A method for deploying UAV base stations using reflected beams to supplement coverage

Abstract

The present invention provides a method for deploying a drone base station using a reflected beam to supplement coverage. Specifically, when the direct beam between the drone and the user is not blocked, the direct beam is used for coverage. When the direct beam is blocked, the reflected beam is used to supplement coverage, thereby giving full play to the coverage potential of the reflected beam, expanding the coverage area of ​​the drone, and effectively reducing the probability of user interruption. First, a model is built to analyze the impact of buildings on the millimeter wave propagation of drones, and the first-order reflection path with the shortest path length is selected for service. The probability density expression of the vertical projection length s of the first-order reflected beam of the service is derived, and finally the coverage rate of the drone base station is obtained. The actual propagation environment parameters of the target coverage area, such as the density and average height of buildings, are brought into the coverage rate expression. According to the relationship between the coverage rate and the drone deployment parameters, the drone deployment height and density that optimize the coverage performance are obtained, providing guidance for the actual deployment of drone base stations.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a deployment method design of an unmanned aerial vehicle (UAV) base station network. Background Art

[0002] Future networks will face significant challenges in addressing user demands for high coverage and capacity. Existing terrestrial cellular networks suffer from long deployment cycles and high costs, making them difficult to meet the demands of highly dynamic scenarios. With significant advancements in unmanned aerial vehicle (UAV) technology, such as improved power management, increased payload capacity, and longer flight endurance, drones can be used for a variety of purposes, including filming, surveillance, transportation, and communications. In terms of communications, due to their high dynamics and line-of-sight transmission links between drones and ground users, drones can serve as temporary base stations to offload traffic from hotspots with high data traffic demands (such as large-scale events, concerts, and sporting events), or provide temporary coverage in areas with poor ground coverage caused by accidents or disasters. Furthermore, drone base stations are potential candidates for providing ubiquitous connectivity in remote areas lacking traditional cellular infrastructure. In scenarios where blind spots and hotspots are being addressed, temporary deployment of drone base stations is more efficient and cost-effective than installing traditional base stations.

[0003] The rapid development of mobile internet and smart devices has led to explosive growth in mobile data services, with data traffic increasing exponentially. Future networks will face significant challenges, including greater capacity and lower latency. Millimeter-wave (mmWave) communications are a key technology for fifth-generation mobile communications, effectively increasing link capacity by leveraging the vast bandwidth of the mmWave range. A significant challenge in implementing mmWave communications is path loss. To compensate for the significant path loss associated with mmWave transmission, mmWave base stations typically employ large-scale antenna arrays for narrow-beam transmission, effectively concentrating transmission energy in a specific area or direction. However, the directional nature of mmWave transmission is highly sensitive to obstruction, which can even lead to connection interruptions. This presents new challenges for mmWave communications.

[0004] Before deploying drones, it is important to first use theoretical modeling to predict and evaluate the system performance after deployment. Then, the deployment parameters are adjusted to find the optimal deployment parameters for the system performance after deployment. This is very instructive for the deployment of drones. Currently, the following methods are used to analyze the coverage performance of drone base stations:

[0005] Because drones can be deployed in the air, they can achieve a high probability of Loss of Sight (LOS). One analytical approach assumes that drones and users are always connected via direct links. However, in dense urban environments, buildings obstruct the coverage of large areas on the ground that are beyond the reach of drone LOS links. Therefore, directly assuming that drones and users can connect via LOS links is unreasonable. Another analytical approach considers the blocking effect of buildings, assuming that users experience a user outage if the direct link is blocked. This assumption fails to account for the coverage potential of reflected beams. In fact, millimeter waves have excellent reflection properties, and the reflected beam power is far greater than the noise level. If the obstacle is of appropriate size, the received power of the reflected beam in an indirect link is very close to the received power of the direct free-space link at the same link distance, demonstrating the feasibility of using reflected beams to serve users. Therefore, theoretical analysis before actual deployment should consider both the blocking and reflection effects of buildings on the beam to ensure a more accurate analytical model. Summary of the Invention

[0006] The present invention proposes a method for deploying UAV base stations using reflected beams to supplement coverage, giving full play to the coverage potential of reflected beams. It considers using reflected beams to supplement coverage of blind spots that cannot be reached by direct beams, and establishes a mathematical model to analyze the coverage performance of UAV base stations in this scenario. According to the relationship between coverage rate and deployment parameters, the UAV deployment height and density that optimize coverage performance are obtained.

[0007] The present invention proposes a method for deploying a UAV base station using reflected beams for supplementary coverage, the method comprising the following steps:

[0008] Step 200: Establish a distribution model of drones and buildings and a propagation environment model of the target area that requires spatial coverage of drone base stations, and analyze the impact of buildings on drone millimeter wave propagation.

