TWI434533B - Method and apparatus for calculating smart antenna weight of spatial division multiple access system - Google Patents

Description

Smart antenna weight calculation method and device for space division multiplexing access system

The present invention relates to a Spatial Division Multiple Access (SDMA) system, and more particularly to a smart antenna weight calculation method and apparatus for a space division multiplexing access system. The smart antenna weight calculation method and device are used to improve the performance degradation caused by intermodulation products caused by nonlinear distortion in the multi-frequency carrier operation of the space division multiplexing access system.

The space division multiplexing access technology is a newly developed division of labor technology. The space division multiplexing access technology can improve the smart antenna gain, make the power control more reasonable and effective, and significantly improve the system capacity. On the other hand, the space division multiplexing technology can also weaken interference from the outside world and reduce interference to other electronic systems.

The core technology for implementing the space division multiplexing access system is the application of beam forming technologies. Ideally, the space division multiplexing access system will require its array antenna to assign a beam to each user. In this way, the space division multiplexing system can distinguish the wireless signals of each user according to the spatial location of the user. In other words, users in different locations can use the same frequency and the same user code at the same time without interfering with each other.

For mobile users, due to the complexity of mobile wireless channels, the algorithms for dynamic capture, recognition and tracking of multi-user signals and channel identification in smart antennas are extremely complex, which makes the digital signal processing operation extremely high. Claim.

The space division multiplexing access technology uses spatial segmentation to form different channels, so that users of different spaces can use the same frequency and the same user code at the same time without interfering with each other. For example, multiple antennas are used on a single satellite, and the beams of each antenna are directed at different areas of the Earth's surface. Earth stations in different regions on the ground will not interfere with each other even if they use the same frequency to communicate with the satellites at the same time.

The beamforming technology of the space division multiplexing system utilizes different spatial locations of the users to distinguish different users, and the array antenna dynamically tracks a plurality of desired users with a plurality of high gain narrow beams. In receive mode, signals from outside the narrow beam are suppressed. However, in the transmit mode, the power of the signal received by the desired user can be maximized while minimizing the interference experienced by undesired users outside the narrow beam illumination range.

The current space division multiplexing access system can be combined with Multiple Input Multiple Output (MIMO) and Multi-Carrier Modulation (MCM) technologies. Therefore, the space division multiplexing access system Both the receiver and the transmitter require a Multi-Carrier Power Amplifier (MCPA).

Please refer to FIG. 1. FIG. 1 is a system block and multi-carrier modulation diagram of a space division multiplexing access system. The space division multiplexing access system 1 is a combination of multiple input multiple output and multi-carrier modulation technology, and includes a base station B1 and three mobile users U11 to U13. The base station B1 and the mobile users U11 to U13 have a plurality of antennas to form a smart antenna. In FIG. 1, by assigning a plurality of antenna weights to a plurality of antennas for each mobile user, it is possible to distinguish mobile users by spatial position, wherein beams P11 to P13 are beams used by mobile users U11 to U13, respectively.

In general, if a communication system using multi-carrier modulation technology does not have a high-linear multi-carrier power amplifier, its multi-carrier signal will have severe nonlinear distortion after passing through the multi-carrier power amplifier. This nonlinear distortion, in turn, can be referred to as ghost interference, which occurs when the amplitude of the multi-carrier signal is too large.

Referring to FIG. 2A and FIG. 2B, FIG. 2A is a graph of input signal power and output signal power and gain of the power amplifier, and FIG. 2B is a graph of the relative phase of the input signal power of the power amplifier and the output signal. The graphs of Figures 2A and 2B are obtained by applying the ADS2009 software analysis power amplifier (Motorola MRF9742), and Figures 2A and 2B are used to explain ghost interference.

In FIG. 2A and FIG. 2B, when the input signal power increases, the output signal power and phase may be nonlinear (refer to the input signal power corresponding to the output signal power and the relative phase curve C101, C111 and the reference input signal power corresponding gain). Curve C105). According to Shimbo's model (Shimbo, O.; "Effects of intermodulation, AM-PM conversion, and additive noise in multicarrier TWT systems," Proceedings of IEEE , vol. 59, Iss. 2, 1971, pp. 230-238) When the analog multi-carrier signal passes through the multi-carrier power amplifier, it can also be known that the output signal power and phase have nonlinear distortion (refer to the input signal power corresponding to the output signal power and the relative phase curve C102 and C112).

