A tunnel edge computing terminal layout method based on channel map construction

Abstract

The present invention relates to the technical field of wireless information transmission, and discloses a tunnel edge computing terminal layout method based on channel map construction, comprising the steps of: constructing a channel impulse response model, generating an adaptive dynamic noise threshold, extracting effective multipath components, calculating sparse channel data, channel data mapping, completing a channel data matrix, displaying a channel map, optimizing the layout, and the like; the tunnel edge computing terminal layout method based on the channel map construction constructs a channel model based on the electromagnetic wave transmission characteristics of a tunnel scene, constructs a complete and accurate three-dimensional channel map of the tunnel scene by accurately and in real time extracting a variety of sparse channel data in the tunnel, obtains channel data such as propagation path information, receiving power, path loss, root mean square delay spread and channel capacity, realizes effective layout optimization of tunnel edge computing terminals, realizes effective extraction of a variety of channel data of tunnel scenes, and realizes complete construction of a three-dimensional channel map of the tunnel scene.

Description

Technical Field

[0001] The present invention relates to the field of wireless information transmission technology, and in particular to a tunnel edge computing terminal layout method based on channel map construction. Background Art

[0002] Highway tunnels are important transportation facilities that connect both sides of geographical barriers. They play an important role in ensuring smooth and convenient traffic and promoting regional economic development. With the completion and opening of special sections such as extra-long tunnels and tunnel groups, tunnel construction, management and maintenance are facing many new problems. The rapid development of emerging technologies such as the Internet of Things, new generation mobile communications, and edge computing will help promote the intelligent and unmanned development of tunnel construction, management and maintenance, which is of great significance for further reducing costs and increasing efficiency, and improving the safety and stability of tunnel operation and management.

[0003] It is difficult to achieve effective and reliable data sharing and linkage between various devices such as data acquisition sensors and control systems with a large number of complex layouts in the tunnel. The redundant data resources are also very complicated for subsequent analysis and processing. The tunnel edge computing terminal is compatible with various transmission protocols and standards, and can effectively and reliably connect various devices in the tunnel and collect, process and analyze data in real time. Some tunnel equipment uses wireless communication technologies such as 4G or 5G for data transmission, which has the advantages of more flexibility, convenience and low cost. However, due to the special closed structure of the tunnel, the transmission characteristics of electromagnetic waves are quite different from those of traditional ground. How to effectively layout edge computing terminals and achieve reliable and stable access to various wireless transmission devices in the tunnel has become a key issue that needs to be solved urgently.

[0004] Traditional edge computing terminal layout methods usually use ultra-low latency transmission of devices as the main constraint to solve the optimal layout quantity and location. Although this can meet the needs of edge computing terminal layout optimization to a certain extent and improve the operation and management efficiency of various devices, due to the special wireless communication environment of the tunnel, this has a significant impact on the reliable and stable connection of wireless transmission equipment. Summary of the invention

[0005] To solve the above problems, the invention proposes a tunnel edge computing terminal layout method based on channel map construction; this method constructs a channel model based on the electromagnetic wave transmission characteristics of the tunnel scene, and constructs a complete and accurate three-dimensional channel map of the tunnel scene through accurate and real-time extraction of various sparse channel data in the tunnel, thereby effectively and conveniently realizing the layout optimization of tunnel edge computing terminals.

[0006] The technical solution adopted by the present invention is:

[0007] A tunnel edge computing terminal layout method based on a channel map comprises the following steps:

[0008] Step 1: Construct a channel impulse response model: Based on the wireless transmission characteristics of electromagnetic waves in the tunnel, the user constructs a channel impulse response model h in the tunnel scenario. p,q (t,τ); The influence of direct path and indirect path is described by channel impulse response model;

[0009] Step 2, generating an adaptive dynamic noise threshold: the user generates an adaptive dynamic noise threshold according to the channel impulse response of the wireless channel in the tunnel scenario;

[0010] Step 3: Extract effective multipath components: Based on the adaptive dynamic noise threshold T p,q [m], extracting the effective multipath component in the channel impulse response of the tunnel wireless channel;

[0011] Step 4: Calculate sparse channel data: The user is based on the effective multipath component MPC in the channel impulse response p,q Calculate sparse channel data in tunnel scenarios;

[0012] Step 5, channel data mapping: the user maps the acquired sparse channel data in the tunnel scenario to the channel map model;

[0013] Step 6: Complete the channel data matrix: The user completes the sparse channel data matrix M in the tunnel scenario. tunnel Complete the channel map in the tunnel scenario.

[0014] Step 7, channel map display and layout optimization: Based on the constructed channel map, intuitively display the electromagnetic wave transmission characteristics in the tunnel scenario, including propagation path power, propagation path delay, received power, path loss, root mean square delay spread and channel capacity; use the channel map to optimize the layout of the tunnel edge computing terminal.

[0015] Furthermore, in step 1, the channel impulse response model h p,q (t,τ) is specifically expressed as the superposition of the direct path component of electromagnetic wave propagation in the tunnel scene and the indirect path component reflected by the tunnel top, ground and side walls;

[0016] As shown below:

[0017]

[0018] In the formula, h p,q (t,τ) represents the channel impulse response from the wireless transmission device in the p-th tunnel to the edge computing terminal in the q-th tunnel, which depends on the time t and the delay τ; where p and q represent the wireless transmission device in the p-th tunnel and the edge computing terminal in the q-th tunnel, respectively, p=1,2,...,P, q=1,2,...,Q; represents the power gain of the direct path; represents the channel fading coefficient of the direct path; δ(τ-τ LoS (t)) and is the impulse function, τ LoS (t) represents the delay of the direct path, Indicates the first NLoS The delay of the indirect path; L NLoS represents the number of indirect paths in the wireless channel of the tunnel scenario, l NLoS =1,2,...,L NLoS ; Indicates the first NLoS The power gain of the indirect path; Indicates the first NLoS The channel fading coefficient of the indirect path.

