An adaptive beam alignment method for millimeter wave communication system based on UCB

Abstract

The present invention discloses an adaptive beam alignment method for a millimeter wave communication system based on UCB, comprising the following steps: S1, construction of a signal model during iterative beam scanning; S2, setting a UCB scanning strategy to eliminate noise; S3, adaptive beam alignment. The method can quickly and accurately lock the optimal beam in different signal-to-noise ratio scenarios through iterative spatial scanning of the UCB strategy. After finding the optimal beam, the beam is rotated by maximum likelihood estimation to further improve the beamforming gain. The method is completely unaware of channel information and noise variance, and does not require a preset non-adaptive scanning time. It is applicable to a wider range of millimeter wave communication scenarios and is therefore more practical.

Description

Technical Field

[0001] The present invention relates to the technical field of digital communication, and in particular to an adaptive beam alignment method for a millimeter wave communication system based on UCB. Background Art

[0002] Millimeter wave communication systems have ample unused spectrum and are gradually becoming one of the important technologies in the mainstream mobile communication field. However, millimeter wave signal transmission in the high-frequency band also suffers from higher path loss and stronger fading. In order to obtain reliable data transmission, the communication transceiver needs to compensate for the higher path loss through beamforming gain. However, the narrower directional beam makes it more difficult to establish and maintain the communication link than the low-frequency communication system that uses omnidirectional (or semi-omnidirectional covering the entire sector) transmission. This is because narrower beams are more sensitive to directional errors. Therefore, mobile millimeter wave communication systems usually need to frequently perform beam searches to ensure the accuracy of beam alignment.

[0003] A common method in analog beam alignment is to use beam search based on spatial scanning. This method scans beams in different directions in the beam codebook to collect signals, and selects the beam that is best aligned with the main path direction of the channel based on the signals. The search time for scanning beams in this type of method is often fixed, and it is prone to non-adaptiveness when applied to communication scenarios. Because the signal-to-noise ratio of users in the communication coverage area is inconsistent, and the fixed search time cannot meet the beam alignment quality of different users, the longer search time prevents high-signal-to-noise ratio users from communicating quickly, and the shorter search time makes it difficult for low-signal-to-noise ratio users to successfully establish a communication link.

[0004] In order to solve the non-adaptive problem of the above-mentioned beam alignment method, it is necessary to design a beam alignment algorithm whose search time is quickly and adaptively adjusted according to the signal-to-noise ratio of different users, which is suitable for a wider range of communication scenarios. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides an adaptive beam alignment method for a millimeter wave communication system based on UCB. The method does not need to know the instantaneous and statistical information of the channel and the noise variance, and can quickly and adaptively adjust the beam search time under different signal-to-noise ratios, which not only reduces the time complexity but also ensures the accuracy of beam alignment.

[0006] In order to solve the above technical problems, the technical solution of the present invention is:

[0007] A UCB-based millimeter wave communication system adaptive beam alignment method comprises the following steps:

[0008] S1. Construction of signal model during beam iterative scanning

[0009] Iterative scanning signal model, each iteration Rx scans a beam, t represents the number of iterations. Considering the narrow bandwidth signal model, in the tth iteration, Rx scans beam l, and the corresponding acquisition signal can be expressed as

[0010]

[0011] in is complex Gaussian noise, s l (t) is the pilot signal sent by Tx in a scanning cycle, and |s l (t)| 2 =PT,P T is the transmit power of Tx; h l Equivalent channel after applying the transmit and receive beams

[0012]

[0013] in is the channel between Tx and Rx, N T and N R are the number of antennas of the base station and UE respectively, and the matched filter output of the collected signal is

[0014]

[0015] The cumulative matched filter output of beam l is [f l (1), ..., f l (m), ..., f l (n l (t))] T , where n l (t) is the cumulative number of scan cycles of Rx scan l, and its joint matched filter output is updated as

[0016]

[0017] S2. Set UCB scanning strategy to eliminate noise

[0018] Each iteration t selects the scan UCB weight function U l The largest beam l in (t) ＊ (t), that is

[0019]

[0020] Where c>0 is an adjustable parameter, is the estimated value of the noise standard deviation;

[0021] S3, Adaptive beam alignment

[0022] S3-1, Rx performs iterative beam scanning initialization;

[0023] S3-2, iterative beam scanning based on UCB scanning strategy;

[0024] S3-3, beam selection.

