Vehicular Network Perception-Assisted Communication Pre-Beam Alignment Method Based on Non-Line-of-Sight Link Identification

Abstract

The present invention discloses a pre-beam alignment method for vehicle-to-everything (V2X) perception-assisted communication based on non-line-of-sight (NLOS) link identification. The communication base station extracts the state information of the target object based on the received echo of the perception signal. The base station combines the perceived object state information with the predicted vehicle state information during the target tracking process to determine whether there is a line-of-sight (LOS) link between the base station and the target vehicle in the current time slot. If the base station determines that the current LOS link is blocked by an obstacle, the communication mode and the target vehicle tracking scheme are adjusted to ensure basic information transmission. If it is determined that the current link is a LOS link, the base station performs beam alignment in advance based on the predicted position of the target vehicle during the target vehicle tracking process to achieve overhead-free pre-beam alignment. The present invention reduces the beam alignment overhead through echo signal perception; at the same time, through NLOS link identification, it effectively solves the communication interruption problem caused by dynamic obstacles in the V2X communication scenario, and improves the link reliability and effective communication rate.

Description

Technical Field

[0001] The present invention belongs to the field of communication technologies, and in particular relates to a method for pre-beam alignment of vehicle-to-everything (V2X) perception-assisted communication based on non-line-of-sight link recognition. Background Art

[0002] Vehicle-to-everything (V2X) communication technology will play an important role in the next generation of autonomous driving vehicles and intelligent transportation systems, and its data transmission requires millisecond-level latency and gigabit-level transmission rates. 5G technology utilizes large-scale multiple-input multiple-output antenna arrays and millimeter-wave frequency bands and is considered a promising solution for implementing V2X communication. On the one hand, the large available bandwidth in the millimeter-wave frequency band can provide the advantage of higher data rates. On the other hand, large-scale multiple-input multiple-output antenna arrays can compensate for the high path loss imposed on millimeter-wave signals by forming beams pointing in the direction of users. To fully exploit the gain of the large-scale antenna array transmitting directional beams on millimeter-wave communication performance, it is necessary to correctly select the beams transmitted by the base station side. The existing beam selection methods mainly include the following:

[0003] (1) The beam training-based scheme, that is, the base station side traverses all beams in its own codebook before communication, and finally selects the beam with the maximum received signal-to-noise ratio of the moving terminal for communication. This method has a large training overhead, resulting in low communication efficiency. Especially when the antenna scale is huge, beam training may occupy all communication resources.

[0004] (2) The hierarchical codebook search-based scheme. Such methods usually use codewords with different resolutions for hierarchical training, and have lower overhead compared to the exhaustive codebook-based scheme. This method has a complex codebook design and does not avoid the codebook training overhead at the root.

[0005] (3) The sensor-assisted scheme. By equipping the base station side with an independent radar sensing device to sense the position information of the terminal, the overhead of beam training at the base station is greatly reduced. This method has a high cost for device configuration and maintenance.

[0006] Compared with the low mobility of terminals in traditional communication networks, the high mobility of vehicles and the uncertainty of their movement trajectories make the timeliness of beam training results poor, and the problem of beam misalignment is serious. This makes millimeter-wave vehicle-mounted communication not only require beam alignment but also require prediction of the beam direction, that is, pre-beam alignment.

[0007] To address the above-mentioned problems and technical requirements in V2X communication, due to the high range resolution brought by the large bandwidth of millimeter waves and the high angular resolution brought by large-scale antenna arrays, the communication system has high-precision sensing capabilities. Recently, the academic community has proposed a new technology with high spectral efficiency and low hardware overhead - ISAC (integrated sensing and communication) technology. Among them, the sensing and communication systems are jointly designed to share the same frequency band and hardware, which can improve spectral efficiency and reduce hardware costs. In a communication-centric ISAC system, the physical characteristics of the surrounding environment, including the target location, can be sensed through wireless signals, thereby reducing the overhead of beam training, obtaining higher performance gains than traditional beam training schemes, and improving information throughput and network efficiency. Based on this idea, recently, some researchers have proposed using DFRC (dual-functional radar and communication) signals to perform beam alignment and beam prediction on dynamic vehicles simultaneously. The echo signal of the base station's directional beam is used to estimate the state information of the target vehicle, and the Kalman filtering algorithm is used to track and predict the target state information. By sending a directional beam to the predicted position of the target, pre-beam alignment is achieved, eliminating the beam training process.