[0019] Further, in step 2, the adaptive dynamic noise threshold generation process includes the following steps:

[0020] Step 2.1, set parameters: the user sets the sliding window size to N W , the number of protected cells is N G , the false alarm probability is p f ;

[0021] False alarm probability p f , and its calculation formula is as follows:

[0022]

[0023] In the formula, i is a counting variable, which ranges from 0 to changes; α represents the threshold factor;

[0024] Step 2.2, Calculate power estimate: Calculate the power estimate P of the lagged cell left [m], and calculate the power estimate P of the leading cell right [m];

[0025] The calculation formula is as follows:

[0026]

[0027] Where m represents the index of the current processing time or space position; h p,q [w,τ] represents the channel impulse response from the wireless transmission device in the pth tunnel to the edge computing terminal in the qth tunnel, which depends on the time w and the delay τ;

[0028] Step 2.3, calculate the background noise power: Calculate the estimated value Z of the background noise power p,q [m];

[0029] The calculation formula is as follows:

[0030] Z p,q [m] = min{P left [m],P right [m]}

[0031] In the formula, min{·} means to find the minimum value;

[0032] Step 2.4, calculate the threshold factor: based on the false alarm probability p f The expression further calculates the threshold factor α;

[0033] The calculation formula is as follows:

[0034]

[0035] Step 2.5, calculate the adaptive dynamic noise threshold: calculate the adaptive dynamic noise threshold T p,q [m];

[0036] The calculation formula is as follows:

[0037] T p,q [m] = α·Z p,q [m].

[0038] Further, in step 3, the effective multipath component extraction process in the channel impulse response includes the following steps:

[0039] Step 3.1, obtaining peak data in the channel impulse response: obtaining peak data in the channel impulse response according to the following formula;

[0040]

[0041] Where peak(·) represents the peak value calculation of the channel impulse response; h p,q [m,τ] represents the channel impulse response from the wireless transmission device in the pth tunnel to the edge computing terminal in the qth tunnel, which depends on the time m and the delay τ; l pea k Variable representing the peak index, the value range is: l peak =1,2,...,L peak ; L peak Represents the number of peaks in the channel impulse response; represents the lth channel impulse response pea k peak value;

[0042] Step 3.2, determine and extract the effective multipath component: According to the following formula, the user is based on the adaptive dynamic noise threshold T p,q[m] Determine the peak values ​​of the channel impulse response in the tunnel scene and extract the effective multipath component MPC p,q ;

[0043]

[0044] In the formula, α l represents the path amplitude of the lth effective propagation path in the wireless channel of the tunnel scenario; τ l represents the delay of the lth effective propagation path in the wireless channel of the tunnel scenario; peak(·) represents the peak value calculation of the channel impulse response; h p,q [m,τ] represents the channel impulse response from the wireless transmission device in the pth tunnel to the edge computing terminal in the qth tunnel, which depends on the time m and the delay τ; T p,q [m,τ] represents the adaptive dynamic noise threshold; l represents the variable of the effective propagation path number index, and the value range is: l=1,2,...,L; L represents the effective propagation path number in the wireless channel of the tunnel scenario.

[0045] Furthermore, in step 3, the peak value must satisfy the following conditions:

[0046]

[0047] In the formula, represents the first lpeak The previous sampling point of the peak value; represents the first lpeak The next sampling point after the peak value.

[0048] Further, in step 4, the sparse channel data calculation in the tunnel scenario includes the following steps:

[0049] Step 4.1, calculation of path power of effective propagation path: Calculate the path power P of the lth effective propagation path in the tunnel scenario according to the following formula: l ;

[0050]

[0051] Where P t Indicates the transmission power of the wireless transmission equipment in the tunnel; h p,q (τ l ) is the channel impulse response of the lth path under the delay τ;

[0052] Step 4.2, signal receiving power calculation: According to the following formula, sum the effective propagation path power at each delay and calculate the signal receiving power P at the tunnel edge computing terminal r ;

[0053]

[0054] Where P t represents the transmission power of the wireless transmission equipment in the tunnel; L represents the number of effective propagation paths in the wireless channel of the tunnel scene; l represents the variable of the effective propagation path number index, and the value range is: l=1,2,...,L; h p,q (τ l ) is the channel impulse response of the lth path under the delay τ;

[0055] Step 4.3, path loss calculation: Calculate the path loss PL in the tunnel scenario according to the following formula;

[0056]

[0057] Where L represents the number of effective propagation paths in the wireless channel of the tunnel scenario; l represents the variable of the effective propagation path number index, and its value range is: l = 1, 2, ..., L; P l represents the power of the lth path;

[0058] Step 4.4, root mean square delay spread calculation: Calculate the root mean square delay spread RMS-DS in the tunnel scenario according to the following formula;

[0059]

[0060] Where L represents the number of effective propagation paths in the wireless channel of the tunnel scenario; l represents the variable of the effective propagation path number index, and its value range is: l = 1, 2, ..., L; τ l represents the delay of the lth path; represents the power-weighted average delay in the channel impulse response;

[0061] Step 4.5, channel capacity calculation: Calculate the channel capacity C in the tunnel scenario according to the following formula;

[0062]

[0063] In the formula, det(·) represents the determinant operation; I represents the unit matrix; ρ represents the signal-to-noise ratio; (·) H represents the conjugate transpose operation; H represents the channel matrix of the constructed tunnel scene channel model, H = [h p,q (t,τ)] P×Q ;

[0064] Step 4.6, define a sparse channel data set: the user defines a sparse channel data set K in a tunnel scenario. The specific expression of the sparse channel data set K in the tunnel scenario is as follows:

[0065]

[0066] Where P l represents the power of the lth path; τ l represents the delay of the lth path; P r is the signal receiving power at the terminal at the tunnel edge; PL is the path loss in the tunnel scenario; RMS-DS is the root mean square delay spread in the tunnel scenario; C is the channel capacity in the tunnel scenario; l represents the variable indexed by the number of effective propagation paths, and its value range is: l=1,2,...,L; L represents the number of effective propagation paths in the wireless channel of the tunnel scenario.