[0025] Preferably, the construction rule of the signal model in step S1 is: considering that the transmitting end Tx sends a pilot signal and the receiving end Rx performs spatial scanning, if Tx is a base station, then Rx is a UE, corresponding to the beam search on the UE side; conversely, when Tx is a UE, the base station is Rx, corresponding to the beam search on the base station side, assuming that the corresponding beam search on the UE side, Tx passes the beam Send pilot signal, Rx beam codebook Perform spatial scanning, beam l corresponds to N T and N R are the number of antennas of the base station and UE respectively.

[0026] Preferably, the step S2 also includes other unknown noise σ 2 The method of elimination is:

[0027] Get σ 2 An unbiased estimate of To eliminate noise interference, at the tth iteration, the arbitrary beam The sample mean of the acquired signal

[0028]

[0029] The corresponding unbiased estimate

[0030]

[0031] Since the noise of each beam is independent and identically distributed, for

[0032]

[0033] Preferably, in step S3-1, the scanning initialization method is: in the initialization stage, scanning All beams in the image are scanned twice, that is, the cumulative number of scanning cycles And the unbiased estimate of the noise variance is obtained by formula (7): The initial value of, the number of iterations t←2, the input parameter c is applied to U l (t) and ε are used as the termination conditions for the scan.

[0034] Preferably, in step S3-2, the iterative beam scanning method is: at each iteration t←t+1, Rx determines the beam l to be scanned by formula (5): ＊(t), the acquisition signal is obtained after beam scanning Update the joint matching filter according to formula (4) And update according to formula (7) Make Continuously approaching the true value, once the beam l ＊ (t) The number of symbol cycles accumulated during scanning Satisfy the termination condition

[0035]

[0036] If yes, stop scanning immediately; otherwise, repeat this step to scan again.

[0037] Preferably, in step S3-3, the beam selection method is: after stopping scanning, the beam with the most scan times is defined as Beam l ＊ The adjacent beam set is l ＊ The beam center and its adjacent beam centers are Φ(l ＊ ), based on the maximum likelihood estimation, the beam center is obtained for

[0038]

[0039] in is the beam center φ∈Φ(l ＊ ) direction vector, and select the beam through formula (9) As a result of beam alignment.

[0040] The present invention has the following characteristics and beneficial effects:

[0041] By adopting the above technical solution, the optimal beam can be quickly and accurately locked in different signal-to-noise ratio scenarios through iterative spatial scanning of the UCB strategy. After finding the optimal beam, the beam is rotated by maximum likelihood estimation to further improve the beamforming gain. This method is completely unknown to the channel information and noise variance, and does not require a preset non-adaptive scanning time. It is applicable to a wider range of millimeter wave communication scenarios, so it is more practical. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0043] Figure 1The present invention is a flowchart of an embodiment of an adaptive beam alignment method for a millimeter wave communication system based on UCB.

[0044] Figure 2 Graphs showing beam alignment performance obtained under linear uniform arrays with different numbers of antennas N in a line-of-sight scenario according to an embodiment of the present invention.

[0045] Figure 3 Graphs showing beam alignment performance obtained under linear uniform arrays with different numbers of antennas N in a non-line-of-sight scenario according to an embodiment of the present invention.

[0046] Figure 4 The figure is a beam alignment performance diagram obtained in a line-of-sight scenario under a planar uniform array with different numbers of antennas N×N according to an embodiment of the present invention.

[0047] Figure 5 The figure is a beam alignment performance diagram obtained by an embodiment of the present invention in a non-line-of-sight scenario under a planar uniform array with different numbers of antennas N×N.