If the input signal includes at least one carrier modulation signal as an example (ω _i = ω ₁ , ω ₂ = ω ₁ + Δω, ω ₃ = ω ₁ + 2 Δω), it can be understood that the P _{1d B} point in the representative linear region will be The input signal contains a carrier signal and changes (refer to the input signal power corresponding to the output signal power and relative phase curve C103, C104, C113 and C114, the curves C103 and C113 correspond to the input signal including the ω ₁ and ω ₃ carrier signals The curves C104 and C114 correspond to the curve of the input signal including the ω ₂ carrier signal).

The conventional method of allocating smart antenna weights is assigned to the i-th antenna of the kth mobile user Ulk (in this example, k is an integer of 1-3), and the antenna weight is a _i,k Where a _ik is the gain of the antenna weight, Ψ _ik is the phase of the antenna weight, i is an integer from 1 to N, and N is the total number of antennas. Because the gain of the antenna weight of each antenna is different. Therefore, when the multi-carrier input signal signal is increased, the nonlinear intensity/intensity (AM/AM) and intensity/phase (AM/PM) conversions are generated by the gain of the antenna weight of each antenna. The amplitude and phase error between the antennas are shown in Figures 3A and 3B.

Please refer to FIG. 3A. FIG. 3A is a graph of input signal power and output signal power of each antenna in a conventional smart antenna system. In FIG. 3A, curve C201 is a curve of the input signal power of each antenna, curve C202 is a curve of the output signal power of each antenna under normal conditions (without intermodulation change), and curve C203 is a nonlinear distortion. A plot of the output signal power of each antenna in the case (which produces intermodulation). It can be seen from Fig. 3A that under normal conditions, the amplitude of the output signal of each antenna will be amplified by the same gain according to the amplitude of the input signal, but the nonlinear AM/AM conversion will be caused by the difference in the gain of the antenna weight of each antenna. error.

Please refer to FIG. 3B. FIG. 3B is a graph of input signal phase and output signal phase of each antenna in the conventional smart antenna system. In FIG. 3B, curve C211 is a curve of the input signal phase of each antenna, and curve C212 is a curve of the output signal phase of each antenna under normal conditions (without intermodulation change), and curve C213 is a nonlinear distortion. The curve of the output signal phase of each antenna in the case (which produces intermodulation). It can be seen from Fig. 3B that under normal conditions, the phase of the input signal of each antenna should be amplified by the amplifier and the relative phase between the antennas should be the same, but the nonlinear AM/PM conversion will be caused by the difference of the antenna weight of each antenna. Phase error.

Please refer to FIG. 4. FIG. 4 is a radiation field diagram of a conventional smart antenna system. Curve 301 is a radiation field map in the case of a simulated uniform distribution, and curve 302 is a conventional smart antenna weight calculation method using a Minimization Mean Square Error (MMSE) algorithm without intermodulation distortion. Radiation field diagram, Fig. 303 is a radiation field diagram of the conventional smart antenna weight calculation method using a minimum mean square error algorithm and intermodulation distortion. From Fig. 4, the influence of the amplitude and phase error between the antennas on the radiation field shape of the smart antenna can be obtained. Under the condition of minimizing the mean square error algorithm and intermodulation distortion, the antenna gain is reduced by about 0.23B, and in the direction of the interference source ( The 20°/40°/60°) side-wave suppression effect is about 10 dB worse. Accordingly, a highly linear multi-carrier power amplifier is required to prevent ghost interference. However, the high linearity of multi-carrier power amplifiers is costly and will result in high manufacturing costs.

