A communication perception integration method based on SLAM and beam management

Abstract

The present invention discloses a communication perception integration method based on SLAM and beam management, wherein the base station and the user first use full-range beam scanning to estimate the optimal beam direction at the current moment. The user uses the angle information obtained in the beam scanning stage and the acceleration measurement output by the inertial measurement unit to execute the SLAM algorithm, obtain the position of the user and environmental features, and complete the user positioning and radio map construction at the current moment. At the next moment, the user uses the inertial measurement unit to predict the user's rough position, generates a priori beam angle search range based on the radio map constructed at the previous moment, and forms an adaptive search codebook for beam tracking. Finally, a switching module is designed to enable the system to switch between the full-range beam scanning and small-range beam tracking modules to ensure the reliability and effectiveness of communication and perception.

Description

Technical Field

[0001] The present invention belongs to the technical field of communication perception integration and wireless positioning, and in particular relates to a communication perception integration method and system based on SLAM and beam management. Background Art

[0002] As one of the emerging research directions of 6G, communication and perception integrated technology is expected to achieve communication functions such as higher data transmission rate, lower data transmission delay and larger user communication capacity, while providing perception services such as high-precision positioning, human posture recognition and environmental mapping.

[0003] Thanks to the higher frequency, wider bandwidth, larger antenna array and denser network deployment of 6G wireless communication systems, communication and perception systems are more similar in hardware architecture, channel characteristics and algorithm design. This enables the full integration of advanced technologies in the fields of communication and perception. Wireless communication systems can provide perception information and even build environmental maps while completing communication tasks. The perception function also provides useful information for wireless communication systems to enhance communication performance.

[0004] There are many problems in completing the joint design of simultaneous localization and mapping (SLAM) and beam management. First, the wireless resources that can be used for beam management are limited. It is difficult to accurately estimate the channel multipath azimuth under the limitation of limited beam pairing measurements. Secondly, the high mobility of users will cause rapid changes in the beam azimuth. It is worth considering how to effectively predict the user's motion state and further predict the beam change. Finally, the transmission ability of high-frequency electromagnetic wave transmission is poor, and its channel multipath components are easily blocked by obstacles, causing the channel multipath to be born and die, resulting in sudden changes in channel state information. How to detect these sudden changes in the beam management process and promptly avoid the negative effects caused by this phenomenon is worth studying. Summary of the invention

[0005] Purpose of the invention: The purpose of the present invention is to provide an integrated communication perception method based on SLAM and beam management to achieve user positioning and environmental perception during beam scanning, and use perception information to assist beam tracking, while ensuring the reliability of data transmission, avoiding the complexity and delay of beam scanning, and greatly saving beam scanning time.

[0006] Technical solution: In order to achieve the above purpose, the present invention adopts the following solution:

[0007] On the one hand, the present invention provides a communication perception integration system based on SLAM and beam management, which is divided into a full-range beam scanning module, a small-range beam tracking module, a SLAM module, an inertial measurement unit (IMU) module, an information conversion module and a function switching module. On the other hand, the present invention provides a communication perception integration method based on SLAM and beam management, which specifically includes the following steps:

[0008] Step 1: The base station and the user first use the full-range beam scanning module to estimate the optimal beam direction at the current moment.

[0009] Step 2: The user uses the beam angle information output by the full-range beam scanning module and the acceleration measurement value output by the IMU module to execute the SLAM module to obtain the position of the user and environmental features, thereby completing the radio map construction and user positioning at the current moment.

[0010] Step 3: At the next moment, the user uses the IMU module measurement and the SLAM module to estimate the user's position at the previous moment to predict the user's rough position at this moment. The user uses the information conversion module to generate a priori beam angle search range based on the radio map constructed at the previous moment and the position prediction at this moment.

[0011] Step 4: Use the prior beam angle search range to form an adaptive search codebook and provide it to the beam tracking module. The user uses the angle measurement obtained by the beam tracking module to continue to execute the SLAM algorithm to update and correct the user positioning and radio map construction.

[0012] Step 5: Design a function switching module so that the communication and perception integrated system switches between the full-range beam scanning and small-range beam tracking modules to ensure the reliability and effectiveness of communication and perception.