[0067] Furthermore, in step 4, the calculation process of the power-weighted average delay τ in the channel impulse response is as follows:

[0068]

[0069] Further, in step 5, the mapping process of sparse channel data in the tunnel scenario includes the following steps:

[0070] Step 5.1, divide the 3D cube area: the user divides the tunnel 3D scene approximately into N total = a three-dimensional cube region composed of X×Y×Z (x∈X, y∈Y, z∈Z) sub-cubes; assuming that the channel data value collected by each sub-cube is the channel data value in the entire sub-cube region;

[0071] Step 5.2, construct the channel data matrix: construct a three-dimensional channel map model of the tunnel scene, specifically represented as the channel data matrix between any two sub-cubes in the three-dimensional area of ​​the tunnel The matrix element m i,j As shown below:

[0072]

[0073] In the formula, K i,j represents the i-th (i=1,2,...,N total )Subcube(x i ,y i ,z i ) and the jth (j=1,2,...,N total )Subcube(x j ,y j ,z j ), when i=j, the two sub-cubes overlap and there is no channel data;

[0074] Step 5.3, mapping sparse channel data: Map the sparse channel data collected in the three-dimensional cube area of ​​the tunnel to the channel data matrix M to obtain the sparse channel data matrix M in the three-dimensional tunnel scene. tunnel ;

[0075] Step 5.4, define the channel dataset: The user defines the channel dataset K in the tunnel channel map model i,j , the specific expression is as follows:

[0076]

[0077] In the formula, K i,j represents the channel dataset between the i-th sub-cube and the j-th sub-cube in the tunnel 3D scene; P i,j；l represents the path power in the lth effective propagation path in the channel impulse response between the i-th sub-cube and the j-th sub-cube; τ i,j；l represents the path delay in the lth effective propagation path in the channel impulse response between the i-th sub-cube and the j-th sub-cube; P r,i,j represents the received power between the i-th sub-cube and the j-th sub-cube; PL i,j represents the path loss between the i-th sub-cube and the j-th sub-cube; RMS-DS i,j represents the RMS delay spread between the i-th subcube and the j-th subcube; C i,j represents the channel capacity between the i-th sub-cube and the j-th sub-cube; l represents the variable indexed by the number of effective propagation paths, and its value range is: l=1,2,...,L; L represents the number of effective propagation paths in the wireless channel of the tunnel scenario.

[0078] Further, in step 6, the tunnel sparse channel data matrix completion process includes the following steps:

[0079] Step 6.1, estimate the channel data in the uncollected sub-cube: the channel data in the uncollected sub-cube in the tunnel 3D cube area It is specifically expressed as the weighted average of the channel data in all the acquisition sub-cubes in the three-dimensional cube area, which is specifically expressed as follows:

[0080]

[0081] In the formula, represents the estimated value of the channel data in the uncollected sub-cube (x0, y0, z0); ω u represents weight; K u represents the uth (u=1,2,...,N) collection sub-cube (x u ,y u ,zu ) is the true value of the channel data; N represents the number of acquisition sub-cubes in the tunnel scene;

[0082] Step 6.2, calculate the Euclidean distance between each acquisition sub-cube: The user calculates the Euclidean distance between each acquisition sub-cube in the tunnel scene. The calculation formula is as follows:

[0083]

[0084] Where, d uv represents the Euclidean distance between the u-th acquisition sub-cube and the v-th acquisition sub-cube in the tunnel scene; the position coordinates of the u-th (u=1,2,...,N) acquisition sub-cube are (x u ,y u ,z u ); the position coordinates of the vth (v=1,2,...,N) acquisition sub-cube are (x v ,y v ,z v );

[0085] Step 6.3, calculate semi-variance: According to the following formula, the user calculates the semi-variance between each acquisition sub-cube;

[0086]

[0087] In the formula, γ uv represents the semi-variance between the u-th acquisition sub-cube and the v-th acquisition sub-cube in the tunnel scene; K u represents the uth (u=1,2,...,N) collection sub-cube (x u ,y u ,z u ) is the true value of the channel data; K v represents the νth (ν=1,2,...,N) collection sub-cube (x ν ,y ν ,z ν ) in the real value of the channel data;

[0088] Step 6.4, fitting the semivariance function: The user fits the semivariance function γ according to the Euclidean distance and semivariance between each acquisition sub-cube in the tunnel scene; as shown in the following formula:

[0089] γ=γ(d)

[0090] Where d represents the Euclidean distance between any two sub-cubes in the tunnel scene;

[0091] Step 6.5, calculate the semivariance between the uncollected sub-cube and the collected sub-cube: The user calculates the semivariance γ between the uncollected sub-cube (x0, y0, z0) and all collected sub-cubes according to the semivariance function γ in the fitted tunnel scene. i0 ; Wherein, i represents the i-th (i=1,2,...,N) acquisition sub-cube in the tunnel scene;

[0092] Step 6.6, calculate the optimal weight: The user calculates the optimal weight according to the following formula

[0093]

[0094] Where φ represents the Lagrange multiplier;

[0095] Step 6.7, calculate the estimated value of the channel data: Calculate the estimated value of the channel data in the uncollected sub-cube in the tunnel three-dimensional cube area according to the following formula:

[0096]

[0097] In the formula, Represents the estimated value of the channel data in the uncollected sub-cube (x0, y0, z0); represents the optimal weight; K u represents the uth (u=1,2,...,N) collection sub-cube (x u ,y u ,z u ) is the true value of the channel data.

[0098] The beneficial effects of the present invention are:

[0099] 1. The tunnel edge computing terminal layout method based on channel map construction can accurately and completely construct the channel map of the three-dimensional electromagnetic scene of the tunnel, obtain channel data such as propagation path information, received power, path loss, root mean square delay spread and channel capacity, and realize the effective layout optimization of tunnel edge computing terminals.

[0100] 2. The adaptive dynamic noise threshold generation process of the tunnel edge computing terminal layout method based on the channel map can generate an adaptive dynamic noise threshold for the channel impulse response of the wireless channel in the tunnel scene, thereby realizing the effective extraction of multiple channel data in the tunnel scene.

