A method for selecting access points for synaesthesia fusion based on Cramer-Rao bound

Abstract

The present invention discloses a method for selecting synaesthesia fusion access points based on Cramer-Rao bounds. This method aims at the situation where the functional allocation of access points with communication and perception functions in a scene does not take into account the impact on the positioning accuracy performance of the perception function. By introducing Cramer-Rao performance limit analysis and using the positioning accuracy heat map verification method to verify the rationality of the selection of access points with perception functions, higher perception positioning accuracy is obtained while ensuring communication performance, so as to solve the problem that most existing technologies do not consider the impact of synaesthesia integrated synaesthesia fusion access point selection on perception performance. The results of the present invention realize a method for selecting synaesthesia fusion access points in the field of mobile communications under non-cellular large-scale MIMO, which effectively improves the perception capability of the access points and promotes the rational allocation of communication resources while ensuring the communication effect.

Description

Technical Field

[0001] The invention relates to a synaesthesia fusion access point selection method based on Cramer-Rao bounds, belonging to the technical field of mobile communications. Background Art

[0002] As mobile communications technology advances to the commercialization of 5G and the ongoing development of 6G, researchers are increasingly studying millimeter-wave and terahertz communications. The shorter wavelengths and higher frequencies of millimeter-wave communications enable finer spatial resolution and higher achievable communication rates. The development of massive MIMO antenna array technology has further enabled finer spatial resolution and beamforming techniques, laying the foundation for spatial multiplexing in mobile communications.

[0003] 4G and 5G mobile communication technologies currently mainly use single antennas or a small number of antenna arrays, and communication technology based on cellular cells. This technology has limited mobile communication speed and energy efficiency, and it is difficult to achieve accurate spatial location perception of mobile users, making it difficult to use the spatial resolution of millimeter wave technology to improve communication efficiency and communication resource allocation.

[0004] The cell-free massive MIMO architecture is a newly proposed communication architecture. In this architecture, rather than using a single AP to divide the service area (as in the cell architecture) to serve users, multiple APs work together to serve users. This technology surpasses the performance limitations of single-AP technology, significantly improving communication rates and spectrum efficiency. Synaesthesia fusion technology can assist in communication based on acquired positioning information, enhancing the spatial resolution of APs and further improving spectrum efficiency and energy utilization. Summary of the Invention

[0005] Technical problem: Aiming at improving the perception capability of mobile communications under a non-cellular massive MIMO architecture, as well as the communication spectrum efficiency and the rational allocation of resources, the present invention proposes a synaesthesia fusion access point selection method that uses synaesthesia fusion technology to assist communication based on positioning information.

[0006] Technical solution: To achieve the above-mentioned purpose, the present invention provides a method for selecting a synaesthesia fusion access point based on the Cramer-Rao bound, comprising the following steps:

[0007] Step S1, simulation preparation, initializing the three-dimensional position distribution of access points AP in the simulation;

[0008] Step S2: During uplink communication, all APs in the non-cellular network provide services to the entire user group, which includes all users using the same orthogonal pilot. Other user groups using different orthogonal pilots can be considered as non-interference users. Each user group can be considered independently in the service scenario. The communication process can be divided into multiple time slots, each of which contains a pilot signal as well as uplink and downlink data signals. One time slot is equivalent to a complete communication process, including positioning, pilot estimation, uplink communication, and downlink communication. After receiving the pilot signal, the AP constructs a signal model as a mathematical expression of the uplink pilot signal.

[0009] In step S3, in a non-cellular network, the massive MIMO antenna arrays on the AP side communicate with each other and share a common control center CPU. The signals received by all APs in the same time slot in step S1 are modeled as noise-free pilot signal models, and then summed up to form a total noise-free uplink pilot signal expression.

[0010] Step S4, converting the received total uplink pilot expression into a CRLB Cramer-Rao bounded Fisher information matrix. The key information of the information matrix can be regarded as the pilot signal of a certain user received by a certain AP. The Cramer-Rao bounded information matrix contains the Fisher information of the pilot signal of each user received by each AP, which is distributed in the entire information matrix.

[0011] Step S5: The Fisher information matrix of each AP's Cramer-Rao bound is integrated to obtain the perception performance bound of the AP group composed of multiple APs. The perception performance bound without a cellular network is calculated based on the perception performance bound according to the formula to achieve the highest positioning accuracy.