[0013] As a further technical solution of the present invention, in step 1, the full-range beam scanning scheme adopted by the base station and the user includes but is not limited to the exhaustive beam scanning scheme and the layered codebook beam scanning scheme. The full-range beam scanning process at the current moment is completed, and the beam angle information obtained includes the discrete measurement values ​​of the angle of arrival (AoA) and the angle of departure (AoD) of the multipath component, and the channel gain estimation value. Therefore, the output information of the full-range beam scanning stage is

[0014] As a further technical solution of the present invention, in step 2, the inertial measurement unit equipped by the user includes an accelerometer, a gyroscope, a magnetometer, etc. The user outputs the acceleration measurement value through a multi-sensor fusion algorithm, wherein the multi-sensor fusion algorithm includes an extended Kalman filter, a matched filter, etc. The acceleration measurement value is used to assist the state transfer module in the SLAM algorithm, and the rough position of the user is calculated as a priori information to improve the accuracy of the algorithm. At this moment, the AoA and AoD discrete measurement values ​​and the acceleration measurement value are input into the SLAM module, and SLAM considers the mirror reflection path to abstract the environmental features into a virtual base station, thereby realizing the estimation of the scatterer position, the virtual base station position and the user position. The SLAM algorithms used include but are not limited to BP-SLAM and PHD-SLAM algorithms.

[0015] As a further technical solution of the present invention, in step 3, the input of the information conversion module is the radio map measurement value of the SLAM module at the previous moment and the position prediction value of the IMU module at the current moment, and the output is the AoA and AoD prior search range at the current moment. According to the different propagation characteristics of the direct path and the indirect path, their prior angle search ranges are determined respectively. For the direct path, the angle search range of the direct path for the beam is determined according to the base station and user positions output by the SLAM module, as well as the base station and position estimation variances; for the indirect path, according to the scatterer position, virtual base station position and user position output by the SLAM module and their estimated variances, a geometric model is established to determine the angle search range of the indirect path for the beam.

[0016] As a further technical solution of the present invention, in step 4, the beam tracking module uses the angle search range information output by the information conversion module to optimize the design of the beam training codebook, including the codebook's angle domain center, codebook range, and codebook layer number, etc., to reduce the beam training overhead. The user uses the multipath AoA and AoD discrete measurement values ​​output by the beam tracking module and the acceleration values ​​output by the IMU module to perform SLAM, and merges them with the SLAM results obtained at the previous moment to update and correct the user positioning and radio map construction.

[0017] As a further technical solution of the present invention, in step 5, the function switching module is used to control the system to switch between the full-range beam scanning and the small-range beam tracking modules. The function switching module can be executed according to a fixed frame structure period, or can be flexibly implemented by a clearance judgment mechanism. The function switching module can be designed based on principles such as received signal energy, signal-to-noise ratio, and the number of multipaths obtained.

[0018] Beneficial effects: The communication perception integration method and system based on SLAM and beam management provided by the present invention has the following advantages:

[0019] 1. The existing millimeter wave beam training method is time-consuming. This scheme generates beam angle prior information through user motion status and radio map. The designed beam tracking algorithm has the advantages of flexible codebook and low overhead.

[0020] 2. Existing perception is usually achieved using additional resources. This solution achieves user positioning and environment mapping through the beam management stage, realizing the integration of communication and perception in hardware and waveform.

[0021] 3. The full-range beam scanning module covers a wide angle range but takes a long time to train, while the small-range beam tracking module consumes less time resources but has a limited angle area of ​​perception. This solution achieves flexible switching between the full-range beam scanning and small-range beam tracking modules by designing a function switching module to adapt to complex and changing wireless environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a flow chart of the present invention.

[0023] Figure 2 It is a module diagram of the present invention. DETAILED DESCRIPTION

[0024] In the following description, for the purpose of explanation rather than limitation, specific details such as specific system structures are provided to provide a thorough understanding of the present invention. However, it should be clear to those skilled in the art that the present invention may also be implemented in other embodiments without these specific details.