[0048] Figure 6 The figure is a comparison diagram of beam alignment performance obtained by an embodiment of the present invention in a line-of-sight scenario with a linear uniform array having N antennas.

[0049] Figure 7 The figure is a comparison diagram of beam alignment performance obtained by an embodiment of the present invention in a line-of-sight scenario with a planar uniform array of N×N antennas. DETAILED DESCRIPTION

[0050] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0051] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", etc. are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", etc. may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0052] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood by specific circumstances.

[0053] The present invention provides a UCB-based millimeter wave communication system adaptive beam alignment method, such as Figure 1 As shown, the following steps are included:

[0054] S1. Construction of signal model during beam iterative scanning

[0055] Consider that the transmitter Tx sends a pilot signal and the receiver Rx performs spatial scanning. If Tx is a base station, Rx is a UE, which corresponds to the beam search on the UE side. Conversely, when Tx is a UE and the base station is Rx, which corresponds to the beam search on the base station side, assuming that the corresponding beam search on the UE side is performed, Tx uses the beam Send pilot signal, Rx beam codebook Perform spatial scanning, beam l corresponds to N T and N R are the number of antennas of the base station and UE respectively.

[0056] Iterative scanning signal model, each iteration Rx scans a beam, t represents the number of iterations. Considering the narrow bandwidth signal model, in the tth iteration, Rx scans beam l, and the corresponding acquisition signal can be expressed as

[0057]

[0058] in is complex Gaussian noise, s l (t) is the pilot signal sent by Tx in a scanning cycle, and |s l (t)| 2 =P T , P T is the transmit power of Tx; h l Equivalent channel after applying the transmit and receive beams

[0059]

[0060] in is the channel between Tx and Rx, N T and N R are the number of antennas of the base station and UE respectively, and the matched filter output of the collected signal is

[0061]

[0062] The cumulative matched filter output of beam l is [f l (1), ..., f l (m), ..., f l (n l (t))] T , where n l (t) is the cumulative number of scan cycles of Rx scan l, and its joint matched filter output is updated as

[0063]

[0064] S2. Set UCB scanning strategy to eliminate noise

[0065] In order to make the iterative scanning more efficient, that is, the scanning is concentrated on the optimal or suboptimal beam, each iteration t selects the scanning UCB weight function U l The largest beam l in (t) ＊ (t), that is

[0066]

[0067] Where c>0 is an adjustable parameter, is the estimated value of the noise standard deviation;

[0068] Because there is an unknown noise variance σ during beam scanning 2 To make the UCB strategy effective, we need to get σ 2 An unbiased estimate of To eliminate noise interference, at the tth iteration, the arbitrary beam The sample mean of the acquired signal

[0069]

[0070] The corresponding unbiased estimate

[0071]

[0072] Since the noise of each beam is independent and identically distributed, for

[0073]

[0074] S3, Adaptive beam alignment

[0075] The specific steps include the following:

[0076] S3-1, Rx performs iterative beam scanning initialization

[0077] Initialization phase, scan All beams in the image are scanned twice, that is, the cumulative number of scanning cycles And the unbiased estimate of the noise variance is obtained by formula (7): The initial value of, the number of iterations t←2, the input parameter c is applied to U l (t) and ε are used as the termination conditions for the scan.

[0078] S3-2. Iterative beam scanning based on UCB scanning strategy

[0079] At each iteration t←t+1, Rx determines the beam l to be scanned by formula (5): ＊ (t), the acquisition signal is obtained after beam scanning Update the joint matching filter according to formula (4) And update according to formula (7) Make Continuously approaching the true value, once the beam l ＊ (t) The number of symbol cycles accumulated during scanning Satisfy the termination condition

[0080]

[0081] If yes, stop scanning immediately; otherwise, repeat this step to scan again.

[0082] S3-3, beam selection

[0083] Define the beam with the most scan times after stopping scanning as Beam l ＊ The adjacent beam set is l ＊ The beam center and its adjacent beam centers are Φ(l ＊ ), based on the maximum likelihood estimation, the beam center is obtained for

[0084]

[0085] in is the beam center φ∈Φ(l ＊ ) direction vector, and select the beam through formula (9) As a result of beam alignment.

