}.Math..Math.167\end{array}$

[0179] The equation 145 could be rearranged into z.sub.1(n) at RX1 corresponding to u.sub.2(n) of TX2, and then the processing vector suppresses the received crosstalk response, as shown in the following equation 168.

[00058]$\begin{array}{cc}{z}_1\left(n\right)=u_2\left(n\right).Math.\left\{p_3\left(n\right).Math.f_{21}\left(n\right)\right\}+u_2\left(n\right).Math.\left\{p_2\left(n\right).Math.f_{22}\left(n\right)\right\}& \mathrm{equation}.Math..Math.168\end{array}$