[0025] The scheme described in this article can be used in various communication systems, including millimeter wave, terahertz, sub-6GHz frequency bands, etc., and the working modes include time division duplex (TDD) and frequency division duplex (FDD). The base station and user end described in this article are suitable for describing any transmitting and receiving pairs.

[0026] In order to make the objectives, technical solutions and advantages of the present invention more clear, a detailed description will be given below with reference to the accompanying drawings.

[0027] Embodiment 1: The communication perception integration method based on SLAM and beam management provided by the embodiment of the present invention is as follows: Figure 1 As shown, the method includes:

[0028] Step 101: The base station and the user first use a full-range beam scanning module to estimate the optimal beam direction at the current moment.

[0029] It should be noted that full-range beam scanning can be implemented by exhaustive scanning, layered codebook scanning, etc. The base station sends a specific beam, the user end receives it through the corresponding beam, and measures the quality of the beam pair. The traditional layered codebook trains beam codebook combinations of different widths, selects the beam combination with the largest gain in each layer of the codebook, and continuously reduces the discrete angle domain of AoA and AoD corresponding to the codebook, and finally obtains the discrete measurement values ​​of AoA and AoD of the multipath component. The following takes the layered codebook as an example to introduce the training process of the l′+1th path at the current moment. It is assumed that when the beam training corresponding to this path starts, the system has obtained the channel parameter estimation values ​​of the l′ paths before the current moment s. in For AoA, For AoD, is the path gain, then the system has a (s) The residual calculation value of It can be expressed as

[0030]

[0031] For the convenience of introduction, the codebook parameter K is set to 2. When the beam training reaches the jth layer, the received energy corresponding to the four different beam combinations is As shown below

[0032]

[0033]

[0034]

[0035]

[0036] Where s is the training symbol of unit amplitude, The numbers of the beam training subsets selected by the UE and BS at the jth layer, respectively. Through the beam training at the jth layer, the system obtains the number of the beam training subset at the j+1th layer. for

[0037]

[0038]

[0039] in and are the numbers of the beam training vectors with the largest receiving energy in the current beam training subset for the UE and BS, respectively, satisfying When the training is performed up to layer j=J, the channel parameter estimation value of the l′th path at time s can be obtained: for

[0040]

[0041]

[0042]

[0043] It is worth noting that the beam training process at time s also requires the current multipath number L (s) Since the four beam combinations of the j=1th layer of the codebook cover the entire angle search domain [0,π], the article assumes that when the j=1th layer of the lth path at the sth time is trained, if its corresponding received signal satisfies The beam training process at the current moment is completed, and the complete channel estimation value at that moment is obtained as follows

[0044]

[0045] in Otherwise, continue training the current l′+1th path.

[0046] Step 102: The user executes the SLAM module using the beam angle information output by the full-range beam scanning module and the acceleration measurement value output by the IMU module to obtain the position of the user and environmental features, thereby completing the radio map construction and user positioning at the current moment.

[0047] It should be noted that the user is equipped with an IMU module and outputs the acceleration measurement value of the user in the xoy coordinate system at time s through the multi-sensor fusion algorithm. The acceleration measurement output by the IMU module at time s-1 is recorded as Then the predicted value of the user's position at time s is With speed prediction value It can be expressed as

[0048]

[0049]

[0050] in The UE position estimate of the SLAM module is used for correction, where T represents the unit time. It can be seen that the position prediction value at this time Only the estimated values ​​of the system at time s-1 and s-2 can be used to effectively reduce the cumulative error of the IMU module. At time s, the system will scan the channel measurement value of the full range beam scanning module User position prediction value with IMU module Input the SLAM module based on AoA and AoD. Based on the mirror reflection characteristics of the NLoS path, the SLAM module abstracts the environmental features into a virtual base station, where the virtual base station is the mirror image of the base station relative to the reflective surface. The positions of the base station and the virtual base station are estimated using the belief propagation and particle filter algorithm.

[0051]

[0052]

[0053] in and represent the estimated positions of the user and base station / virtual base station respectively, and and Represents a total number P of UE estimated position particles and PA / VA estimated position particles.

[0054] Step 103: At the next moment, the user uses the IMU module measurement value and the SLAM module to estimate the user's position at the previous moment to predict the user's rough position at this moment. The user uses the information conversion module to generate a priori beam angle search range based on the radio map constructed at the previous moment and the position prediction at this moment.