[0008] However, although the above pre-beam alignment design based on DFRC signals can greatly reduce the overhead of pre-beam alignment in millimeter-wave vehicle communication, the assumption that there is always a line-of-sight link between the base station and the target vehicle is difficult to hold. Since millimeter waves are sensitive to blockages, the impact of blockages on the communication link should be considered in real communication scenarios. This requires the base station to judge the blockage state of the current link in advance before transmitting a directional beam to communicate with the target vehicle, so as to select a more appropriate communication strategy. Existing non-line-of-sight link identification methods mainly include: 1) methods based on distance estimation; 2) methods based on channel impulse response; 3) methods based on surrounding environment context awareness. These methods usually have high implementation complexity and require the results of link identification to be fed back from the terminal to the base station, with high implementation overhead and are not suitable for high-dynamic vehicle communication scenarios that frequently require non-line-of-sight link identification.

[0009] Based on the analysis of the above pre-beam alignment scheme in vehicle communication based on integrated sensing and communication technology, it can be seen that there is an urgent need for an efficient and reliable pre-beam alignment scheme in vehicle communication that can overcome dynamic link scenarios with non-line-of-sight links. Summary of the Invention

[0010] The object of the present invention is to provide a vehicle-to-everything (V2X) perception-assisted communication pre-beam alignment method based on non-line-of-sight (NLOS) link identification. By judging the current link state before target tracking and pre-beam alignment and timely adjusting the communication mode, the communication performance of a highly mobile vehicle network in the case of NLOS links can be improved, so as to solve the technical problem that reliable pre-beam alignment cannot be carried out in a dynamic link scenario with NLOS links.

[0011] To solve the above technical problems, the specific technical solution of the present invention is as follows:

[0012] A vehicle-to-everything (V2X) perception-assisted communication pre-beam alignment method based on non-line-of-sight (NLOS) link identification, comprising the following steps:

[0013] Step 1: When the target vehicle enters the base station coverage area, the base station uses a traditional beam training method to estimate the target state information and sends a directional beam to the target vehicle to establish a communication link between the base station and the target vehicle; wherein the target state information includes distance, azimuth, speed, and signal reflection coefficient, and the directional beam is reflected back to the base station by the target vehicle to form an echo signal.

[0014] Step 2: The base station obtains the observation of the target state information from the echo signal of the directional beam. By establishing a state transition model of the moving target, the base station executes a target tracking algorithm to track and predict the target state information.

[0015] Step 3: The base station uses the target state information extracted from the echo signal for NLOS link identification; specifically: the base station compares the one-step prediction value of the target state information in the previous time slot with the estimated value of the target state information obtained from the echo signal in the current time slot to judge whether there is a line-of-sight (LOS) link between the base station and the target vehicle in the current time slot.

[0016] Step 4: The base station adjusts the link with the target vehicle according to the result of NLOS link identification; specifically: if the link in the current time slot is identified as an NLOS link, the base station adjusts the communication mode to ensure the necessary data communication with the target vehicle; if the link in the current time slot is judged to be a LOS link, there is no need to adjust the link, and the base station emits a directional beam in the direction of the predicted target vehicle position to achieve pre-beam alignment.

[0017] Step 5: The base station judges whether the target vehicle is still in the base station coverage area according to the state information. If not, the base station hands over the target vehicle to an adjacent base station and ends the communication; if so, the base station repeats steps 2 to 4 until the communication ends.

[0018] Further, in step 2, for each echo signal, a matched filtering algorithm is used to estimate the distance, speed, azimuth, and reflection coefficient of the target vehicle.

[0019] Further, in step 2, based on the state transition model of the moving target, the base station takes the observation of the target state information extracted from the echo signal as the input of the target tracking algorithm, corrects the estimation of the target state information in the current time slot, and gives the predicted value of the target state information in the next time slot.