[0012] In step S6, after obtaining the perception performance limit, a perception accuracy heat map is drawn based on the performance limit. The heat map describes the positioning accuracy limit that can be achieved for each user on the map based on the distribution of APs in a three-dimensional plane. The positioning performance of the AP group for users in different locations in a non-cellular network architecture is understood through the heat map. By changing the method for selecting synaesthesia fusion access points in this method, perception accuracy heat maps under different synaesthesia fusion access point selection distributions are obtained. By applying mean analysis, variance analysis, or sampling point analysis to the heat map, the synaesthesia fusion access point selection distribution that best suits the perception accuracy is found, thereby achieving the purpose of maximizing the AP's positioning accuracy for users.

[0013] in,

[0014] The mathematical expression of the uplink pilot signal in step S2 is:

[0015]

[0016]

[0017] In formulas (1) and (2), superscript (·) r Indicates received signal, superscript (·) p Indicates that the received signal is a pilot signal, the subscript (·) n represents the nth AP, subscript (·) k represents the kth user, subscript (·) l represents the lth path (the same superscripts and subscripts have the same meaning in the following text), t represents the tth moment, Represents the signal received at the AP, represents the pilot signal sent by the user, τ n,k,l Indicates the transmission delay, represents the pilot transmission power, is the antenna array used by the user to send pilot signals, represents additive Gaussian noise, Represents the root length of the pilot symbol, H n,k,l represents the channel information, α n,k,l is the large-scale fading factor, which describes the power decay during signal transmission; is the signal receiving deflection angle matrix of the MIMO antenna array at the AP end, with the subscript (·) r represents the receiving AP, where Represents the phase shift of the antenna receiving signal in the horizontal direction, Represents the longitudinal phase shift of the antenna received signal.

[0018] The noise-free uplink pilot signal model in step S3 is:

[0019] The kth user pilot signal received by the nth AP is:

[0020]

[0021] Formula (3) is the noise-free case of formula (2),

[0022] The total noise-free uplink pilot signal expression is:

[0023]

[0024] In formula (4), superscript (·) p Indicates that the received signal is a pilot signal, K indicates that there are K users, and N indicates that there are N APs.

[0025] The step S4 specifically includes:

[0026] Step S401: Calculate the Cramer-Rao bounded Fisher information matrix based on the noise-free pilot signal model obtained in step S3:

[0027]

[0028] In formula (5), is the specific form of the Fisher information matrix of the Cramer-Rao bound, ηn,k,l represents the complex field vector, The included Fisher information is given by the following formula:

[0029]

[0030] In formula (6), superscript (·) r represents the received signal, τ n,k,l Indicates the transmission delay, represents the real part of the large-scale fading factor, represents the imaginary part of the large-scale fading factor, with the superscript (·) * Indicates conjugation, superscript (·) H represents conjugate transpose, l' represents signals of the same or different paths, l=l' represents signals of the same path, l≠l' represents signals of different paths, Indicates taking the real part of a complex number, α l is the large-scale fading factor, They are expressed by equations (7) and (8) respectively:

[0031] γ p =N rx N tx N p / P n (7)

[0032] In formula (7), N rx is the number of receiving antennas, N tx is the number of transmitting antennas, N p is the pilot symbol length, P n is the transmitted pilot power;

[0033]

[0034] In formula (8), |P(f)| 2 is the probability distribution function of power, B is the bandwidth;

[0035] Step S402: Obtain the error performance bound of the pilot signal parameters of the Cramer-Rao bounded Fisher information matrix of a single AP using the calculated Fisher information matrix:

[0036]

[0037] In formula (9), superscript express The estimator of , the following formulas have the same meaning;

[0038]

[0039]

[0040] In formula (11), B eff Indicates bandwidth,

[0041]

[0042] The step S5 specifically includes:

[0043] In step S501, the perception performance limit is calculated using the Fisher information matrix of the Cramer-Rao bound obtained in step S4. The range domain positioning accuracy limit and the angle domain positioning accuracy limit are derived from formulas (13a) and (13b), respectively:

[0044]

[0045]

[0046] The subscript (·) in formula (13) sp Indicates the positioning accuracy limit in the range domain, subscript (·) so Indicates the positioning accuracy limit in the angle domain, with a superscript (·) -1 represents the inverse of the matrix, where p k represents the Cartesian coordinate vector of the user, Represents the phase shift of the antenna receiving signal in the horizontal direction, represents the longitudinal phase shift of the antenna receiving signal, It is expressed by formula (14):

[0047]

[0048] The superscript (·) in formula (14) T represents the matrix transpose, Expressed as formula (15a), T n,k Expressed as formula (15b):

[0049]

[0050]

[0051] In formula (15), only the LOS diameter is considered for the convenience of calculation, that is, where l=1, They are expressed as formula (16a) and (16b) respectively, Expressed as formula (17a-17d):

[0052]

[0053]

[0054] In formula (16)

[0055]

[0056]

[0057]

[0058]

[0059] The elements in formula (17) are all the calculation results of formula (6).