Please refer to FIG. 5. FIG. 5 is a block diagram of a mobile user of the space division multiplexing access system. The block diagram of the base station of the space division multiplexing access system is substantially the same as the block diagram of the mobile user 2. Therefore, only the block diagram of the mobile user 2 will be described here as an example. The mobile user 2 is, for example, a mobile phone or other handheld device, and includes a weight calculation unit 21, a plurality of antennas T1 to TN, a receiving/transmitting switch SW1 to SWN, transceivers TxRx1 to TxRxN, analog digital converters ADC1 to ADCN, and a digital analogy. Converters DAC1 to DACN, signal processor 22 and upper layer circuit 23. The upper layer circuit 23 includes a communication protocol processor 231 and an application layer circuit 232. The coupling relationship of each component in the mobile user 2 is as shown in FIG. 5, so it will not be described again.

The antennas T1 to TN receive or transmit N radio signals, respectively, and the reception/transmission switches SW1 to SWN are switched in accordance with reception/transmission of the transceivers TxRx1 to TxRxN. In the receiving mode (the receiving/transmitting switches SW1 to SWN are switched to the receiving state), the transceivers TxRx1 to TxRxN receive N radio signals from the antennas T1 to TN, respectively, and process the N radio signals to generate N. The analog signals are given to analog converters ADC1~ADCN. In the transmission mode (the receiving/transmitting switches SW1 to SWN are switched to the transmitting state), the transceivers TxRx1 to TxRxN respectively process the N analog signals, and generate N wireless signals to the antennas T1 to TN.

The analog-to-digital converters ADC1 to ADCN are used to convert N analog signals into digital signals, respectively. The digital analog converters DAC1 to DACN are used to convert N digital signals into analog signals, respectively. The signal processor 22 is configured to receive N digital signals of the analog-to-digital converters ADC1 - ADCN, and obtain a plurality of receiving bits according to the decoding of the N digital signals, and encode the plurality of transmitting bits to generate N digital signals. Give digital analog converters DAC1 ~ DACN.

The protocol processor 231 is configured to process signals below the application layer above the physical layer, and the application layer circuit 232 processes signals at the application layer. In more detail, the protocol processor 231 processes a plurality of received bits and generates processing information for the application layer circuit 232 to cause the application layer circuit 232 to execute the corresponding application based on the processing information. At the same time, after executing the application, the application layer circuit 232 generates corresponding processing information to the protocol processor 231, and the communication protocol processor 231 generates a plurality of transmission bits accordingly.

In FIG. 5, the weight calculation unit 21 is configured to calculate a plurality of antenna weights, and the antenna weights are respectively transmitted to the transceivers TxRx1 to TxRxN, so that the mobile user performs beam forming on the plurality of signals, thereby Correct the phase and vibration error caused by the interference signal and noise caused by the transmission process. Simply put, the analog signal of each transceiver TxRx1 ~ TxRxN will be multiplied by the corresponding antenna weight. As described above, the gain of the antenna weight corresponding to each of the transceivers TxRx1 to TxRxN is different, and therefore the transceivers TxRx1 to TxRxN are required to have a high linearity multi-carrier power amplifier.

The weight calculation unit 21 can use a conventional smart antenna weight calculation method, and the conventional smart antenna weight calculation method adopts a minimized mean square error algorithm, and the detailed content thereof is explained as follows. Please refer to FIG. 5 and FIG. 6. FIG. 6 is a flowchart of a conventional smart antenna weight calculation method. First, in step S31, the weight calculation unit 21 receives N received signals of the N transceivers TxRx1 to TxRxN (assuming that the mobile user 2 is the kth mobile user), wherein the N received signals may be referred to as a received signal vector. X _k =[X _1,k X _2,k ...X _N,k ] ^T .

In step S32, the weight calculation unit 21 calculates a covariance matrix of the covariance of the received signal vector X _k . . Then, in step S33, a cross-correlation vector between the received signal vector X _k and the training signal vector D _k is calculated. . Thereafter, in step S34, according to the inverse matrix of the covariance of the received signal vector X _k Cross-correlation vector between the received signal vector X _k and the training signal vector D _k Get the best antenna weight vector ,among them . In addition, for the weight unit of the base station, it calculates the antenna weight vector of the kth mobile user. The way is also like the above.

Optimal antenna weight vector Each antenna weight a _i,k The gain a _ik (where i is an integer from 1 to N) is not the same, so the received signal will be amplified by the multi-carrier power amplifier, and nonlinear distortion may occur, which may affect the performance of the entire space division multiplexing system. .