[0020] Further, in step 3, a hypothesis testing method is used to determine whether there is a line-of-sight link between the base station and the target vehicle in the current time slot; specifically, if there is a line-of-sight link in the current time slot, the difference between the predicted value and the estimated value of the target vehicle state is considered to be a Gaussian white noise variable, and if the current time slot is a non-line-of-sight link, since the echo signal is reflected from an obstacle rather than the target vehicle, the difference between the predicted value and the estimated value of the state is not a Gaussian white noise variable.

[0021] Further, in step 4, if the current link is determined to be a non-line-of-sight link, the base station adjusts the communication mode to ensure the necessary data communication with the target vehicle, such as basic safety message (BSM); at the same time, to avoid the deviation brought by the non-line-of-sight link state to the target state information tracking, the target tracking algorithm predicts the target state information in the next time slot while skipping the estimated value of the target state information in the current time slot.

[0022] Further, before the base station uses the target tracking algorithm to correct the estimation of the target state information in the current time slot and gives the predicted value of the target state information in the next time slot, it further includes:

[0023] The state transition noise covariance matrix Q corresponding to the state transition model of the moving target s , the observation noise covariance matrix Q for extracting target state information from the echo signal m .

[0024] Further, before the base station uses the hypothesis testing method to determine the current link state, it further includes:

[0025] The probability distribution of the difference in state information contained in the echo signals of the obstacle and the target vehicle: the probability density function p(ρ n b ) of the reflection coefficient offset, the probability density function p(d n b ) of the distance offset, the probability density function p(v n b ) of the speed offset; the threshold γ of the hypothesis testing Bayesian decision.

[0026] A vehicle-to-everything (V2X) perception-assisted communication pre-beam alignment method based on non-line-of-sight link identification according to the present invention has the following advantages:

[0027] (1) This method uses the content of step 2, can make full use of the sensing ability of millimeter-wave signals, and the target state information extracted from the echo signal reflects the actual motion state of the target, providing the best beam emission direction for the base station, thus greatly reducing the overhead of beam alignment.

[0028] (2) This method uses the content of step 3, can make full use of the target state information provided by the echo signal and the vehicle tracking algorithm to identify non-line-of-sight links, has no uplink feedback overhead, low complexity, and is suitable for high-dynamic vehicle communication scenarios.

[0029] (3) This method uses the content of step 4, can adaptively adjust the communication link and vehicle tracking algorithm according to the non-line-of-sight link identification result, and improves the robustness and reliability of vehicle communication in an environment with obstacles. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 is a schematic diagram of vehicle communication in an urban environment provided by an embodiment of the present invention;

[0031] Figure 2 is a schematic diagram of a vehicle motion state transition model provided by an embodiment of the present invention;

[0032] Figure 3 is a schematic diagram of the process of a vehicle-to-everything (V2X) perception-assisted communication pre-beam alignment method based on non-line-of-sight link identification provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0033] In order to better understand the purpose, structure and function of the present invention, the following further describes in detail a vehicle-to-everything (V2X) perception-assisted communication pre-beam alignment method based on non-line-of-sight link identification of the present invention with reference to the accompanying drawings.

[0034] Those skilled in the art of the present technology can understand that, unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meaning as the general understanding of those of ordinary skill in the art to which the present invention belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted with an idealized or overly formal meaning unless defined as here.

[0035] A vehicle - to - everything (V2X) perception - assisted communication pre - beam alignment method based on non - line - of - sight (NLOS) link identification provided by an embodiment of the present invention. The base station determines the current link state based on the echo signal of the directional beam, enabling the base station to adjust the communication strategy in a timely manner according to the link state, thereby avoiding link interruption in the NLOS link state and achieving reliable communication in a vehicle - borne network with obstacles. At the same time, a target tracking algorithm designed based on the perceived target state information can give a prediction of the future state information of the target vehicle, thus assisting the base station in selecting the transmitting beam and greatly reducing the pre - beam alignment overhead in a highly dynamic V2X network scenario.