[0009] The impact of buildings on the millimeter wave propagation of drones mainly includes three aspects: the blocking effect of buildings on direct beams, the reflection effect of buildings on direct beams, and the blocking effect of buildings on reflected beams. The distribution of drones is modeled by a Poisson point process distributed at the same height H, with a density of λ. A building can be represented by a four-tuple {C = (x, y), l, w, h, θ}, where (x, y) are the coordinates of the center of the bottom surface of the building, l represents the length of the long side of the bottom surface of the building, w represents the length of the wide side of the bottom surface of the building, h is the height of the building, and θ is defined as the counterclockwise angle between the vector from the vertical projection point of the drone to the typical user and the l side of the building. Assume that (x, y) obeys a Poisson point process (PPP) on a two-dimensional plane, the building density is μ, and the length, width and height of the building independently obey a specific distribution, and the probability density functions are f respectively. L (l), fW (w) and f H (h), θ is uniformly distributed in (0,π].

[0010] The number of buildings that block the direct beam between the drone and the user is N LB The density function of is:

[0011]

[0012] Where η represents the probability that a building does not block the direct beam in three-dimensional space, given that it blocks the vertical projection of the direct path from the drone to the user on a two-dimensional plane. It can be expressed as calculate. By the formula Calculation, D represents the horizontal distance from the user to the drone, E l and E w are the mean values ​​of the length and width of the building respectively.

[0013] Buildings block direct beams and generate reflected beams, which cause energy loss during the reflection process. Therefore, we do not consider higher-order reflections and scattering effects. Because buildings vary in location and orientation, different reflected beams are formed. To better describe the reflection process, we assume that the mechanism of reflected beam formation follows the specular reflection theorem, which states that the angle of incidence equals the angle of reflection. The ellipse model can well characterize the first-order reflection path in P2P links. It is based on the fact that, given a transmitting point and a receiving point, a first-order reflection path is completely determined by its reflection point. Given the length s of the first-order reflection path, all possible reflection points form an ellipse with the transmitting and receiving points as foci.

[0014] Based on the elliptical model, a building is fixed towards θ, and the vertical projection length of the first-order reflection path is between D and s. The number of first-order reflection paths N is R (s) is a non-homogeneous Poisson distributed random variable with an intensity function of

[0015]

[0016] In addition to blocking direct beams, buildings may also block reflected beams. A reflection path can be divided into two sections: from the transmitting point to the reflection point and from the reflection to the receiving point. If one of the two is blocked, the reflection path is blocked. Generally speaking, there is a certain correlation between whether the two paths are blocked and the amount of blockage, especially when the two paths are close to each other. The correlation will be stronger, but there are literatures that prove that ignoring the path correlation in this case will not cause too much error in the results, especially when the size of the building is much smaller than the length of the first-order reflection path, the impact of this correlation on the results is even smaller, so we assume that the two links are independent. Then in three-dimensional space, given a building orientation θ and the vertical projection length s of the first-order reflection path, the number of buildings that can block the first-order reflection path is a Poisson distributed random variable with a mean of

[0017] E NRB (s,θ)=(1-η)μS RB (s,θ) / π (3)

[0018] Among them, S RB (s,θ) can be calculated using the following formula:

[0019]

[0020] Therefore, the probability that a first-order reflection path of length s is not blocked is Therefore, use The conditional probability of diluting the non-homogeneous Poisson distribution N R (s) can be obtained: Fixed building orientation θ, the number of first-order reflection paths N that exist and are not blocked with vertical projection lengths between D and s PR (s) is a non-homogeneous Poisson distributed random variable, and its intensity function is:

[0021]

[0022] Step 210: Use the first-order reflection path to supplement coverage. When the direct beam between the UAV and the user is blocked, consider using the shortest first-order reflection beam for coverage and derive the probability density expression of the vertical projection length s of the service first-order reflection beam.

[0023] The shortest length of the first-order reflection path S>s, which means N PR (s) = 0, so the complementary cumulative distribution function (CCDF) of the shortest first-order reflection path length S is derived as follows:

[0024]

[0025] Among them E NPR (s) represents N PR The mean of (s), that is

[0026]

[0027] The probability density function (PDF) of the shortest first-order reflection path length s can be calculated by the following formula:

[0028]

[0029] Step 220: Establish a beam channel model from the drone to the user. Based on the influence of buildings on the drone beam and the probability density expression of the vertical projection length s of the first-order reflection beam of the service, obtain the coverage performance of the drone base station. According to the changes in coverage performance with the deployment height, deployment density and building environmental parameters of the drone base station, obtain the optimal drone base station group deployment plan.