Embodiments of the present invention provide a smart antenna weight calculation method for a space division multiplexing access system. This method is used to calculate the antenna weight vector of the kth mobile user in the space division multiplexing access system. And perform the following steps. Step a: Obtain the received signal vector of the kth mobile user Where the received signal vector Each of the received signals comes from the corresponding antenna of the kth mobile user. Step b: According to the phase vector of the kth mobile user Receive signal vector Get the received signal vector Analytical signal vector after beamforming , where i represents the number of iterations of the iteration. Step c: Calculate the analytical signal vector Squared distance from the training signal vector D _{k of} the kth mobile user . Step d: Calculate the distance squared Phase vector Gradient . Step e: According to the phase vector , step parameters ξ and gradient Get the phase vector in the next iteration of the iteration . Step f: Determine whether to end the iterative operation. Step g: if at least one condition of ending the iterative operation is satisfied, according to the phase vector Output the antenna weight vector of the kth mobile user Antenna weight vector The gain of each of the antenna weights is the same as each other.

According to an embodiment of the present invention, the at least one condition includes at least one of the following: whether the current iteration operation loop number i satisfies a predetermined iteration operation loop number; and a phase vector Phase vector The distance between them has already reached the predetermined distance.

According to an embodiment of the invention, the above method further comprises the following steps. Step h: If it is judged that the iterative operation is not ended, the number of times of the iteration operation is increased by 1, and step a is set to step f again. Step i: Initialize the iteration operation loop number i is 1, and initialize the phase vector .

Embodiments of the present invention provide a smart antenna weight calculation apparatus for a space division multiplexing access system. The device is used to calculate the antenna weight vector of the kth mobile user in the space division multiplexing access system And includes a distance square calculation device, a gradient calculation device, a step parameter generator, a step parameter product, a vector subtractor, a decision maker, and a buffer. The distance square computing device is used to calculate the received signal vector of the kth mobile user Analytical signal vector after beamforming Squared distance from the training signal vector D _{k of} the kth mobile user ,among them It is the phase vector of the kth mobile user, and i represents the number of iterations. Gradient calculation device for calculating distance squared Phase vector Gradient . The step parameter generator is used to provide the step parameter ξ. Scalar multiplier for gradient Multiply by the step parameter ξ to obtain the multiplication result . Vector subtractor for phase vector Subtract multiplication result To get the phase vector in the next iteration of the iteration . The decision maker determines whether to end the iterative operation if it satisfies at least one condition of ending the iterative operation, according to the phase vector Output the antenna weight vector of the kth mobile user Antenna weight vector The gain of each of the antenna weights is the same as each other. If it is judged that the iterative operation is not ended, the phase vector is obtained. Send to the buffer. The buffer is used to send its stored phase vector to the next iteration of the iteration loop.

In summary, the smart antenna weight calculation method for the space division multiplexing access system provided by the embodiment of the present invention can generate a fixed gain antenna weight vector, thereby avoiding the occurrence of a multi-carrier power amplifier of each transceiver. Linear distortion (ie, reduced ghost interference).

The detailed description of the present invention and the accompanying drawings are to be understood by the claims The scope is subject to any restrictions.

[Mobile User/Base Station Example]

Please refer to FIG. 7. FIG. 7 is a block diagram of a mobile user of a space division multiplexing access system according to an embodiment of the present invention. The block diagram of the base station of the space division multiplexing access system is substantially the same as the block diagram of the mobile user 4. Therefore, only the block diagram of the mobile user 4 will be described here as an example. The mobile user 4 is, for example, a mobile phone or other handheld device, and includes a fixed weight calculation unit 41, a plurality of antennas T1 to TN, receiving/transmitting switches SW1 to SWN, transceivers TxRx1 to TxRxN, analog digital converters ADC1 to ADCN, and digital digits. Analog converters DAC1 to DACN, signal processor 42 and upper layer circuit 43. The upper layer circuit 43 includes a communication protocol processor 431 and an application layer circuit 432.