[0036] In one embodiment, referring to Figure 3 the flow schematic diagram shown, a V2X perception - assisted communication pre - beam alignment method based on NLOS link identification proposed by the present invention includes the following main steps:

[0037] a) When the target vehicle enters the coverage area of the base station, the base station uses a traditional beam training method to estimate the target state information and sends a directional beam to the target vehicle to establish a communication link between the base station and the target vehicle; where the target state information includes distance, azimuth, speed, and signal reflection coefficient, and the directional beam is reflected back to the base station by the target vehicle to form an echo signal;

[0038] b) The base station obtains the observation of the target state information from the echo signal of the directional beam. Based on the moving target state transition model, the base station executes a target tracking algorithm to track and predict the target state information;

[0039] c) The base station uses the target state information extracted from the echo signal for NLOS link identification. Specifically: The base station compares the one - step prediction value of the target state information in the previous time slot with the estimated value of the target state information obtained from the echo signal in the current time slot, and uses a hypothesis - testing method to determine whether there is a line - of - sight (LOS) link between the base station and the target vehicle in the current time slot;

[0040] d) The base station adjusts the link with the target vehicle according to the result of the NLOS link identification. Specifically: If the link in the current time slot is identified as an NLOS link, the base station adjusts the communication mode to ensure the necessary data communication with the target vehicle, such as basic safety message (BSM). If the link in the current time slot is judged to be an LOS link, there is no need to adjust the link, and the base station transmits a directional beam in the direction of the predicted target vehicle position to achieve pre - beam alignment;

[0041] e) The base station determines whether the target vehicle is still in the coverage area of the base station according to the state information. If not, the base station hands over the target vehicle to an adjacent base station and ends the communication; if so, the base station repeats steps b) - d) until the communication ends.

[0042] Wherein,

[0043] This pre-beam alignment method uses echo signals for target state perception, making full use of the high range resolution brought by the large bandwidth of millimeter-wave signals and the high angular resolution brought by large-scale antennas. Based on the reflected echo signals of objects, the non-line-of-sight link identification method used determines whether the current echo signal comes from the target vehicle, and adjusts the communication link accordingly.

[0044] In step a), when the target enters the coverage area of the base station, the optimal beam can be obtained through beam scanning, and then the initial motion state of the object can be perceived by transmitting the directional beam and receiving the echo signal, including the distance, azimuth angle, speed, and signal transmission coefficient of the target.

[0045] In step b), when the base station receives the echo signal, the radar matched filtering is performed on the echo signal using the transmitted signal with time delay and Doppler shift to obtain the time delay and Doppler frequency of the target vehicle. At the same time, the obtained time delay and Doppler frequency are used to perform pulse compression on the echo signal, and the observation vector about the azimuth angle and reflection coefficient of the target can be obtained. Establish a target vehicle state transition model, take the observations of time delay, Doppler frequency, azimuth angle, and reflection coefficient extracted from the echo signal as inputs, and use the low-complexity extended Kalman filtering method to track and predict the target state, so as to achieve pre-beam alignment of the base station to the target vehicle.

[0046] In step c), when performing non-line-of-sight link identification, it can be achieved by using the binary hypothesis testing method to utilize the difference between the estimated value of the echo signal state information and the predicted value of the target tracking state information. Specifically, in a specific vehicle communication scenario, since the echo signal may come from an obstacle or the target vehicle, which respectively correspond to the non-line-of-sight link state and the line-of-sight link state. The difference in the state information contained in the echo signals of the obstacle and the target vehicle can be modeled as a random variable, and its probability density function can be determined by on-site measurement and data fitting methods.

[0047] In step d), when the current link is a non-line-of-sight link and link adjustment is required, the base station can activate the Sub-6G communication mode and communicate with the target vehicle using the rich multipath links in the low-frequency band to avoid link interruption problems in the non-line-of-sight link situation. At the same time, since the echo signal is reflected from an obstacle in the non-line-of-sight link state, the obtained state information cannot reflect the true target motion state, so the state information observations extracted from the echo signal will not be used as the input for the Kalman filtering in the next time slot.

[0048] The method is not only applicable to single-vehicle communication scenarios, but also to multi-vehicle communication scenarios. For multi-vehicle communication scenarios, communication and perception of multiple vehicles can be achieved simultaneously through digital beamforming methods.

[0049] Figure 1 It is a schematic diagram of a scenario of a pre-beam alignment method for vehicle-to-everything (V2X) perception-assisted communication based on non-line-of-sight (NLOS) link identification shown according to an exemplary embodiment. t represents time. It can be seen that the line-of-sight (LOS) link between the base station and the target vehicle will be blocked by obstacles at t = t2 and t = t3, resulting in link interruption.