[0060] The mean analysis described in step S6 is to take the mean of the entire heat map, and the result obtained is the average positioning accuracy information, which represents the average positioning accuracy of the AP group in the area; the variance analysis described in step S6 is to take the mean square deviation of the entire heat map, and the result obtained is the degree of fluctuation of the positioning accuracy, which represents the degree of fluctuation of the positioning accuracy of the AP group in the area; the sampling point analysis described in step S6 is to take the mean of the key points in the heat map, and the result obtained is the average positioning accuracy of the key points, which represents the positioning capability at the key sampling points.

[0061] Beneficial effects: The synaesthesia fusion access point selection method based on positioning information-assisted communication provided by the present invention can improve the spatial resolution of the AP, thereby further improving spectrum efficiency and energy utilization efficiency.

[0062] The synaesthesia fusion access point selection method for assisting communication based on positioning information of the present invention can make communication APs with perception functions more reasonably distributed in positions, and determine the spatial position of users more accurately on the basis of ensuring communication. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 A diagram showing the steps of the synaesthesia fusion access point selection method of the present invention;

[0064] Figure 2This is a schematic diagram of the AP placement for the non-cellular architecture of the present invention, where BS1-BS6 are the placement positions of the six APs, and MS1-MS8 are the set perception accuracy sampling points. The sampling results of the target points are used to compare the performance boundaries of perception accuracy under different AP placement conditions.

[0065] Figure 3 The thermal map images of perception accuracy are drawn according to the performance limit of the present invention. The upper one is the angle domain perception accuracy thermal map, and the lower one is the distance domain perception accuracy thermal map. DETAILED DESCRIPTION

[0066] To further understand the content of the present invention, a method for selecting a synaesthesia fusion access point based on the Cramer-Rao bound of the present invention is described in detail with reference to the accompanying drawings and specific embodiments.

[0067] like Figure 1 As shown, the present invention adopts the synaesthesia fusion technology to assist communication according to the positioning information. The steps of the synaesthesia fusion access point selection method are as follows:

[0068] Step 1: Simulation preparation, initialize the three-dimensional position distribution of AP in the simulation, such as Figure 2 As shown in the figure, BS1-BS6 are the locations of the APs, and their three-dimensional coordinates are (0, 0, 10), (0, 100, 10), (0, 200, 10), (50, 0, 10), (50, 100, 10), and (50, 200, 10), respectively, in meters. The coordinates of the positioning accuracy sampling points are (10, 12, 1), (65, 12, 1), (122, 12, 1), (180, 12, 1), (20, 38, 1), (77, 38, 1), (135, 38, 1), and (190, 38, 1), in meters.

[0069] In step 2, during uplink communications, all APs in the non-cellular network provide service to the entire user group, which includes all users using the same orthogonal pilot. Other user groups using different orthogonal pilots are considered non-interference users, and each user group can be considered independently within the service scenario. The communication process can be divided into multiple time slots, each of which contains a pilot signal as well as uplink and downlink data signals. A time slot is equivalent to a complete communication process, including positioning, pilot estimation, uplink communication, and downlink communication. After receiving the pilot signal, the AP constructs a signal model as a mathematical expression for the uplink pilot.

[0070]

[0071]

[0072] In formulas (1) and (2), superscript (·)r Indicates received signal, superscript (·) p Indicates that the received signal is a pilot signal, the subscript (·) n represents the nth AP, subscript (·) k represents the kth user, subscript (·) l represents the lth path (the same superscripts and subscripts have the same meaning in the following text), t represents the tth moment, Represents the signal received at the AP, represents the pilot signal sent by the user, τ n,k,l Indicates the transmission delay, represents the pilot transmission power, is the antenna array used by the user to send pilot signals, represents additive Gaussian noise, Represents the root length of the pilot symbol, H n,k,l represents the channel information, α n,k,l is the large-scale fading factor, which describes the power decay during signal transmission; is the signal receiving deflection angle matrix of the MIMO antenna array at the AP end, with the subscript (·) r represents the receiving AP, where Represents the phase shift of the antenna receiving signal in the horizontal direction, Represents the longitudinal phase shift of the antenna received signal.