The plurality of antennas T1 to TN are respectively coupled to the receiving/transmitting switches SW1 to SWN, and the receiving/transmitting switches SW1 to SWN are respectively coupled to the transceivers TxRx1 to TxRxN. The transceivers TxRx1 to TxRxN are respectively coupled to the analog-to-digital converters ADC1 to ADCN, and the transceivers TxRx1 to TxRxN are also coupled to the digital analog converters DAC1 to DACN, respectively. The analog-to-digital converters ADC1 - ADCN and the digital analog converters DAC1 - DDAC are coupled to the signal processing circuit 42 , and the signal processing circuit is coupled to the upper layer circuit 43 . The fixed gain weight calculation unit 41 is coupled to the transceivers TxRx1 to TxRxN, the signal processor 42 and the upper layer circuit 43.

The antennas T1 to TN receive or transmit N radio signals, respectively, and the reception/transmission switches SW1 to SWN are switched in accordance with reception/transmission of the transceivers TxRx1 to TxRxN. In the receiving mode (the receiving/transmitting switches SW1 to SWN are switched to the receiving state), the transceivers TxRx1 to TxRxN receive N radio signals from the antennas T1 to TN, respectively, and process the N radio signals to generate N. The analog signals are given to analog converters ADC1~ADCN. In the transmission mode (the receiving/transmitting switches SW1 to SWN are switched to the transmitting state), the transceivers TxRx1 to TxRxN respectively process the N analog signals, and generate N wireless signals to the antennas T1 to TN.

The analog-to-digital converters ADC1 to ADCN are used to convert N analog signals into digital signals, respectively. The digital analog converters DAC1 to DACN are used to convert N digital signals into analog signals, respectively. The signal processor 42 is configured to receive N digital signals of the analog-to-digital converters ADC1 - ADCN, and obtain a plurality of receiving bits according to decoding of the N digital signals, and encode the plurality of transmitting bits to generate N digital signals. Give digital analog converters DAC1 ~ DACN.

The protocol processor 431 is configured to process signals below the application layer above the physical layer, and the application layer circuit 432 processes signals at the application layer. In more detail, the protocol processor 431 processes a plurality of received bits and generates processing information for the application layer circuit 432 to cause the application layer circuit 432 to execute the corresponding application based on the processing information. At the same time, after executing the application, the application layer circuit 432 generates corresponding processing information to the protocol processor 431, and the protocol processor 431 generates a plurality of transfer bits accordingly.

In FIG. 7, the fixed gain weight calculation unit 41 is configured to calculate a plurality of antenna weights (where the gains of the antenna weights are the same as each other), and the antenna weights are respectively transmitted to the transceivers TxRx1 to TxRxN, so that the mobile user is more The signals are beam forming, thereby correcting the phase and vibration errors caused by the interference signals and noise received by the transmission process. Simply put, the analog signal of each transceiver TxRx1 ~ TxRxN will be multiplied by the corresponding antenna weight. As described above, the gain of the antenna weight of each transceiver corresponding to TxRx1~TxRxN is the same, so the possibility of nonlinear distortion of the multi-carrier power amplifier of each transceiver TxRx1~TxRxN can be reduced.

Accordingly, the mobile user 4 using the fixed gain weight calculation unit 41 can eliminate the need for a highly linear multi-carrier power amplifier compared to the mobile user of the space division multiplexing system, and therefore, the manufacturing cost of the mobile user 4 can be reduced. The hardware implementation of the fixed gain weight calculation unit 41 is as shown in FIG. 8, or the fixed gain weight calculation unit 41 can also be implemented using the microprocessor or firmware to perform the method of FIG. In summary, the fixed gain weight calculation unit 41 is not intended to limit the present invention.

[Embodiment of Fixed Gain Weight Calculation Unit]

Please refer to FIG. 8. FIG. 8 is a detailed block diagram of a fixed gain weight calculation unit according to an embodiment of the present invention. The fixed gain weight calculation unit 5 includes a distance square calculation means 51, a gradient calculation means 52, a step size generator 53, a scalar multiplier 54, a vector subtractor 55, a decision maker 56, and a buffer 57. The distance calculation device 51 is coupled to the gradient calculation device 52 , and the gradient calculation device 52 is coupled to the scalar multiplier 54 . The scalar multiplier 54 is coupled to the step parameter generator 53 and the vector subtractor 55, and the vector subtractor 55 is coupled to the decider. The decision maker 56 is coupled to the buffer 57, and the buffer 57 is coupled to the distance square computing device 51.