[0050] Figure 2 It is an example diagram of a target vehicle state transition model shown according to an exemplary embodiment. Among them, the target vehicle is driving on a straight road, and the antenna array of the base station is parallel to the road.

[0051] Considering a single time slot with a duration of ΔT, the antennas at the base station side are a uniform linear array, and N t represents the number of transmitting antennas at the base station side, and N r is used to represent the number of receiving antennas at the base station side. p n is the signal transmission power in the nth time slot, μ n is the Doppler frequency in the nth time slot, τ n is the echo delay in the nth time slot, θ n is the target azimuth angle in the nth time slot, s n (t) is the downlink signal transmitted by the base station in the nth time slot. Then, the echo signal received by the base station in the nth time slot is expressed as

[0052]

[0053] where nΔT ≤ t < (n + 1)ΔT, is the conjugate transpose of a t (θ n ), β n is the signal reflection coefficient in the nth time slot, a t is the transmit steering vector, a r is the receive response vector, f n is the beamforming vector in the nth time slot, z n (t) represents the additive white Gaussian noise in the nth time slot. According to the basic equation of radar, the numerical value of the reflection coefficient β n marked in dB (decibels) is ρ n , specifically

[0054]

[0055] where, is the wavelength of the millimeter-wave signal, f0 is the center frequency of the millimeter-wave signal, c is the speed of light propagation, is the distance between the base station and the vehicle in the nth time slot, σ nis the radar cross-section of the target in the nth time slot. When the echo signal of the directional beam is reflected from an obstacle rather than the target vehicle, since the obstacle and the target vehicle may correspond to different reflection coefficients, distances (corresponding to time delays), and speeds (corresponding to Doppler frequencies), this property can be used for non-line-of-sight link identification. Specifically, the vector defined for non-line-of-sight link identification is s n =[ρ n ,τ n ,μ n T , where [·] T denotes the transpose of the vector, v n is the running speed of the target in the nth time slot.

[0056] For the vehicle state transition model, such as Figure 2 , the state vector of the target vehicle in the nth time slot is defined as x n =[d n ,v n ,θ n ,ρ n T . According to Figure 2 the vehicle state transition model described in, it can be considered that the running speed of the vehicle remains unchanged within a single time slot ΔT (usually a few tens of milliseconds). Then, through simple geometric analysis, the vehicle state transition model of the target vehicle can be obtained as

[0057]

[0058] where, w d ,w v ,w θ ,w ρ is the additive Gaussian state transition noise. The above model can be expressed in vector form as x n =g(x n-1 )+w n , where g(·) is the mapping function representing the target state transition, w=[w d ,w v ,w θ ,w ρ T is the state transition noise vector, and its covariance matrix is Q s .

[0059] In addition, at the nth time slot, the base station can extract the observation of the target state information from the echo signal. Specifically, for the received echo signal r n (t), the estimates of the time delay and Doppler frequency are obtained using the matched filtering method as:

[0060] ​​​

[0061] where z τ , z μ is the parameter estimation noise, which follows a Gaussian distribution with a mean of 0. Based on the estimation of the time delay and Doppler frequency, the observation vector for the target azimuth angle and reflection coefficient can be obtained as:

[0062]

[0063] where s n *(·) is the conjugate signal of s n (·), φ is a constant value within the base station coverage area, G is the matched filtering gain, and z r is Gaussian white noise. Further, the target azimuth angle θ n can be estimated by the maximum likelihood method, expressed as Then the estimated value for the reflection coefficient is

[0064]

[0065] where z ρ is the parameter estimation noise, which follows a Gaussian distribution with a mean of 0.

[0066] Based on the above description and definitions, the specific implementation steps of the exemplary embodiments of the proposed method can be summarized as follows:

[0067] (1) Initial access phase. The base station uses the beam scanning method to obtain the initial state information of the target vehicle x0 = [d0, v0, θ0, ρ0] T , and sends a directional beam in the direction of the initial position to establish a communication link.