[0030] Assuming that the UAV beam is affected by small-scale fading and propagation loss, when the user is covered by the direct beam, the signal-to-noise ratio expression on the user side is: in, is the large-scale fading index, α is the small-scale fading coefficient, and obeys the Gaussian distribution CN(0,1). When the user is covered by the first-order reflected beam, the signal-to-noise ratio expression on the user side is σ is the reflection loss.

[0031] According to the influence of buildings on the drone beam, the number of buildings N that can actually block the direct beam between the drone and the user LB Is a parameter The probability that the direct beam between the user and the drone is not blocked is The probability of NLOS is The coverage rate is defined as the probability that the user's signal-to-interference-plus-noise ratio (SINR) value is greater than the threshold T. When the threshold is T, the coverage rate is:

[0032]

[0033] Where P(SNR LOS >T) can be calculated by the following formula:

[0034]

[0035] P(SNR Ref >T) can be calculated by the following formula:

[0036]

[0037] Obtain the propagation environment parameters of the target coverage area (such as the average building height and density), substitute them into the coverage rate expression, and draw a graph showing the relationship between the coverage rate and the drone deployment parameters (drone deployment height and density). Find the maximum coverage rate from the graph. The deployment parameters corresponding to the maximum coverage rate are the optimal deployment plan for the target area.

[0038] Beneficial effects

[0039] This paper addresses the coverage blind spots of direct beams in drone cellular networks by proposing a method for deploying drone base stations that utilizes reflected beams for supplemental coverage. By utilizing reflected beams to provide supplemental coverage in blind spots not reached by direct beams, this method leverages the reflected beam's potential, expands the drone's coverage area, and effectively reduces the probability of user interruptions. A mathematical model is developed to analyze the coverage performance of drone base stations in this scenario. The actual propagation environment parameters of the target coverage area are incorporated into the coverage expression. Based on the relationship between coverage and deployment parameters, the optimal drone deployment height and density are determined, providing guidance for actual drone base station deployment. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a schematic diagram of an application scenario of a drone base station deployment using reflected beams to supplement coverage of the present invention;

[0041] Figure 2 It is a flowchart of the algorithm implementation of the present invention;

[0042] Figure 3 This is the relationship between SINR coverage and drone height change

[0043] Figure 4 This is the relationship between SINR coverage and drone density; DETAILED DESCRIPTION

[0044] This invention aims to address the coverage blind spots of the direct beam of the drone cellular network and proposes a method for deploying drone base stations that uses reflected beams to supplement coverage. In dense urban areas, buildings cause serious obstruction to millimeter wave beams, such as the Figure 1 As shown in the figure, the direct beam from the drone to the user is blocked, causing the user to be in the coverage blind spot of the drone's direct beam. At this time, deploying an additional drone to serve this blind spot can solve the problem of the direct beam not being covered, but it also brings about an increase in cost. However, buildings generate abundant reflections while blocking the beam, as shown in the following figure. Figure 1 The first-order reflection path 1 and the first-order reflection path 2 in the image are just aimed at the user and can provide services to the user. This involves the problem of selecting the first-order reflection path. We assume that the drone selects the first-order reflection path with the shortest path length to provide services to the user.

[0045] Consider using reflected beams to supplement coverage in blind spots beyond the reach of direct beams. This leverages the reflected beam's potential, expands the drone's coverage area, and effectively reduces user interruption probability. A mathematical model is developed to analyze the coverage performance of drone base stations in this scenario. The actual propagation environment parameters of the target coverage area are incorporated into the coverage ratio expression. Based on the relationship between coverage ratio and deployment parameters, the optimal drone deployment height and density are determined, providing guidance for actual drone base station deployment.

[0046] The algorithm flow of this case is as follows Figure 2 As shown, the specific implementation steps are:

[0047] Step 300, establish the distribution model of drones and buildings and the propagation environment model of the target area that requires spatial coverage of drone base stations, and analyze the impact of buildings on the millimeter wave propagation of drones. The impact of buildings on the millimeter wave propagation of drones mainly includes three aspects: the blocking effect of buildings on direct beams, the reflection effect of buildings on direct beams, and the blocking effect of buildings on reflected beams. Use the Poisson point process distributed at the same height H to model the distribution of drones, with a density of λ. A building can be represented by a four-tuple {C = (x, y), l, w, h, θ}, where (x, y) is the coordinate of the center of the bottom surface of the building, l represents the length of the long side of the bottom surface of the building, w represents the length of the wide side of the bottom surface of the building, h is the height of the building, and θ is defined as the counterclockwise angle between the vector from the vertical projection point of the drone to the typical user and the l side of the building. Assume that (x, y) obeys the Poisson point process (PPP) on a two-dimensional plane, the building density is μ, and the length, width and height of the building independently obey a specific distribution, and the probability density function is f respectively. L (l), f W (w) and f H (h), θ is uniformly distributed in (0,π], and the number of buildings N that block the direct beam between the drone and the user is derived. LB The density function of .