[0086] Based on the above technical solution, the following specific millimeter wave system beam alignment implementation cases are given:

[0087] Case 1

[0088] Rx is equipped with a linear uniform antenna array, and the DFT codebook is The beam where φ l is the beam center, and is the angle domain covered by the beam codebook, and the array antenna spacing is d=λ / 2.

[0089] Execute step S3 to scan the beam with the most scan times. Adjacent beam Adjacent beam center in Indicates l ＊ The beam center of , Δ represents the half beam width. According to the calculation formulas (9) and (10), the maximum likelihood beam center is obtained Select Beam As a result of beam alignment.

[0090] In this case, the Tx transmission power P is set T =1, c=1, ε=3, the simulation results are the statistical average of 5000 random channels.

[0091] Case 2

[0092] Rx equipped The planar uniform antenna array, the codebook is The beam

[0093]

[0094] where μ k and θ j are the elevation and azimuth of the beam, and μ k ,θ j ∈[-90°, 90°] is the angle domain covered by the beam codebook, is the Kronecker product, The array antenna spacing is d=λ / 2.

[0095] Execute step S3 to scan the beam with the most scan times. Adjacent beam

[0096] Corresponding adjacent beam centers Δ k and Δ j denote the half beamwidth in elevation and azimuth respectively.

[0097] Similar to case 1, fixed elevation angle μ k , maximum likelihood azimuth beam center By calculating (9) and (10), we get Maximum Likelihood Elevation Beam Center Similarly, if the maximum likelihood beam center is obtained Compare middle The strength of the joint matched filter output of two adjacent beams, Towards the stronger beam, or Get the adjacent maximum likelihood beam center Select Beam As a result of beam alignment.

[0098] like Figure 2 and Figure 3 As shown in the figure, it can be seen that the beam alignment method based on UCB provided by the present invention has a relationship between the beam alignment performance and the number of linear uniform array ULA antennas N∈{8, 16, 32, 64} under line-of-sight propagation and non-line-of-sight propagation. The horizontal axis is the signal-to-noise ratio SNR∈[-15dB, 0dB] before beamforming, and the vertical axes of (a) and (b) correspond to the relative spectrum efficiency E{R} / E{R ＊} (alignment accuracy) and pilot symbol number overhead (search time). For the selected beam The corresponding actual spectrum efficiency is The optimal beam center u is assumed to be known. ＊ =v(φ ＊ ) can achieve the ideal spectrum efficiency. As shown in the figure, the search time can be adaptively adjusted with the change of signal-to-noise ratio at different antenna numbers N, while maintaining a high alignment accuracy.

[0099] like Figure 4 and Figure 5 As shown, it can be seen that the adaptive beam alignment method for a millimeter-wave communication system based on UCB provided by the present invention has a relationship between the beam alignment performance and the number of planar uniform array UPA antennas N×N, N∈{8, 12, 16} under line-of-sight and non-line-of-sight channels. The horizontal axis and the vertical axis are the same as above, representing the alignment accuracy and the search time, respectively. As shown in the figure, the search time can automatically adapt to the change of the signal-to-noise ratio, that is, the beam search time corresponding to a low signal-to-noise ratio is long, and the beam search time corresponding to a high signal-to-noise ratio is short; in addition, even if the elevation angle and the azimuth angle need to be estimated simultaneously, the spectrum efficiency achievable by the beam alignment of the present invention remains close to the ideal spectrum efficiency.

[0100] like Figure 6 and Figure 7As shown, it can be seen that the adaptive beam alignment method for a millimeter wave communication system based on UCB provided by the present invention has a beam alignment performance in a line-of-sight propagation channel when the number of ULA antennas is N=32 and the number of UPA antennas is N×N=8×8, and the parameters are set the same as the above case. The compared algorithms refer to: an adaptive millimeter wave beam search method based on spatial scanning, referred to as ABS, wherein the threshold γ=2 is set for ULA and γ=1 for UPA. The horizontal and vertical axes are the same as above, corresponding to the alignment accuracy and search time. As shown in the figure, both inventions can realize the adaptation of the search time, but the search time required to achieve similar alignment accuracy of ABS in the present invention is further reduced, and the beam alignment is realized faster.