[0068] (2) Target tracking and state prediction phase. In the nth time slot, the base station extracts information from the echo signal to obtain the observation vector of the target state information The relationship between it and x n can be vectorized as y n = h(x n ) + z n , where h(·) represents the mapping function from the target state information to the observation vector, is the observation noise vector, and its covariance matrix is Q m . The base station corrects the estimation of the target state information in the current time slot based on the observation vector y n and the target vehicle state transition model x n = g(x n-1 ) + w n , and predicts the future position of the target vehicle based on the link state recognition result. Specifically, the process of vehicle tracking and state prediction based on the extended Kalman filter is divided into the following six steps:

[0069] a) Prediction of target state information: where is the estimated value of the target state information at the (n - 1)-th time slot, is the predicted value of the target state information at the current time slot at the (n - 1)-th time slot.

[0070] b) Linearization of the state observation equation:

[0071] c) Prediction of the error covariance matrix: where M 0∣0 needs to be initialized and can generally be set as a diagonal matrix of the same order of magnitude as the actual state estimation variance.

[0072] d) Calculation of the Kalman filter gain:

[0073] e) Update of the target state information:

[0074]

[0075] where b n is the result of the non-line-of-sight link identification at the current time slot, and b n = 1 indicates that the current link is in the non-line-of-sight link state. It can be seen that if the current link is a non-line-of-sight link, then the echo signal's state observation vector y n of the target will not be used for the update of the target state information. The process of using binary hypothesis testing for non-line-of-sight link state identification will be specifically described in step (3).

[0076] f) Update of the error covariance matrix: M n∣n = (I - K n H n )M n∣n-1 . where I is the identity matrix.

[0077] (3) Non-line-of-sight link identification phase. At the n-th time slot, given the predicted value of the base station for the current time slot state information at the (n - 1)-th time slot we can obtain Given the estimated value of the target state based on the echo signal at the current time slot, we can obtain Define The binary test problem is expressed as:

[0078]

[0079] where the hypothesis indicates that the link between the target and the base station is a line-of-sight link, and the hypothesis indicates that the link between the target and the base station is a non-line-of-sight link. is the difference between the predicted value and the estimated value of s under the line-of-sight link state, which is Gaussian white noise. Define n as the parameter deviation factor caused by obstacle blockage, where Using the Bayesian method of hypothesis testing with unknown parameters, this problem can

[0080] be reduced to the decision on the decision maker L(·):

[0081] where γ is the threshold of the Bayesian decision of hypothesis testing. Given the probability density distribution p(b

[0082]

[0083] (including n ) and the decision threshold γ, this binary hypothesis testing problem can be solved. (4) Link adjustment phase. If the link state in the nth time slot is judged as b

[0084] = 1, that is, the non-line-of-sight link state, in order to ensure reliable communication with the vehicle, the base station can enable the Sub-6G communication module to communicate with the target vehicle using low-frequency communication signals to achieve necessary data transmission; if the link state in the nth time slot is judged as b n = 0, that is, the line-of-sight link state, the base station turns off the Sub-6G communication module. At the same time, based on the position prediction of the vehicle in the next time slot, a directional beam is transmitted, that is, the transmit beamforming vector n where is the predicted value of the target azimuth angle in the current time slot at the (n - 1)th time slot. (5) Link handover phase. The base station judges whether the target vehicle is still in the coverage area of the base station according to the target state information. If not, the base station hands over the target vehicle to the adjacent base station and ends the communication; if so, the base station repeats steps (2)-(4) until the communication ends.

[0085] The above pre-beam alignment method uses the echo signal for target state perception and target tracking, making full use of the high range resolution brought by the large bandwidth of the millimeter wave signal and the high angle resolution brought by the large-scale antenna (step (2)). On the other hand, the non-line-of-sight link identification method used in the present invention is only based on the state information extracted from the echo signal and the state prediction in the target tracking process (step (4)), and its low overhead is suitable for high-dynamic vehicle communication scenarios.

[0086]

[0087] ​In summary, the present invention greatly reduces the beam alignment overhead through echo signal sensing; at the same time, through non-line-of-sight link identification, it effectively solves the communication interruption problem caused by dynamic obstacles in the vehicle-to-everything (V2X) communication scenario, and greatly improves the link reliability and effective communication rate.