[0055] It should be noted that the UE position prediction value at time s using the IMU module The radio map estimate of the SLAM module at time s-1 Generate AoA, AoD prior information that is useful for beam training at time s. For the LoS path (l = 0), it is easy to know that its prior angle value is

[0056]

[0057]

[0058] in They are The prior angle value is ±π, which is designed to satisfy the range assumption of (0,π). The prior angle search range is set according to the estimated variance of the base station and virtual base station, and its mean and variance can be calculated as

[0059]

[0060]

[0061] by As the center of the circle, A circular area with a radius of can be used to cover the PA / VA particles, where C angle is a constant. angleWhen >3, the characteristics of Gaussian distribution show that if the positions of BS and UE are accurate, all PA / VA particles will be covered with a confidence level of more than 99%. Then the a priori search range of the LoS path can be calculated as

[0062]

[0063]

[0064] For the lth NLoS path (l≥1), the virtual base station location corresponds to There is a mirror reflection relationship. According to the mirror reflection geometric relationship, the AoA corresponding to the lth VA can be derived. Its prior angle value Search range with a priori angle for

[0065]

[0066]

[0067]

[0068]

[0069] Step 104: Use the prior beam angle search range to form an adaptive search codebook and provide it to the beam tracking module. The user uses the angle measurement obtained by the beam tracking module to continue to execute the SLAM algorithm, update and correct the user positioning and radio map construction.

[0070] It should be noted that, considering the design of the beam tracking codebook for the lth path, the mth beam training vector [F (j,φ) ] :,m Satisfy the equation

[0071]

[0072] in It can be seen that the code is is the center of the angle domain, and the approximate solution of the equation is

[0073]

[0074] Among them G (j,φ) is an N×2 0-1 matrix whose position in the mth column and uth row satisfies When , the element at that position is set to 1, otherwise it is set to 0. The UE can perform the same codebook design to obtain [W (j,θ) ] :,m . Secondly, the number of initial training layers for the lth path is

[0075]

[0076]

[0077] in and are the initial training layers at the BS and UE, respectively. scale With θ scale There is a difference. When the codebook on one side of the UE and the BS reaches the bottom layer first, the side will continue to participate in the training with the optimal beam training vector at the bottom layer of the codebook until the training process is completed when the codebooks on both sides reach the bottom layer. After the multipath training is completed, the output of this submodule at time s is consistent with formula (), where Therefore, the small-range beam tracking module can reduce the beam training overhead while maintaining the same angular resolution as the full-range beam scanning.

[0078] Step 105: Design a function switching module so that the communication and perception integrated system switches between the full-range beam scanning and small-range beam tracking modules to ensure the reliability and effectiveness of communication and perception.

[0079] It should be noted that at the beginning of time s, the function switching module is used to determine whether to execute the full-range beam scanning or the small-range beam tracking module at time s. After the beam tracking module is executed, the function switching module is used again to determine the validity of the beam tracking module and decide whether to execute the beam scanning module additionally at time s. The function switching module can be designed based on the channel energy. The channel energy E (s) Channel Energy Difference It can be calculated as

[0080]

[0081]

[0082] in is a constant. Since the LoS path can be Perform detection and calculate the channel energy E (s) The energy of the LoS path is considered to be a constant, so The perception of LoS path changes will be reduced, and changes in NLoS paths can be easily detected. In the function switching module, define the variable D (s) Indicates the judgment result of the switching module, expressed as

[0083]

[0084] in With E minis a constant threshold, The minimum energy threshold corresponding to the CSI mutation; E min is the minimum channel energy threshold that can meet the user's communication needs. At the beginning of time s, if D (s-1) =1, the system will select the full-range beam scanning module to implement beam management and SLAM at time s, otherwise it will execute the small-range beam tracking module. At the same time, after the beam tracking module is executed, when D (s) = 1, the system will additionally execute a full-range beam scanning module at time s to compensate for the result of the beam tracking module.