The fixed gain weight calculation unit 5 calculates the antenna weight vector of the kth mobile user using an iteration method. Antenna weight vector The phase vector is expressed as Antenna weight vector Phase vector Respectively as follows,

The gains of the antennas are equal to each other, that is, a _{1, k} = a _{2, k} = ... = a _{N, k} .

The distance square computing device 51 receives a received signal vector including a plurality of received signals ( =[X _1,k X _2,k ...X _N,k ], each received signal comes from the corresponding transceiver or antenna) and the phase vector obtained from the ith iteration of the loop And calculate the received signal vector accordingly Analytical signal vector after beamforming Squared distance from the training signal vector D _k , that is, calculation Wherein the training signal is represented as a vector D _{k D k = [D 1, k} D 2, k ... D N, k] T, D _k and the training signal vector signal vector is known in advance.

Next, the gradient calculation device 52 calculates the distance squared Correct Gradient . Thereafter, the step parameter generator 53 generates a step parameter ξ, and the scalar multiplier 53 sets the distance squared. Pair Gradient Multiply by the step parameter ξ to obtain the multiplication result .

The step parameter ξ can ideally be the distance squared The decreasing amount of the maximum decreasing direction, in other words, the step parameter ξ ideally may be the magnitude of the steepest slope direction. The step parameter generator 53 can define the size of the step parameter ξ by the user (the step parameter ξ is, for example, 0.01), or obtain the better step parameter 透过 through the step parameter calculation method.

Next, the vector subtractor 55 returns the phase vector obtained by the ith iteration of the iteration. Subtract multiplication result To obtain the phase vector obtained by the i++1st iteration loop In other words, . Then, the decider 56 determines whether the number of times of the iteration operation loop has satisfied the number of times of the predetermined iteration operation loop (the number of times of the predetermined iteration operation loop is, for example, 3), or the i-th and i+1-time iteration operation loop Obtained phase vector versus Whether the distance between them has already reached the predetermined distance.

If the number of iterations of the iteration operation has already met the number of times of the predetermined iteration, or the phase vector obtained by the i-th and i+1 iterations versus The distance between the distances has been met by the decision maker 56, and the phase vector obtained by the decision maker 56 according to the i+1th iteration operation loop Output antenna weight vector ,at this time, .

Conversely, if the number of iterations of the iteration does not satisfy the number of times of the predetermined iteration, and the phase vector obtained by the i-th and i+1 iterations versus If the distance between the distances does not satisfy the predetermined distance, the decision maker 56 returns the phase vector obtained by the i+1th iteration operation loop. It is sent to the buffer 57. The buffer 57 outputs its stored phase vector for use in the next iteration loop.

[Embodiment of smart antenna weight calculation method for space division multiplexing access system]

Please refer to FIG. 9. FIG. 9 is a flowchart of a smart antenna weight calculation method for a space division multiplexing access system according to an embodiment of the present invention. The gain of each antenna weight of the antenna weight vector calculated by the smart antenna weight calculation method of FIG. 9 is the same as each other. In addition, the smart antenna weight calculation method of FIG. 9 is only one embodiment of the present invention, which is not intended to limit the present invention. invention. The following is to calculate the antenna weight vector of the kth mobile user. Give an example for explanation.

First, in step S61, a received signal vector is obtained. . Then, initialize the number of iterations of the iteration operation i (i = 1) and the phase vector (i=1). Then, in step S63, according to the phase vector Receive signal vector Get the received signal vector Analytical signal vector after beamforming . Then, in step S64, the analytical signal vector is calculated. Squared distance from the training signal vector D _k , that is, calculation .

Next, in step S65, the analytical signal vector is calculated. Squared distance from the training signal vector D _k Correct Gradient . Then, in step S66, according to the current phase vector , step parameters ξ and analytical signal vectors Squared distance from the training signal vector D _k Correct Gradient Get the phase vector in the next iteration of the iteration , where the phase vector in the next iteration of the iteration .