[0073] In step 3, in a non-cellular network, the massive MIMO antenna arrays on the AP side communicate with each other. They have a common control center CPU, which can add the signals received by all APs in the same time slot in step 1 and model them as the total uplink pilot expression.

[0074] The nth user pilot signal received by the kth AP (noise-free):

[0075]

[0076] Formula (3) is the noise-free case of formula (2).

[0077] The total noise-free uplink pilot signal expression is:

[0078]

[0079] Step 4: Convert the received total uplink pilot expression into a CRLB Cramer-Rao bounded Fisher information matrix. The key information of the information matrix can be regarded as the pilot signal of a certain user received by a certain AP. The Cramer-Rao bounded Fisher information matrix contains the pilot signal information of each user received by each AP, which is distributed in the entire Fisher information matrix.

[0080] The calculation of Fisher information matrix is ​​obtained by the following formula:

[0081]

[0082] In formula (5), is the specific form of the Fisher information matrix of the Cramer-Rao bound, The following formula is obtained (calculation The formula is given as formula (6):

[0083]

[0084]

[0085] In formula (6), superscript (·) r represents the received signal, τ n,k,l Indicates the transmission delay, represents the real part of the large-scale fading factor, represents the imaginary part of the large-scale fading factor, with the superscript (·) * Indicates conjugation, superscript (·) H represents conjugate transpose, l' represents signals of the same or different paths, l=l' represents signals of the same path, l≠l' represents signals of different paths, Indicates taking the real part of a complex number, α l is the large-scale fading factor, They are expressed by equations (7) and (8) respectively:

[0086] γ p =N rx N tx N p / P n (7)

[0087] In formula (7), N rx is the number of transmitting antennas, N tx is the number of receiving antennas, N p is the pilot symbol length, P n is the transmitted pilot power.

[0088]

[0089] In formula (8), |P(f)| 2 is the probability distribution function of power.

[0090] The error performance bound of the pilot signal parameters of the Cramer-Rao bounded Fisher information matrix of a single AP can be obtained by calculating the Fisher information matrix.

[0091]

[0092]

[0093]

[0094] In formula (11), B eff Indicates bandwidth.

[0095]

[0096] In step 5, the Fisher information matrix of each AP's Cramer-Rao bound is combined to obtain the perception performance bound of the AP group composed of multiple APs. The perception performance bound without a cellular network can be calculated using the formula, which is the highest achievable positioning accuracy. The following perception performance bounds are simulated using mathematical tools. The perception performance bounds are calculated using the Cramer-Rao bound's Fisher information matrix. The range domain positioning accuracy bound and the angle domain positioning accuracy bound are derived from formulas (13a) and (13b), respectively:

[0097]

[0098]

[0099] The subscript (·) in formula (13) sp Indicates the positioning accuracy limit in the range domain, subscript (·) so Indicates the positioning accuracy limit in the angle domain, with a superscript (·) -1 represents the inverse of the matrix, where p k represents the Cartesian coordinate vector of the user, Represents the phase shift of the antenna receiving signal in the horizontal direction, represents the longitudinal phase shift of the antenna receiving signal, It is expressed by formula (14):

[0100]

[0101] The superscript (·) in formula (14) T represents the matrix transpose, Expressed as formula (15a), T n,k Expressed as formula (15b):

[0102]

[0103]

[0104] In formula (15), only the LOS diameter is considered for the convenience of calculation, that is, where l=1, They are expressed as formula (16a) and (16b) respectively, Expressed as formula (17a-17d):

[0105]

[0106]

[0107] In formula (16)

[0108]

[0109]

[0110]

[0111] The elements in formula (17) are all the calculation results of formula (6).

[0112] In step 6, after obtaining the perception performance bounds, a perception accuracy heat map can be plotted based on the performance bounds. This heat map describes the positioning accuracy bounds that can be achieved for each user on the map based on the distribution of APs within a three-dimensional plane. This heat map can be used to describe the positioning performance of the AP group for users in different locations in a non-cellular network architecture. By varying the selection of synaesthesia fusion access points within this method, perception accuracy heat maps can be obtained for different synaesthesia fusion access point selection distributions. By analyzing the heat map sampling points, the mean sampling point values ​​for this simulation meet the angular performance bound of 0.1523° and the distance accuracy bound of 0.0648m. The synaesthesia fusion access point selection distribution meets the positioning accuracy requirements.