[0101] 3. The process of completing the sparse channel data of the tunnel scene in the tunnel edge computing terminal layout method based on the channel map can accurately complete the sparse channel data and realize the complete construction of the three-dimensional tunnel scene channel map. BRIEF DESCRIPTION OF THE DRAWINGS

[0102] Figure 1 is a flow chart of a tunnel edge computing terminal layout method based on a channel map;

[0103] Figure 2 is a flow chart of the process of generating an adaptive noise threshold in the present invention;

[0104] Figure 3 It is a flow chart of the sparse channel data completion process in the tunnel scenario of the present invention. DETAILED DESCRIPTION

[0105] The following will be combined with the accompanying drawings of the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0106] This embodiment provides a tunnel edge computing terminal layout method based on channel map construction, such as Figure 1 As shown, the tunnel edge computing terminal layout method based on the channel map includes the following steps:

[0107] Step 1: Construct a channel impulse response model:

[0108] Based on the wireless transmission characteristics of electromagnetic waves in the tunnel, the user constructs the channel impulse response model h in the tunnel scenario in this example. p,q (t,τ); the channel impulse response model is used to describe the impact of the direct path and the indirect path.

[0109] Channel impulse response model h p,q (t,τ) is specifically expressed as the superposition of the direct path component of electromagnetic wave propagation in the tunnel scene and the indirect path component reflected by the tunnel top, ground and side walls, as shown in the following formula:

[0110]

[0111] In the formula, h p,q (t,τ) represents the channel impulse response from the wireless transmission device in the p-th tunnel to the edge computing terminal in the q-th tunnel, which depends on the time t and the delay τ; where p and q represent the wireless transmission device in the p-th tunnel and the edge computing terminal in the q-th tunnel, respectively, p=1,2,...,P, q=1,2,...,Q;

[0112] represents the power gain of the direct path;

[0113] represents the channel fading coefficient of the direct path;

[0114] δ(τ-τ LoS (t)) and is the impulse function, τ LoS (t) represents the delay of the direct path, Indicates the first NLoS The delay of the indirect path;

[0115] L NLoS represents the number of indirect paths in the wireless channel of the tunnel scenario, l NLoS =1,2,...,L NLoS ;

[0116] Indicates the first NLoS The power gain of the indirect path;

[0117] Indicates the first NLoS The channel fading coefficient of the indirect path.

[0118] The channel impulse response model combines the influence of the direct path and multiple indirect paths in the tunnel environment, and describes the propagation characteristics of electromagnetic waves by using power gain and channel fading coefficient. The channel impulse response model can more accurately reflect the propagation of wireless signals in the tunnel, providing a basis for further communication system design and performance analysis.

[0119] Step 2, adaptive dynamic noise threshold generation:

[0120] like Figure 2 As shown, the user generates an adaptive dynamic noise threshold based on the channel impulse response of the wireless channel in the tunnel scenario.

[0121] The adaptive dynamic noise threshold generation process includes the following steps:

[0122] Step 2.1, set parameters:

[0123] In this example, the user sets the sliding window size to N W , the number of protected cells is N G , the false alarm probability is p f .

[0124] False alarm probability p f , and its calculation formula is as follows:

[0125]

[0126] In the formula, i is a counting variable, which ranges from 0 to changes; α represents the threshold factor.

[0127] Step 2.2, calculate the power estimate:

[0128] Calculate the power estimate P of the lag cell left [m], and calculate the power estimate P of the leading cell right [m]; the calculation formula is as follows:

[0129]

[0130] Where m represents the index of the current processing time or space position;

[0131] h p,q [w,τ] represents the channel impulse response from the wireless transmission device in the pth tunnel to the edge computing terminal in the qth tunnel, which depends on the time w and the delay τ.

[0132] Step 2.3, calculate the background noise power:

[0133] Calculate an estimate of the background noise power Z p,q [m]; the calculation formula is as follows:

[0134] Z p,q [m] = min{P left [m],P right [m]}

[0135] In the formula, min{·} means to find the minimum value;

[0136] Step 2.4, calculate the threshold factor:

[0137] Based on the false alarm probability p f The expression of further calculates the threshold factor α; the calculation formula is as follows:

[0138]

[0139] Step 2.5, calculate the adaptive dynamic noise threshold:

[0140] Calculate the adaptive dynamic noise threshold T p,q [m]; the calculation formula is as follows:

[0141] T p,q [m] = α·Z p,q [m]

[0142] The above steps describe in detail how to generate an adaptive dynamic noise threshold that can be used for noise management in a wireless channel.

[0143] Step 3, extract the effective multipath components:

[0144] Based on adaptive dynamic noise threshold T p,q[m], extract the effective multipath components in the channel impulse response of the tunnel wireless channel.

[0145] The effective multipath component extraction process in the channel impulse response includes the following steps:

[0146] Step 3.1, obtain the peak data in the channel impulse response:

[0147] The peak data in the channel impulse response is obtained according to the following formula;

[0148]

[0149] Where peak(·) represents the peak value calculation of the channel impulse response;

[0150] h p,q [m,τ] represents the channel impulse response from the wireless transmission device in the pth tunnel to the edge computing terminal in the qth tunnel, which depends on the time m and the delay τ;

[0151] l peak Variable representing the peak index, the value range is: l peak =1,2,...,L peak ;

[0152] L peak Represents the number of peaks in the channel impulse response;

[0153] represents the lth channel impulse response peak A peak value.

[0154] Furthermore, the peak value must satisfy the following conditions:

[0155]

[0156] In the formula, represents the lth channel impulse response peak The previous sampling point of the peak value; represents the lth channel impulse response peak The next sampling point after the peak value;

[0157] Step 3.2, determine and extract the effective multipath components:

[0158] According to the following formula, users can use the adaptive dynamic noise threshold T p,q [m] Determine the peak values ​​of the channel impulse response in the tunnel scene in this example and extract the effective multipath component MPC p,q ;

[0159]

[0160] In the formula, αl represents the path amplitude of the lth effective propagation path in the wireless channel of the tunnel scenario;

[0161] τ l represents the delay of the lth effective propagation path in the wireless channel of the tunnel scenario;

[0162] peak(·) represents the peak value calculation of the channel impulse response;

[0163] h p,q [m,τ] represents the channel impulse response from the wireless transmission device in the pth tunnel to the edge computing terminal in the qth tunnel, which depends on the time m and the delay τ;

[0164] T p,q [m,τ] represents the adaptive dynamic noise threshold;

[0165] l represents the variable of the index of the number of effective propagation paths, and its value range is: l=1,2,...,L;

[0166] L represents the number of effective propagation paths in the wireless channel of the tunnel scenario.