[0085] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A communication perception integration method based on SLAM and beam management, characterized in that: The integrated system includes a full-range beam scanning module, a small-range beam tracking module, a SLAM module, an inertial measurement unit IMU module, an information conversion module and a function switching module. The method includes the following steps: Step 1: The base station and the user first use the full-range beam scanning module to estimate the optimal beam direction at the current moment. Step 2: The user uses the beam angle information output by the full-range beam scanning module and the acceleration measurement value output by the IMU module to execute the SLAM module to obtain the position of the user and environmental features, thereby completing the radio map construction and user positioning at the current moment. Step 3: At the next moment, the user uses the IMU module measurement value and the SLAM module to estimate the user's position at the previous moment and predict the user's rough position at this moment. The user uses the information conversion module to generate a priori beam angle search range based on the radio map constructed at the previous moment and the position prediction at this moment. Step 4: Use the prior beam angle search range to form an adaptive search codebook and provide it to the beam tracking module. The user uses the angle measurement obtained by the beam tracking module to continue to execute the SLAM algorithm, update and correct the user positioning and radio map construction, Step 5: Design a function switching module to enable the communication and perception integrated system to switch between the full-range beam scanning and small-range beam tracking modules to ensure the reliability and effectiveness of communication and perception.

2. The communication perception integration method based on SLAM and beam management according to claim 1, characterized in that: In step 1, the full-range beam scanning scheme adopted by the base station and the user includes the beam scanning scheme and the layered codebook beam scanning scheme, and the full-range beam scanning process at the current moment is completed. The beam angle information obtained includes the discrete measurement values ​​of the arrival angle AoA and the departure angle AoD of the multipath component, as well as the channel gain estimation value. Therefore, the output information of the full-range beam scanning stage is 3. According to a communication perception integration method based on SLAM and beam management as claimed in claim 1, it is characterized in that: In step 2, the inertial measurement unit equipped by the user includes an accelerometer, a gyroscope and a magnetometer. The user outputs the acceleration measurement value through a multi-sensor fusion algorithm, wherein the multi-sensor fusion algorithm includes an extended Kalman filter and a matched filter. The acceleration measurement value is used to assist the state transfer module in the SLAM algorithm, and the rough position of the user is calculated as prior information to improve the algorithm accuracy. At the current moment, the AoA and AoD discrete measurement values ​​and the acceleration measurement value are input into the SLAM module. SLAM considers the mirror reflection path to abstract the environmental features into a virtual base station, thereby realizing the estimation of the scatterer position, the virtual base station position and the user position. The SLAM algorithms used include BP-SLAM and PHD-SLAM algorithms.

4. A communication perception integration method based on SLAM and beam management according to claim 1, characterized in that: In step 3, the input of the information conversion module is the radio map measurement value of the SLAM module at the previous moment and the position prediction value of the IMU module at the current moment, and the output is the AoA and AoD prior search ranges at the current moment. According to the different propagation characteristics of the direct path and the indirect path, their prior angle search ranges are determined respectively. For the direct path, the angle search range of the direct path for the beam is determined according to the base station and user positions output by the SLAM module, as well as the base station and position estimation variances; for the indirect path, according to the scatterer position, virtual base station position and user position and their estimated variances output by the SLAM module, a geometric model is established to determine the angle search range of the indirect path for the beam.

5. A communication perception integration method based on SLAM and beam management according to claim 1, characterized in that: In step 4, the beam tracking module uses the angle search range information output by the information conversion module to optimize the design of the beam training codebook, including the angle domain center of the codebook, the codebook range, and the number of codebook layers, to reduce the beam training overhead. The user uses the multipath AoA and AoD discrete measurement values ​​output by the beam tracking module and the acceleration values ​​output by the IMU module to perform SLAM, which is integrated with the SLAM results obtained at the previous moment to update and correct the user positioning and radio map construction.

6. The communication perception integration method based on SLAM and beam management according to claim 1, characterized in that: In step 5, the function switching module is used to control the system to switch between the full-range beam scanning and small-range beam tracking modules. The function switching module is executed according to a fixed frame structure period, or is flexibly implemented through a judgment mechanism. The function switching module is designed based on the received signal energy, signal-to-noise ratio and the number of multipaths obtained.

Images