[0048] Buildings block direct beams and generate reflected beams, which cause energy loss during the reflection process, so we do not consider high-order reflection and scattering effects. Because the location and orientation of buildings are different, different reflected beams will be formed. Using the elliptical model, we deduce the number of first-order reflection paths N whose vertical projection length is between D and s for a fixed building orientation θ. R (s) intensity function.

[0049] In addition to blocking direct beams, buildings may also block reflected beams. A reflection path can be divided into two sections: from the transmitting point to the reflection point and from the reflection point to the receiving point. If one of the two is blocked, the reflection path is blocked. Ignoring the correlation between the two links and assuming that they are independent of each other, derive the mean number of buildings that can block the first-order reflection path in three-dimensional space, given a building orientation θ and a vertical projection length s of the first-order reflection path. Then derive the number N of first-order reflection paths that exist and are not blocked with a vertical projection length between D and s for a fixed building orientation θ. PR (s) intensity function.

[0050] Step 310: Use the first-order reflection path to supplement coverage. When the direct beam between the drone and the user is blocked, consider using the first-order reflection beam to provide service to the user. When multiple first-order reflection paths are available, select the first-order reflection beam with the shortest path length for coverage. Derive the probability density expression of the vertical projection length s of the service first-order reflection beam.

[0051] Step 320: Establish a beam channel model from the drone to the user. Based on the influence of buildings on the drone beam and the probability density expression of the vertical projection length s of the first-order reflection beam of the service, obtain the coverage performance of the drone base station. According to the changes in coverage performance with the deployment height, deployment density and building environmental parameters of the drone base station, obtain the optimal drone base station group deployment plan.

[0052] Assuming that the UAV beam is affected by small-scale fading and propagation loss, when the user is covered by the direct beam, the signal-to-noise ratio expression on the user side is: in, is the large-scale fading index, α is the small-scale fading coefficient, and obeys the Gaussian distribution CN(0,1). When the user is covered by the first-order reflected beam, the signal-to-noise ratio expression on the user side is σ is the reflection loss.

[0053] Coverage is defined as the probability that a user's signal-to-interference-plus-noise ratio (SINR) exceeds a threshold value, T. Based on the impact of buildings on drone beams, the coverage is derived when the threshold value is T. The propagation environment parameters of the target coverage area (such as average building height and density) are obtained and substituted into the coverage expression. A graph is plotted showing how coverage varies with drone deployment parameters (drone deployment height and density). The maximum coverage value is found in the graph. The deployment parameters corresponding to the maximum coverage value are the optimal deployment solution for the target area.

[0054] The simulation results are shown in the attached Figure 3 and attached Figure 4 shown.

[0055] Attachment Figure 3Given the SINR threshold of -3dB and the drone density λ = 0.00006 / m 2 , user density is γ=0.06 / m 2 , building density is μ=0.0005 / m 2 The following graph shows the relationship between SINR coverage and drone height when the average building height is 40m and 80m, respectively. As can be seen, coverage initially increases with increasing drone altitude, but then decreases. This is because when the drone's altitude is initially low, increasing altitude significantly improves the probability of loss of service (LOS) between the drone and the user. However, when the drone's altitude is very high, the LOS probability is already high. Further increases in altitude slow the increase in LOS probability, but the signal experiences significant increases in large-scale fading, causing SINR to decrease with increasing altitude. The first gain in the figure represents drone deployment gain. Considering only direct beams, a drone altitude of zero is equivalent to a ground base station. Adjusting the drone's altitude to maximize coverage yields a 159% gain compared to ground base stations. This demonstrates the advantages of using drones for user coverage in dense urban environments. The second gain is reflected beam gain, which maximizes coverage by optimizing the drone's altitude after considering the first-order reflected beam. This gain is 22% compared to considering direct beams.