[0113] In summary, the present invention addresses the problem of selecting synaesthesia-fusion access points (APs) based on positioning information for synaesthesia-fusion-assisted communication in a non-cellular massive MIMO architecture. This method introduces a Cramer-Rao bounded perception accuracy heat map method for synaesthesia-fusion AP selection, using simulation to calculate the Cramérau bound. This method addresses the problem that existing technologies fail to consider the location allocation of synaesthesia-fusion APs, making it difficult to determine the performance bounds for user positioning accuracy for jointly operating APs. Furthermore, by using the perception accuracy heat map as a reference for synaesthesia-fusion AP selection, the positioning accuracy performance of the communication APs is factored into performance considerations. This method improves the positioning accuracy of synaesthesia-fusion APs while ensuring communication speed, reducing power consumption and possessing practical significance.

[0114] Anything not described in detail in the present invention is well known to those skilled in the art.

[0115] The above is a schematic description of the present invention and its embodiments, which is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. Therefore, if a person skilled in the art is inspired by this and, without departing from the purpose of the present invention, designs a structure and embodiment similar to this technical solution without inventiveness, they shall fall within the scope of protection of the present invention.

Claims

1. A method for selecting synaesthesia fusion access points based on the Cramer-Rao bound, characterized in that: The method comprises the following steps: Step S1, simulation preparation, initializing the three-dimensional position distribution of access points AP in the simulation; Step S2: During uplink communication, all APs in the non-cellular network provide services to the entire user group, which includes all users using the same orthogonal pilot. Other user groups using different orthogonal pilots are non-interference users. Each user group is considered independently in the service scenario. The communication process can be divided into multiple time slots, each of which contains a pilot signal as well as uplink and downlink data signals. One time slot is equivalent to a complete communication process, including positioning, pilot estimation, uplink communication, and downlink communication. After receiving the pilot signal, the AP constructs a signal model as a mathematical expression of the uplink pilot signal. In step S3, in a non-cellular network, the massive MIMO antenna arrays on the AP side communicate with each other and share a common control center CPU. The signals received by all APs in the same time slot in step S1 are modeled as noise-free pilot signal models, and then summed up to form a total noise-free uplink pilot signal expression. Step S4: Convert the received total uplink pilot expression into a CRLB Cramer-Rao bounded Fisher information matrix. The key information of this information matrix is ​​considered to be the pilot signal of a certain user received by a certain AP. The Cramer-Rao bounded information matrix contains the Fisher information of the pilot signal of each user received by each AP, distributed throughout the information matrix. Step S5: The Fisher information matrix of each AP's Cramer-Rao bound is combined to obtain the sensing performance bound of the AP group composed of multiple APs. The sensing performance bound without a cellular network is calculated based on the sensing performance bound according to the formula to achieve the highest positioning accuracy. Step S6: After obtaining the perception performance limit, a perception accuracy heat map is drawn based on the performance limit. The heat map describes the positioning accuracy limit that can be achieved for each user on the map based on the distribution of APs in a three-dimensional plane. The positioning performance of the AP group for users in different locations in a non-cellular network architecture is obtained through the heat map. By changing the method for selecting synaesthesia fusion access points in this method, perception accuracy heat maps under different synaesthesia fusion access point selection distributions are obtained. By applying mean analysis, variance analysis, or sampling point analysis to the heat map, the synaesthesia fusion access point selection distribution that best suits the perception accuracy is found, thereby achieving the purpose of maximizing the AP positioning accuracy for users.

2. The method for selecting synaesthesia fusion access points based on the Cramer-Rao bound according to claim 1, characterized in that: The mathematical expression of the uplink pilot signal in step S2 is: In formula (1), superscript (·) r The superscript (·) indicates that the parameter with this superscript is the received signal related to the receiving AP. p Indicates that the parameter with this superscript is related to the pilot signal Related parameters, subscript (·) n represents the nth AP, subscript (·) k represents the kth user, subscript (·) l represents the lth path, t represents the tth moment, Represents the signal received at the AP, represents the pilot signal sent by the user, τ n,k,l Indicates the transmission delay, represents the pilot transmission power, is the antenna array used by the user to send pilot signals, represents additive Gaussian noise, Represents the root length of the pilot symbol, H n,k,l represents the channel information, α n,k,l is the large-scale fading factor, which describes the power decay during signal transmission; is the signal receiving deflection angle matrix of the MIMO antenna array at the AP end, with the subscript (·) r represents the receiving AP, where Represents the phase shift of the antenna receiving signal in the horizontal direction, Represents the longitudinal phase shift of the antenna received signal.