[0101] To sum up the simulation results, the present invention can quickly and adaptively adjust the time in millimeter wave communication scenarios with different signal-to-noise ratios, and has completely unknown channel instantaneous and statistical information and noise variance, and can obtain relatively ideal spectrum efficiency.

[0102] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions and variations of these embodiments including components are made without departing from the principles and spirit of the present invention, and still fall within the scope of protection of the present invention.

Claims

1. A UCB-based millimeter wave communication system adaptive beam alignment method, characterized in that: The steps include: S1. Construction of signal model during beam iterative scanning Iterative scanning signal model, each iteration Rx scans a beam, t represents the number of iterations, considering the narrow bandwidth signal model, then in the tth iteration, Rx scans beam l, the corresponding acquisition signal is expressed as in is complex Gaussian noise, s l (t) is the pilot signal sent by Tx in a scanning cycle, and |s l (t)| 2 =P T , P T is the transmit power of Tx; h l Equivalent channel after applying the transmit and receive beams in is the channel between Tx and Rx, N T and N R are the number of antennas of the base station and UE respectively, The matched filter output of the collected signal is The cumulative matched filter output of beam l is [f l (1), ..., f l (m), ..., f l (n l (t))] T , where n l (t) is the cumulative number of scan cycles of Rx scan l, and its joint matched filter output is updated as Signal model construction rules: Consider that the transmitter Tx sends a pilot signal and the receiver Rx performs spatial scanning. If Tx is a base station, Rx is a UE, which corresponds to the beam search on the UE side. Conversely, when Tx is a UE and the base station is Rx, which corresponds to the beam search on the base station side, assuming that the corresponding beam search on the UE side is performed, Tx uses the beam Send pilot signal, Rx beam codebook Perform spatial scanning, beam l corresponds to N T and N R are the number of antennas of the base station and UE respectively; S2. Set UCB scanning strategy to eliminate noise Each iteration t selects the scan UCB weight function U l The largest beam l in (t) ＊ (t), that is Where c>0 is an adjustable parameter, is the estimated value of the noise standard deviation; For other unknown noise σ 2 The method of elimination is: Get σ 2 An unbiased estimate of To eliminate noise interference, at the tth iteration, the arbitrary beam The sample mean of the acquired signal The corresponding unbiased estimate Since the noise of each beam is independent and identically distributed, for S3, Adaptive beam alignment S3-1, Rx performs iterative beam scanning initialization; The scanning initialization method is: in the initialization phase, scanning All beams in the image are scanned twice, that is, the cumulative number of scanning cycles And the unbiased estimate of the noise variance is obtained by formula (7): The initial value of, the number of iterations t←2, the input parameter c is applied to U l (t), ε is used as the termination condition for scanning; S3-2, iterative beam scanning based on UCB scanning strategy; The iterative beam scanning method is: in each iteration t←t+1, Rx determines the beam l to be scanned by formula (5): ＊ (t), the acquisition signal is obtained after beam scanning Update the joint matching filter according to formula (4) And update according to formula (7) Make Continuously approaching the true value, once the beam l ＊ (t) is the number of symbol cycles accumulated during scanning Satisfy the termination condition If yes, stop scanning immediately, otherwise repeat this step to scan; S3-3, beam selection The beam selection method is: define the beam with the most scan times after stopping scanning as Beam l ＊ The adjacent beam set is l ＊ The beam center and its adjacent beam centers are Φ(l ＊ ), based on the maximum likelihood estimation, the beam center is obtained for in is the beam center φ∈Φ(l ＊ ) direction vector, and select the beam through formula (9) As a result of beam alignment.

Images