Next, in step S67, it is determined whether the iterative operation is ended, if the number of times of the iteration operation loop has satisfied the number of times of the predetermined iteration operation loop, or the phase vector versus If the distance between the distances has already reached the predetermined distance, step S69 is performed; otherwise, step S68 is performed. In step S68, the number of iterations is updated, even if i=i+1. In step S69, according to the phase vector in the next iteration loop obtained in step S66 Output antenna weight vector ,at this time, .

[Possible efficacy of the embodiment]

According to the embodiment of the present invention, the smart antenna weight calculation method and apparatus for the space division multiplexing access system of the above embodiment give the same gain to each antenna, thereby reducing the multi-carrier in the base station or the mobile user's transceiver. The occurrence of nonlinear distortion of the power amplifier can reduce the ghost interference of the entire space division multiplexing access system and improve system performance (including the improvement of the bit error rate and the suppression of the side lobe intensity).

Referring to FIG. 10A, FIG. 10A is a graph of input signal power and output signal power for each antenna in the embodiment. Curve C401 is a plot of input signal power, curve C402 is a plot of output signal power without intermodulation distortion, and curve C403 is a plot of output signal power with intermodulation distortion. This embodiment only needs to change the phase of the weight without changing its size. Therefore, the input signal amplitude of each antenna is injected into each amplifier with the same amplitude (assuming each amplifier has the same performance), and the ideal input signal amplitude should be The input signal amplitude is multiplied by the gain of the amplified gas. Even if there is nonlinear distortion such as AM/AM conversion, it will only increase or decrease the signal amplitude of each antenna.

Please refer to FIG. 10B. FIG. 10B is a graph of input signal phase and output signal phase of each antenna in the embodiment. Curve C411 is a plot of the phase of the input signal, curve C412 is a plot of the phase of the output signal without intermodulation distortion, and curve C413 is a plot of the phase of the output signal with intermodulation distortion. For the amplifier, the signal changes will cause phase error, so even if there is nonlinear distortion, it will only increase or decrease the signal phase of each antenna.

Please refer to FIG. 11, which is a radiation field diagram of the embodiment. Curve C501 is a radiation field diagram in the absence of intermodulation distortion, and curve C502 is a radiation field diagram in the presence of intermodulation distortion. Since the amplitude and phase between the antennas are synchronously increased or decreased, there is no error between the antennas due to nonlinear distortion of the amplifier. Therefore, it can be known from the comparison curves C501 and C502 that this embodiment can reduce the influence of the amplifier nonlinear misalignment system.

The above description is only an embodiment of the present invention, and is not intended to limit the scope of the invention.

1. . . Space division multiplexing access system

B1. . . Base station

U11~U13, 2, 4. . . Mobile users

P11～P13. . . Beam

C101～C105, C111～C114, C201～C203, C211～C213, C301～C303, C401～C403, C411～C413, C501, C502. . . curve

T1 ~ TN. . . antenna

SW1~SWN. . . Receive/transmit switch

TxRx1~TxRxN. . . Transceiver

ADC1~ADCN. . . Analog digital converter

DAC1 ~ DACN. . . Digital analog converter

twenty one. . . Weight calculation unit

twenty two. . . Signal processor

twenty three. . . Upper circuit

231. . . Protocol processor

232. . . Application layer circuit

41, 5. . . Fixed gain weight calculation unit

51. . . Distance square computing device

52. . . Gradient computing device

53. . . Step parameter generator

54. . . Scalar multiplier

55. . . Vector subtractor

56. . . Decision maker

57. . . buffer

S61～S69. . . Step flow

1 is a system block and multi-carrier modulation map of a space division multiplexing access system.

2A is a graph of input signal power and output signal power and gain for a power amplifier.

Figure 2B is a plot of the input signal power of the power amplifier versus the phase of the output signal.

3A is a graph of input signal power and output signal power for each antenna in a conventional smart antenna system.

Figure 3B is a graph of input signal phase and output signal phase for each antenna in a conventional smart antenna system.

4 is a radiation field diagram of a conventional smart antenna system.

Figure 5 is a block diagram of a mobile subscriber of a space division multiplexed access system.

6 is a flow chart of a conventional smart antenna weight calculation method.