[0167] Through the above steps, the effective multipath components in the channel impulse response of the tunnel wireless channel can be effectively extracted, providing basic data for the subsequent optimization and analysis of the wireless communication system.

[0168] Step 4, calculate sparse channel data:

[0169] User-based effective multipath component MPC in channel impulse response p,q Compute sparse channel data in tunnel scenarios.

[0170] The sparse channel data calculation in the tunnel scenario includes the following steps:

[0171] Step 4.1, path power calculation of effective propagation path:

[0172] According to the following formula, the path power P of the lth effective propagation path in the tunnel scenario is calculated: l ;

[0173] P l =P t |h p,q (τ l )| 2

[0174] Where P t Indicates the transmission power of the wireless transmission equipment in the tunnel in this example;

[0175] h p,q (τ l ) is the lThe channel impulse response of the path under the delay τ.

[0176] Step 4.2, signal receiving power calculation:

[0177] According to the following formula, the effective propagation path power at each delay is summed to calculate the signal receiving power P at the tunnel edge computing terminal in this example: r ;

[0178]

[0179] Where P t Indicates the transmission power of the wireless transmission equipment in the tunnel in this example;

[0180] L represents the number of effective propagation paths in the wireless channel of the tunnel scenario;

[0181] l represents the variable of the index of the number of effective propagation paths, and its value range is: l=1,2,...,L;

[0182] h p,q (τ l ) is the channel impulse response of the lth path under the delay τ.

[0183] Step 4.3, path loss calculation:

[0184] According to the following formula, the path loss PL in the tunnel scenario is calculated;

[0185]

[0186] Where L represents the number of effective propagation paths in the wireless channel of the tunnel scenario;

[0187] l represents the variable of the index of the number of effective propagation paths, and its value range is: l=1,2,...,L;

[0188] P l represents the power of the lth path.

[0189] Step 4.4, RMS delay spread calculation:

[0190] The root mean square delay spread RMS-DS in the tunnel scenario is calculated according to the following formula:

[0191]

[0192] Where L represents the number of effective propagation paths in the wireless channel of the tunnel scenario;

[0193] l represents the variable of the index of the number of effective propagation paths, and its value range is: l=1,2,...,L;

[0194] τ lrepresents the delay of the lth path;

[0195] It represents the power-weighted average delay in the channel impulse response.

[0196] Furthermore, the power-weighted average delay in the channel impulse response is The calculation process is as follows:

[0197]

[0198] Step 4.5, channel capacity calculation:

[0199] According to the following formula, the channel capacity C in the tunnel scenario is calculated;

[0200]

[0201] In the formula, det(·) represents the determinant operation;

[0202] I represents the identity matrix;

[0203] ρ represents the signal-to-noise ratio;

[0204] (·) H represents the conjugate transpose operation;

[0205] H represents the channel matrix of the tunnel scenario channel model constructed in this example, H = [h p,q (t,τ)] P×Q .

[0206] Step 4.6, define the sparse channel dataset:

[0207] The user defines the sparse channel dataset K in the tunnel scenario in this example. The specific expression of the sparse channel dataset K in the tunnel scenario is as follows:

[0208]

[0209] Where P l represents the power of the lth path;

[0210] τ l represents the delay of the lth path;

[0211] P r Calculate the signal received power at the terminal for the tunnel edge;

[0212] PL is the path loss in the tunnel scenario;

[0213] RMS-DS is the root mean square delay spread in the tunnel scenario;

[0214] C is the channel capacity in the tunnel scenario;

[0215] l represents the variable of the index of the number of effective propagation paths, and its value range is: l=1,2,...,L;

[0216] L represents the number of effective propagation paths in the wireless channel of the tunnel scenario.

[0217] The above calculation process of sparse channel data in the tunnel scenario covers path power, received power, path loss, RMS delay spread, channel capacity and the final data set definition.

[0218] Step 5, channel data mapping:

[0219] The user maps the acquired sparse channel data in the tunnel scenario in this example to the channel map model.

[0220] The mapping process of sparse channel data in the tunnel scenario includes the following steps:

[0221] Step 5.1, divide the three-dimensional cube area:

[0222] The user divides the 3D tunnel scene in this example into N total = A three-dimensional cube area composed of X×Y×Z (x∈X, y∈Y, z∈Z) sub-cubes; this example assumes that the channel data values ​​collected by each sub-cube are the channel data values ​​in the entire sub-cube area.

[0223] Step 5.2, construct the channel data matrix:

[0224] Construct a 3D channel map model of the tunnel scene, which is specifically represented as a channel data matrix between any two sub-cubes in the 3D area of ​​the tunnel. The matrix element m i,j As shown below:

[0225]

[0226] In the formula, K i,j represents the i-th (i=1,2,...,N total )Subcube(x i ,y i ,z i ) and the jth (j=1,2,...,N total )Subcube(x j ,y j ,z j ), when i=j, the two sub-cubes overlap and there is no channel data.

[0227] Step 5.3, map sparse channel data:

[0228] Map the sparse channel data collected in the three-dimensional cube area of ​​the tunnel to the channel data matrix M to obtain the sparse channel data matrix M in the three-dimensional tunnel scene tunnel .

[0229] Step 5.4, define the channel data set:

[0230] The user defines the channel dataset K in the tunnel channel map model in this example. i,j , the specific expression is as follows:

[0231]

[0232] In the formula, K i,j Represents the channel data set between the i-th sub-cube and the j-th sub-cube in the tunnel three-dimensional scene;

[0233] P i,j；l represents the path power in the lth effective propagation path in the channel impulse response between the i-th sub-cube and the j-th sub-cube;

[0234] τ i,j；l represents the path delay in the lth effective propagation path in the channel impulse response between the i-th sub-cube and the j-th sub-cube;

[0235] P r,i,j represents the received power between the i-th sub-cube and the j-th sub-cube;

[0236] PL i,j represents the path loss between the i-th sub-cube and the j-th sub-cube;

[0237] RMS-DS i,j represents the RMS delay spread between the i-th subcube and the j-th subcube;

[0238] C i,j represents the channel capacity between the i-th subcube and the j-th subcube;

[0239] l represents the variable of the index of the number of effective propagation paths, and its value range is: l=1,2,...,L;

[0240] L represents the number of effective propagation paths in the wireless channel of the tunnel scenario.