3. The method for selecting synaesthesia fusion access points based on the Cramer-Rao bound according to claim 1, characterized in that: The noise-free uplink pilot signal model in step S3 is: The kth user pilot signal received by the nth AP is: Formula (2) is the noise-free case of formula (1), The total noise-free uplink pilot signal expression is: In formula (4), superscript (·) p Indicates that the parameter with this superscript is related to the pilot signal Related parameters, K means there are K users in total, and N means there are N APs in total.

4. The method for selecting synaesthesia fusion access points based on the Cramer-Rao bound according to claim 1, characterized in that: The step S4 specifically includes: Step S401: Calculate the Cramer-Rao bounded Fisher information matrix based on the noise-free pilot signal model obtained in step S3: In formula (4), is the specific form of the Fisher information matrix of the Cramer-Rao bound, η n,k,l represents a complex field vector, The included Fisher information is given by the following formula: In formula (6), superscript (·) r Indicates that the parameter with this superscript is the received signal related to the receiving end AP, τ n,k,l Indicates the transmission delay, represents the real part of the large-scale fading factor, represents the imaginary part of the large-scale fading factor, with the superscript (·) * Indicates conjugation, superscript (·) H represents conjugate transpose, l' represents signals of the same or different paths, l=l' represents signals of the same path, l≠l' represents signals of different paths, Indicates taking the real part of a complex number, α l is the large-scale fading factor, γ p 、 They are expressed by equations (7) and (8) respectively: γ p =N rx N tx N p / P n (6) In formula (6), N rx is the number of receiving antennas, N tx is the number of transmitting antennas, N p is the pilot symbol length, P n is the transmitted pilot power; In formula (7), |P(f)| 2 is the probability distribution function of power, B is the bandwidth; Step S402: Obtain the error performance bound of the pilot signal parameters of the Cramer-Rao bounded Fisher information matrix of a single AP using the calculated Fisher information matrix: In formula (8), superscript express The estimator of , the following formulas have the same meaning; In formula (11), B eff Indicates bandwidth, 5. The method for selecting synaesthesia fusion access points based on the Cramer-Rao bound according to claim 1, characterized in that: The step S5 specifically includes: In step S501, the perception performance limit is calculated using the Fisher information matrix of the Cramer-Rao bound obtained in step S4. The range domain positioning accuracy limit and the angle domain positioning accuracy limit are derived from formulas (13a) and (13b), respectively: The subscript (·) in formulas (12a) and (12b) s p represents the positioning accuracy limit in the range domain, and the subscript (·) so Indicates the positioning accuracy limit in the angle domain, with a superscript (·) -1 represents the inverse of the matrix, where p k represents the Cartesian coordinate vector of the user, Represents the phase shift of the antenna receiving signal in the horizontal direction, represents the longitudinal phase shift of the antenna receiving signal, It is expressed by formula (13): The superscript (·) in formula (13) T represents the matrix transpose, Expressed as formula (14a), T n,k Expressed as formula (14b): In formula (14), only the LOS diameter is considered for the convenience of calculation, that is, where l=1, They are expressed as formula (15a) and (15b) respectively, Expressed as formula (16a-16d): Official (16) The elements in formulas (16a), (16b), (16c), and (16d) are all the calculation results of formula (5).

6. The method for selecting synaesthesia fusion access points based on the Cramer-Rao bound according to claim 1, characterized in that: The mean analysis described in step S6 is to take the mean of the entire heat map, and the result obtained is the average positioning accuracy information, which represents the average positioning accuracy of the AP group in the area; the variance analysis described in step S6 is to take the mean square deviation of the entire heat map, and the result obtained is the degree of fluctuation of the positioning accuracy, which represents the degree of fluctuation of the positioning accuracy of the AP group in the area; the sampling point analysis described in step S6 is to take the mean of the key points in the heat map, and the result obtained is the average positioning accuracy of the key points, which represents the positioning capability at the key sampling points.

Images