FIG. 7 is a block diagram of a mobile user of a space division multiplexing access system according to an embodiment of the present invention.

FIG. 8 is a detailed block diagram of a fixed gain weight calculation unit according to an embodiment of the present invention.

FIG. 9 is a flowchart of a smart antenna weight calculation method for a space division multiplexing access system according to an embodiment of the present invention.

Figure 10A is a graph of input signal power and output signal power for each antenna in the embodiment.

Figure 10B is a graph of the phase of the input signal and the phase of the output signal for each antenna in the embodiment.

Figure 11 is a radiation field diagram of the embodiment.

5. . . Fixed gain weight calculation unit

51. . . Distance square computing device

52. . . Gradient computing device

53. . . Step parameter generator

54. . . Scalar multiplier

55. . . Vector subtractor

56. . . Decision maker

57. . . buffer

Claims (10)

A smart antenna weight calculation method for a space division multiplexing access system for calculating an antenna weight vector of one of the kth mobile users in a space division multiplexing access system , including: step a: obtaining one of the kth mobile users to receive a signal vector The received signal vector Each of the received signals is from a corresponding antenna of the kth mobile user; step b: according to a phase vector of the kth mobile user And the received signal vector Obtain the received signal vector An analytical signal vector after beamforming Where i represents the number of turns of a iteration operation; step c: calculates the vector of the resolved signal a distance squared from the training signal vector D _k of one of the kth mobile users Step d: calculate the square of the distance The phase vector a gradient Step e: according to the phase vector , one-step parameter ξ and the gradient Get the phase vector in the next iteration of the iteration Step f: determining whether to end the iterative operation; and step g: if at least one condition of ending the iterative operation is satisfied, according to the phase vector Outputting the antenna weight vector of the kth mobile user The antenna weight vector The gain of each of the antenna weights is the same as each other. The method for calculating a smart antenna weight for a space division multiplexing access system according to claim 1, wherein the distance is squared . The method for calculating a smart antenna weight for a space division multiplexing access system according to claim 1, wherein the phase vector . The method for calculating a smart antenna weight for a space division multiplexing access system according to claim 1, wherein the at least one condition includes at least one of the following: whether the current number of times of the iteration operation is satisfied a predetermined iteration operation loop number; and the phase vector Phase vector The distance between them has already reached a predetermined distance. The method for calculating a smart antenna weight for a space division multiplexing access system according to claim 1, wherein the step parameter is a magnitude of a steepest slope direction. The method for calculating a smart antenna weight for a space division multiplexing access system according to claim 1, further comprising: step h: if it is judged that the iterative operation is not ended, the number of times of the iteration operation is i Add 1 and repeat step a to set step f. The smart antenna weight calculation method for the space division multiplexing access system according to claim 1, further comprising: step i: initializing the iteration operation loop number i is 1, and initializing the phase vector . A smart antenna weight calculation device for a space division multiplexing access system for calculating an antenna weight vector of one of the kth mobile users in a space division multiplexing access system The method includes: a distance square computing device, configured to calculate a received signal vector of the kth mobile user An analytical signal vector after beamforming a distance squared from the training signal vector D _k of one of the kth mobile users ,among them a phase vector of the kth mobile user, and i represents the number of times of the iteration of the iteration; a gradient calculation device for calculating the square of the distance The phase vector a gradient a one-step parameter generator for providing a one-step parameter ξ; a scalar multiplier for the gradient Multiply the step parameter ξ to obtain a multiplication result a vector subtractor for the phase vector Subtract the result of the multiplication To get the phase vector in the next iteration of the iteration a decision maker that determines whether to end the iterative operation, and if at least one condition that ends the iterative operation is satisfied, according to the phase vector Outputting the antenna weight vector of the kth mobile user The antenna weight vector The gain of each of the antenna weights is the same as each other, and if it is judged that the iterative operation is not ended, the phase vector is Sent to a buffer; and the buffer is used to send its stored phase vector to the next iteration loop. A smart antenna weight calculation device for a space division multiplexing access system according to claim 8, wherein the distance is squared . The smart antenna weight calculation device for a space division multiplexing access system according to claim 8, wherein the step parameter is a magnitude of a steepest slope direction.