[0241] The above steps provide a detailed methodology for mapping sparse channel data in tunnel scenarios, and channel performance can be effectively analyzed and optimized through structured data.

[0242] Step 6, complete the channel data matrix:

[0243] like Figure 3As shown, the user has a sparse channel data matrix M in the tunnel scenario in this example tunnel Complete the channel map in the tunnel scenario.

[0244] The tunnel sparse channel data matrix completion process includes the following steps:

[0245] Step 6.1, estimate the channel data in the uncollected sub-cube:

[0246] Channel data in the sub-cubes are not collected in the tunnel 3D cube area It can be specifically expressed as the weighted average of the channel data in all the acquisition sub-cubes in the three-dimensional cube area, which is specifically expressed as shown in the following formula:

[0247]

[0248] In the formula, Indicates the estimated value of the channel data in the uncollected sub-cube (x0, y0, z0) in this example;

[0249] ω u represents weight;

[0250] K u Indicates the uth (u=1,2,...,N) collection sub-cube (x u ,y u ,z u ) in the real value of the channel data;

[0251] N represents the number of collected sub-cubes in the tunnel scene of this example.

[0252] Step 6.2, calculate the Euclidean distance between each collected sub-cube:

[0253] The user calculates the Euclidean distance between each collection sub-cube in the tunnel scene of this example. The calculation formula is as follows:

[0254]

[0255] Where, d uv represents the Euclidean distance between the u-th collection sub-cube and the v-th collection sub-cube in the tunnel scene of this example; the position coordinates of the u-th (u=1,2,...,N) collection sub-cube are (x u ,y u ,z u ); the position coordinates of the vth (v=1,2,...,N) acquisition sub-cube are (x v ,y v ,z v ).

[0256] Step 6.3, calculate the semivariance:

[0257] According to the following formula, the user calculates the semi-variance between each acquisition sub-cube;

[0258]

[0259] In the formula, γ uv represents the semi-variance between the u-th acquisition sub-cube and the v-th acquisition sub-cube in the tunnel scenario of this example;

[0260] K u Indicates the uth (u=1,2,...,N) collection sub-cube (x u ,y u ,z u ) in the real value of the channel data;

[0261] K v Indicates the νth (ν=1,2,...,N) collection sub-cube (x ν ,y ν ,z ν ) is the true value of the channel data.

[0262] Step 6.4, fitting the semivariogram function:

[0263] The user fits the semivariance function γ based on the Euclidean distance and semivariance between each acquisition sub-cube in the tunnel scene in this example; as shown in the following formula:

[0264] γ=γ(d)

[0265] Where d represents the Euclidean distance between any two sub-cubes in the tunnel scene.

[0266] Step 6.5, calculate the semi-variance between the uncollected sub-cube and the collected sub-cube:

[0267] The user calculates the semivariance γ between the uncollected sub-cube (x0, y0, z0) and all collected sub-cubes according to the semivariance function γ in the tunnel scene fitted in this example. i0 ; Wherein, i represents the i-th (i=1,2,...,N) acquisition sub-cube in the tunnel scene.

[0268] Step 6.6, calculate the optimal weight:

[0269] The user calculates the optimal weight according to the following formula

[0270]

[0271] In the formula, the left side of the matrix represents the product of the semivariance matrix and the weight vector, in which each element reflects the relationship between different sub-cubes and their contribution to the final estimation result; the right side of the matrix includes the expected semivariance result; φ represents the Lagrange multiplier.

[0272] Step 6.7, calculate the channel data estimate:

[0273] According to the following formula, the estimated value of the channel data in the uncollected sub-cube in the tunnel 3D cube area in this example is calculated:

[0274]

[0275] In the formula, Indicates the estimated value of the channel data in the uncollected sub-cube (x0, y0, z0) in this example; represents the optimal weight; K u Indicates the uth (u=1,2,...,N) collection sub-cube (x u ,y u ,z u ) is the true value of the channel data.

[0276] The above steps constitute a complete channel data completion for tunnel scenarios. By weighting the collected channel data and extracting related information, the channel data of uncollected areas can be effectively estimated. This method not only improves the integrity of the channel data, but also optimizes the prediction performance of channel performance.

[0277] Step 7: Channel map display and layout optimization:

[0278] Based on the constructed channel map, the electromagnetic wave transmission characteristics in the tunnel scenario are intuitively displayed, including propagation path power, propagation path delay, received power, path loss, root mean square delay spread and channel capacity; finally, the channel map is used to optimize the layout of tunnel edge computing terminals to ensure that the wireless transmission from various wireless transmission devices to tunnel edge computing terminals has high reliability and stability.

[0279] In summary, the tunnel edge computing terminal layout method based on channel map construction can accurately and completely construct the channel map of the three-dimensional electromagnetic scene of the tunnel, obtain channel data such as propagation path information, received power, path loss, root mean square delay spread and channel capacity, and realize the effective layout optimization of the tunnel edge computing terminal; the adaptive dynamic noise threshold generation process can generate an adaptive dynamic noise threshold for the channel impulse response of the wireless channel in the tunnel scene, and realize the effective extraction of multiple channel data in the tunnel scene; the completion process of sparse channel data in the tunnel scene can generate an adaptive dynamic noise threshold for the channel impulse response of the wireless channel in the tunnel scene, and realize the effective extraction of multiple channel data in the tunnel scene.