[0088] It can be understood that the present invention is described through some embodiments. Those skilled in the art know that various changes or equivalent replacements can be made to these features and embodiments without departing from the spirit and scope of the present invention. Additionally, under the teaching of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application belong to the scope protected by the present invention.

Claims

1. A pre-beam alignment method for vehicle-to-everything (V2X) perception-assisted communication based on non-line-of-sight (NLOS) link identification, characterized in that It includes the following steps: Step 1: When the target vehicle enters the base station coverage area, the base station uses the traditional beam training method to estimate the target state information and sends a directional beam to the target vehicle to establish a communication link between the base station and the target vehicle; The target state information includes distance, azimuth, speed, and signal reflection coefficient, and the directional beam is reflected back to the base station by the target vehicle to form an echo signal; Step 2: The base station obtains the observation of the target state information from the echo signal of the directional beam. By establishing a state transition model of the moving target, the base station executes a target tracking algorithm to track and predict the target state information; Step 3: The base station uses the target state information extracted from the echo signal to identify the non-line-of-sight link; Specifically: The base station compares the one-step predicted value of the target state information in the previous time slot with the estimated value of the target state information obtained from the echo signal in the current time slot to determine whether there is a line-of-sight link between the base station and the target vehicle in the current time slot; In Step 3, the hypothesis testing method is used to determine whether there is a line-of-sight link between the base station and the target vehicle in the current time slot; Specifically, if there is a line-of-sight link in the current time slot, the difference between the target vehicle state prediction value and the estimated value is considered a Gaussian white noise variable. If the current time slot is a non-line-of-sight link, since the echo signal is reflected from an obstacle rather than the target vehicle, the difference between the state prediction value and the estimated value is not a Gaussian white noise variable; Step 4: The base station adjusts the link with the target vehicle according to the result of the non-line-of-sight link identification; specifically: if the link in the current time slot is identified as a non-line-of-sight link, the base station adjusts the communication mode to ensure the necessary data communication with the target vehicle; if the link in the current time slot is determined to be a line-of-sight link, there is no need to adjust the link, and the base station emits a directional beam in the direction of the predicted target vehicle position to achieve pre-beam alignment; Step 5: The base station determines whether the target vehicle is still in the base station coverage area according to the state information. If not, the base station hands over the target vehicle to an adjacent base station and ends the communication; If so, the base station repeats Step 2 to Step 4 until the communication ends; Before the base station uses the hypothesis testing method to determine the current link state, it also includes: Probability distribution of the difference in state information contained in the echo signals of the obstacle and the target vehicle: Probability density function of reflection coefficient offset , Probability density function of distance offset , Probability density function of velocity offset ; Threshold for hypothesis testing Bayesian decision .

2. The vehicle network perception-assisted communication pre-beam alignment method based on non-line-of-sight link recognition according to claim 1, characterized in that In Step 2, for each echo signal, the matching filtering algorithm is used to estimate the distance, speed, azimuth, and reflection coefficient of the target vehicle.

3. The vehicle network perception-assisted communication pre-beam alignment method based on non-line-of-sight link recognition according to claim 1, characterized in that In Step 2, based on the state transition model of the moving target, the base station takes the observation of the target state information extracted from the echo signal as the input of the target tracking algorithm, corrects the estimation of the target state information in the current time slot, and gives the predicted value of the target state information in the next time slot.

4. The vehicle network perception-assisted communication pre-beam alignment method based on non-line-of-sight link recognition according to claim 1, characterized in that In Step 4, if the current link is determined to be a non-line-of-sight link, the base station adjusts the communication mode to ensure the necessary data communication with the target vehicle. The necessary data includes basic safety information BSM; at the same time, to avoid the deviation brought by the non-line-of-sight link state to the target state information tracking, the target tracking algorithm predicts the target state information in the next time slot while skipping the estimated value of the target state information in the current time slot.

5. The method for pre-beam alignment of vehicle network perception-assisted communication based on non-line-of-sight link identification according to claim 3, wherein Before the base station uses the target tracking algorithm to correct the estimation of the target state information in the current time slot and gives the predicted value of the target state information in the next time slot, it also includes: The state transition noise covariance matrix corresponding to the state transition model of the moving target , and the observation noise covariance matrix for extracting the target state information from the echo signal .

Images