[0056] Attachment Figure 4 Given the SINR threshold of -3dB, the drone height H is 100m, and the user density is γ = 0.06 / m 2 , building density is μ=0.0005 / m 2 , when the average height of the buildings is 40m and 80m respectively, the relationship between SINR coverage and drone density is shown. It can be seen that the greater the drone density, the greater the SINR coverage. This is because the greater the drone density, the smaller the horizontal distance between the user and the associated drone, so the probability of the direct beam being blocked is smaller. Even if it is blocked, the length of the reflection path is relatively shorter, and the large-scale fading experienced will also be smaller, so a higher coverage rate can be achieved. However, the dense deployment of drones will also bring about a sharp increase in deployment costs. In addition, we can also clearly see that the introduction of the first-order reflection beam has brought a significant improvement in coverage. For example, when the drone density is 0.00011 / m 2When the average building height is 40m, the gain from the first-order reflection beam is 39%, while when the average building height is 80m, the gain is 84%. This shows that the introduction of the first-order reflection beam can achieve high coverage even when the drone density is low, and the first-order reflection beam can save costs without affecting performance. We can also see that the gain percentage from the first-order reflection beam increases with the average building height and the more obstructed the environment.

Claims

1. A method for deploying a UAV base station using reflected beams to supplement coverage, characterized in that: include: When the direct beam between the UAV and the user is not blocked, the direct beam is used for coverage. When the direct beam is blocked, the first-order reflected beam is used to provide supplementary coverage for the blind area that the direct beam cannot reach. When there are multiple available first-order reflection paths, the first-order reflected beam with the shortest path length is selected for coverage. The probability density expression of the vertical projection length s of the service first-order reflected beam is derived as follows: The distribution function of the vertical projection length s of the first-order reflection beam of the service is obtained to expand the coverage area of ​​the drone and effectively reduce the probability of user interruption. A mathematical model based on the blocking and reflection effects of buildings on the beam and the distribution function of the vertical projection length s of the first-order reflection beam of the service is established to analyze the coverage performance of the drone base station in this scenario. The density and average height of the buildings in the target coverage area are substituted into the coverage expression. According to the relationship between the coverage rate and the variation of the drone deployment parameters, the drone deployment height and density that optimize the coverage performance are obtained, providing guidance for the actual deployment of drone base stations.

2. The method according to claim 1, characterized in that First, it is necessary to establish a distribution model of drones and buildings, and model and analyze the impact of buildings on the millimeter wave propagation of drones; the impact of buildings on the millimeter wave propagation of drones mainly includes three aspects: the blocking effect of buildings on direct beams, the reflection effect of buildings on direct beams, and the blocking effect of buildings on reflected beams; the distribution of drones is modeled by a Poisson point process distributed at the same height H, with a density of λ, and a four-tuple {C = (x, y), l, w, h, θ} to represent the building, where (x, y) is the coordinate of the center of the bottom surface of the building, l represents the length of the long side of the bottom surface of the building, w represents the length of the wide side of the bottom surface of the building, h is the height of the building, and θ is defined as the counterclockwise angle between the vector from the vertical projection point of the drone to the typical user and the l side of the building; assuming that (x, y) obeys a Poisson point process (PPP) on a two-dimensional plane, the building density is μ, and the length, width and height of the building independently obey a specific distribution, and the probability density functions are f L (l), f W (w) and f H (h), θ is uniformly distributed in (0,π]; then the number of buildings that block the direct beam between the drone and the user is N LB The density function of is: Where η represents the probability that a building does not block the direct beam in three-dimensional space, given that it blocks the vertical projection of the direct path from the drone to the user on a two-dimensional plane. It can be expressed as calculate; By the formula Calculation, D represents the horizontal distance from the user to the drone, E l and E w Represent the mean of the length and width of the building respectively; when the building is facing θ, the number of first-order reflection paths N that exist and are not blocked and have a vertical projection length between D and s is PR (s) is a non-homogeneous Poisson distributed random variable, and its intensity function is: Among them, S RB (s,θ) represents the area of ​​the center of the building on the two-dimensional plane that can block the first-order reflection path. The impact of buildings on the UAV beam is analyzed based on the distribution of the number of buildings that block the UAV's direct beam and the intensity function of the number of available first-order reflection paths.

3. According to the method of claim 1 or 2, a UAV-to-user beam channel model is established. Based on the influence of buildings on the UAV beam and the probability density expression of the vertical projection length s of the first-order reflection beam of the service, the coverage performance expression of the UAV base station is derived as follows: ， Where P(SNR LOS >T) can be calculated by the following formula: P(SNR Ref >T) can be calculated by the following formula: Obtain the propagation environment parameters of the target coverage area, substitute them into the coverage rate expression, and draw a graph showing how the coverage rate varies with the deployment parameters of the drone. Find the maximum coverage rate from the graph, and the deployment parameters corresponding to the maximum coverage rate are the optimal deployment plan for the target area.

Images