[0280] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A tunnel edge computing terminal layout method based on channel map construction, characterized in that: The following steps are involved: Step 1: Construct a channel impulse response model: Based on the wireless transmission characteristics of electromagnetic waves in the tunnel, the user builds a channel impulse response model h in the tunnel scenario. p,q (t,τ); The influence of direct path and indirect path is described by channel impulse response model; Step 2, adaptive dynamic noise threshold generation: The user generates an adaptive dynamic noise threshold based on the channel impulse response of the wireless channel in the tunnel scenario; Step 3: Extract effective multipath components: Based on adaptive dynamic noise threshold T p,q [m], extracting the effective multipath component in the channel impulse response of the tunnel wireless channel; Step 4, calculate sparse channel data: User-based effective multipath component MPC in channel impulse response p,q Calculate sparse channel data in tunnel scenarios; Step 5, channel data mapping: The user maps the acquired sparse channel data in the tunnel scenario to the channel map model; Step 6, complete the channel data matrix: The sparse channel data matrix M of the user in the tunnel scenario tunnel Complete the channel map in the tunnel scenario. Step 7: Channel map display and layout optimization: Based on the constructed channel map, the electromagnetic wave transmission characteristics in the tunnel scenario are intuitively displayed, including propagation path power, propagation path delay, received power, path loss, root mean square delay spread and channel capacity; the channel map is used to optimize the layout of tunnel edge computing terminals.

2. The tunnel edge computing terminal layout method based on channel map construction according to claim 1 is characterized in that: In step 1, the channel impulse response model h p,q (t,τ) is specifically expressed as the superposition of the direct path component of electromagnetic wave propagation in the tunnel scene and the indirect path component reflected by the tunnel top, ground and side walls; As shown below: In the formula, h p,q (t,τ) represents the channel impulse response from the wireless transmission device in the p-th tunnel to the edge computing terminal in the q-th tunnel, which depends on the time t and the delay τ; where p and q represent the wireless transmission device in the p-th tunnel and the edge computing terminal in the q-th tunnel, respectively, p=1,2,...,P, q=1,2,...,Q; represents the power gain of the direct path; represents the channel fading coefficient of the direct path; δ(τ-τ LoS (t)) and is the impulse function, τ LoS (t) represents the delay of the direct path, Indicates the first NLoS The delay of the indirect path; L NLoS represents the number of indirect paths in the wireless channel of the tunnel scenario, l NLoS =1,2,...,L NLoS ; Indicates the first NLoS The power gain of the indirect path; Indicates the first NLoS The channel fading coefficient of the indirect path.

3. The tunnel edge computing terminal layout method based on channel map construction according to claim 1 is characterized in that: In step 2, the adaptive dynamic noise threshold generation process includes the following steps: Step 2.1, set parameters: the user sets the sliding window size to N W , the number of protected cells is N G , the false alarm probability is p f ; False alarm probability p f , and its calculation formula is as follows: In the formula, i is a counting variable, which ranges from 0 to changes; α represents the threshold factor; Step 2.2, Calculate power estimate: Calculate the power estimate P of the lagged cell left [m], and calculate the power estimate P of the leading cell right [m]; The calculation formula is as follows: Where m represents the index of the current processing time or space position; h p,q [w,τ] represents the channel impulse response from the wireless transmission device in the pth tunnel to the edge computing terminal in the qth tunnel, which depends on the time w and the delay τ; Step 2.3, calculate the background noise power: Calculate the estimated value Z of the background noise power p,q [m]; The calculation formula is as follows: Z p,q [m]=min{P left [m],P right [m]} In the formula, min{·} means to find the minimum value; Step 2.4, calculate the threshold factor: based on the false alarm probability p f The expression further calculates the threshold factor α; The calculation formula is as follows: Step 2.5, calculate the adaptive dynamic noise threshold: calculate the adaptive dynamic noise threshold T p,q [m]; The calculation formula is as follows: T p,q [m]=α·Z p,q [m]。 4. The tunnel edge computing terminal layout method based on channel map construction according to claim 1 is characterized in that: In step 3, the effective multipath component extraction process in the channel impulse response includes the following steps: Step 3.1, obtaining peak data in the channel impulse response: obtaining peak data in the channel impulse response according to the following formula; Where peak(·) represents the peak value calculation of the channel impulse response; h p,q [m,τ] represents the channel impulse response from the wireless transmission device in the pth tunnel to the edge computing terminal in the qth tunnel, which depends on the time m and the delay τ; l peak Variable representing the peak index, the value range is: l peak =1,2,...,L peak ; L peak Represents the number of peaks in the channel impulse response; represents the lth channel impulse response peak peak value; Step 3.2, determine and extract the effective multipath component: According to the following formula, the user is based on the adaptive dynamic noise threshold T p,q [m] Determine the peak values ​​of the channel impulse response in the tunnel scene and extract the effective multipath component MPC p,q ; In the formula, α l represents the path amplitude of the lth effective propagation path in the wireless channel of the tunnel scenario; τ l represents the delay of the lth effective propagation path in the wireless channel of the tunnel scenario; peak(·) represents the peak value calculation of the channel impulse response; T p,q [m,τ] represents the adaptive dynamic noise threshold from the pth wireless transmission device to the qth edge computing terminal under the time index m and the delay index τ; L represents the number of effective propagation paths in the wireless channel of the tunnel scenario; l represents the variable of the effective propagation path number index, and the value range is: l=1,2,...,L.

5. The tunnel edge computing terminal layout method based on channel map construction according to claim 4 is characterized in that: In step 3, the peak value must meet the following conditions: In the formula, represents the lth channel impulse response peak The previous sampling point of the peak value; represents the lth channel impulse response peak The next sampling point after the peak value.

6. The tunnel edge computing terminal layout method based on channel map construction according to claim 1 is characterized in that: In step 4, the sparse channel data calculation in the tunnel scenario includes the following steps: Step 4.1, calculation of path power of effective propagation path: Calculate the path power P of the lth effective propagation path in the tunnel scenario according to the following formula: l ; P l =P t |h p,q (t l )| 2 Where P t Indicates the transmission power of the wireless transmission equipment in the tunnel; h p,q (τ l ) is the channel impulse response of the lth path under the delay τ; Step 4.2, signal receiving power calculation: According to the following formula, sum the effective propagation path power at each delay and calculate the signal receiving power P at the tunnel edge computing terminal r ; Where, L represents the number of effective propagation paths in the wireless channel of the tunnel scenario; l represents the variable of the effective propagation path number index, and the value range is: l = 1, 2, ..., L; Step 4.3, path loss calculation: Calculate the path loss PL in the tunnel scenario according to the following formula; Where P l represents the power of the lth path; Step 4.4, root mean square delay spread calculation: Calculate the root mean square delay spread RMS-DS in the tunnel scenario according to the following formula; In the formula, τ l represents the delay of the lth path; represents the power-weighted average delay in the channel impulse response; Step 4.5, channel capacity calculation: Calculate the channel capacity C in the tunnel scenario according to the following formula; In the formula, det(·) represents the determinant operation; I represents the unit matrix; ρ represents the signal-to-noise ratio; (·) H represents the conjugate transpose operation; H represents the channel matrix of the constructed tunnel scenario channel model, Step 4.6, define a sparse channel data set: the user defines a sparse channel data set K in a tunnel scenario. The specific expression of the sparse channel data set K in the tunnel scenario is as follows:

7. The tunnel edge computing terminal layout method based on channel map construction according to claim 6 is characterized in that: In step 4, the power-weighted average delay τ in the channel impulse response is calculated as follows:

8. The tunnel edge computing terminal layout method based on channel map construction according to claim 1 is characterized in that: In step 5, the mapping process of sparse channel data in the tunnel scenario includes the following steps: Step 5.1, divide the 3D cube area: the user divides the tunnel 3D scene approximately into N total = a three-dimensional cube region composed of X×Y×Z (x∈X, y∈Y, z∈Z) sub-cubes; assuming that the channel data value collected by each sub-cube is the channel data value in the entire sub-cube region; Step 5.2, construct the channel data matrix: construct a three-dimensional channel map model of the tunnel scene, specifically represented as the channel data matrix between any two sub-cubes in the three-dimensional area of ​​the tunnel The matrix element m i,j As shown below: In the formula, K i,j represents the i-th (i=1,2,...,N total )Subcube(x i ,y i ,z i ) and the jth (j=1,2,...,N total )Subcube(x j ,y j ,z j ), when i=j, the two sub-cubes overlap and there is no channel data; Step 5.3, mapping sparse channel data: Map the sparse channel data collected in the three-dimensional cube area of ​​the tunnel to the channel data matrix M to obtain the sparse channel data matrix M in the three-dimensional tunnel scene. tunnel ; Step 5.4, define the channel dataset: The user defines the channel dataset K in the tunnel channel map model i,j , the specific expression is as follows: In the formula, K i,j represents the channel dataset between the i-th sub-cube and the j-th sub-cube in the tunnel 3D scene; P i,j；l represents the path power in the lth effective propagation path in the channel impulse response between the i-th sub-cube and the j-th sub-cube; τ i,j；l represents the path delay in the lth effective propagation path in the channel impulse response between the i-th sub-cube and the j-th sub-cube; P r,i,j represents the received power between the i-th sub-cube and the j-th sub-cube; PL i,j represents the path loss between the i-th sub-cube and the j-th sub-cube; RMS-DS i,j represents the RMS delay spread between the i-th subcube and the j-th subcube; C i,j represents the channel capacity between the i-th sub-cube and the j-th sub-cube; l represents the variable indexed by the number of effective propagation paths, and its value range is: l=1,2,...,L; L represents the number of effective propagation paths in the wireless channel of the tunnel scenario.

9. The tunnel edge computing terminal layout method based on channel map construction according to claim 1 is characterized in that: In step 6, the tunnel sparse channel data matrix completion process includes the following steps: Step 6.1, estimate the channel data in the uncollected sub-cube: the channel data in the uncollected sub-cube in the tunnel 3D cube area It is specifically expressed as the weighted average of the channel data in all the acquisition sub-cubes in the three-dimensional cube area, which is specifically expressed as follows: In the formula, represents the estimated value of the channel data in the uncollected sub-cube (x0, y0, z0); ω u represents weight; K u represents the uth (u=1,2,...,N) collection sub-cube (x u ,y u ,z u ) is the true value of the channel data; N represents the number of acquisition sub-cubes in the tunnel scene; Step 6.2, calculate the Euclidean distance between each acquisition sub-cube: The user calculates the Euclidean distance between each acquisition sub-cube in the tunnel scene. The calculation formula is as follows: Where, d uv represents the Euclidean distance between the u-th acquisition sub-cube and the v-th acquisition sub-cube in the tunnel scene; the position coordinates of the u-th (u=1,2,...,N) acquisition sub-cube are (x u ,y u ,z u ); the position coordinates of the vth (v=1,2,...,N) acquisition sub-cube are (x v ,y v ,z v ); Step 6.3, calculate semi-variance: According to the following formula, the user calculates the semi-variance between each acquisition sub-cube; In the formula, γ uv represents the semi-variance between the u-th acquisition sub-cube and the v-th acquisition sub-cube in the tunnel scene; K v represents the νth (ν=1,2,...,N) collection sub-cube (x ν ,y ν ,z ν ) in the real value of the channel data; Step 6.4, fitting the semivariance function: The user fits the semivariance function γ according to the Euclidean distance and semivariance between each acquisition sub-cube in the tunnel scene; as shown in the following formula: γ=γ(d) Where d represents the Euclidean distance between any two sub-cubes in the tunnel scene; Step 6.5, calculate the semivariance between the uncollected sub-cube and the collected sub-cube: The user calculates the semivariance γ between the uncollected sub-cube (x0, y0, z0) and all collected sub-cubes according to the semivariance function γ in the fitted tunnel scene. i0 ; Wherein, i represents the i-th (i=1,2,...,N) acquisition sub-cube in the tunnel scene; Step 6.6, calculate the optimal weight: The user calculates the optimal weight according to the following formula Where φ represents the Lagrange multiplier; Step 6.7, calculate the estimated value of the channel data: Calculate the estimated value of the channel data in the uncollected sub-cube in the tunnel three-dimensional cube area according to the following formula: In the formula, represents the optimal weight.

Images