A UAV flying network transmission method for channel sorting by predicting delay

Abstract

The present invention discloses a UAV flying network transmission method for selecting channels according to estimated delay, belonging to the field of UAV self-organizing network; the method comprises the following steps: firstly, dividing a UAV group into a master machine and a slave machine, the master machine receiving the same data packet at a specified frequency, detecting the transmission rates of different channels in real time and sorting the channels in descending order according to the propagation rate; then, when receiving multiple data packets, the master machine selects channels of corresponding levels for different transmission data by priority judgment; the estimated delay is measured according to the real-time environment such as the used channel, so as to define a time gap for the data transmission of the slave machine, if the transmission time exceeds the time gap, it will be forcibly cut off, and the timed data transmission will be allocated to a channel with a lower rate to continue the transmission, the present invention avoids the conflict of channels through real-time detection and dynamic adjustment, can effectively utilize the available spectrum resources, and improve the transmission efficiency and capacity of the whole network.

Description

Technical Field

[0001] The present invention belongs to the field of unmanned aerial vehicle self-organizing networks, and in particular relates to a unmanned aerial vehicle flying network transmission system that predicts delays for channel sorting. Background Art

[0002] Compared with traditional communication networks, drone self-organizing networks have the characteristics of free node movement, frequent topology changes, and freedom to enter and leave the network. Traditional communication protocols cannot meet the high-speed movement of drone self-organizing network nodes due to their shortcomings such as easy to cause network node congestion, large control overhead, and long network latency. In complex situations such as battlefields, some important information delays are even required to reach 2ms, and the changes in high-speed moving nodes will inevitably cause network node congestion or information conflicts for a period of time. Therefore, it is necessary to classify data and transmit information of different importance in a hierarchical manner.

[0003] How to select priorities and reasonably allocate priorities to data determines how the drone network can work normally and complete tasks, so the priority setting is particularly important.

[0004] The priority protocol is essentially a media access control (MAC) protocol. Currently, the research directions of related MAC protocols are mainly divided into time division multiple access (TDMA), carrier sense multiple access / collision avoidance (CSMA / CA) and statistical priority-based multiple access (SPMA).

[0005] As a time-based resource division protocol, TDMA cannot guarantee high priority and low latency. CSMA / CA is a random competition protocol and cannot effectively differentiate services. Although SPMA is also a priority access protocol, it still cannot avoid the problem of increased collision probability of high priority data when the maximum channel capacity is exceeded.

[0006] The main modes of existing drone mesh network transmission signals usually include fixed priority and dynamic priority.

[0007] Fixed priority mode: In this mode, each node or device is assigned a fixed priority, and data packets are transmitted according to the node's priority. Usually, the priority in this mode is pre-set during network configuration or node deployment and will not be adjusted according to real-time environmental conditions.

[0008] Dynamic priority mode: In this mode, the priority of nodes is dynamically adjusted according to real-time network conditions and environmental changes. For example, if the signal quality of a node suddenly drops, its priority may be lowered so that other nodes can transmit data faster.

[0009] Dynamic Spectrum Management (DSM) is an intelligent spectrum resource management method that can dynamically adjust and optimize the allocation and use of spectrum resources according to the real-time radio environment and usage requirements. The goal of DSM technology is to improve the utilization of spectrum resources and the performance of communication systems while ensuring fairness and reliability among different users and services.

[0010] DSM technology is widely used and can be used in various wireless communication systems, such as cellular mobile communications, wireless local area networks, satellite communications, etc. In drone cluster communications, DSM technology can play an important role in improving the communication efficiency and stability of drone clusters and promoting the development and application of drone cluster technology.

[0011] However, DSM technology also has the following problems:

[0012] 1) High complexity: Dynamic spectrum management requires real-time monitoring and analysis of the radio environment, involving a large amount of data collection and processing. This makes the design and implementation of the DSM system relatively complex, requiring high-performance hardware and complex algorithm support.

[0013] 2) High real-time requirements: DSM technology needs to make decisions and adjustments based on real-time spectrum usage and demand, so the system is required to have high real-time and response speed. In a rapidly changing communication environment, if the system cannot make accurate decisions in time, it may affect the stability and performance of the communication.

[0014] 3) Security issues: Since DSM technology involves the dynamic allocation and management of spectrum resources, there may be some security vulnerabilities and attack risks. For example, malicious users may obtain additional spectrum resources through interference or deception, thus affecting the normal operation and security of the system.

[0015] 4) Multi-user collaboration: In scenarios where multiple users or services share spectrum resources, it is a challenge to achieve fair and efficient spectrum allocation and management. DSM technology needs to comprehensively consider the needs and priorities of different users to ensure fair allocation and effective use of spectrum resources. Summary of the invention

[0016] The present invention proposes a UAV flying network transmission method that predicts the delay for channel sorting, adopts dual priority judgment, and divides the priorities of different data in more detail. While saving control overhead, it can ensure efficient and reliable data transmission and provide effective protection for UAV self-organizing networks.

[0017] The method for transmitting data over a drone network by predicting the delay and performing channel sorting has the following specific steps:

[0018] Step 1: For the network composed of drone swarms, select the drone located at the center of the network as the master drone, and the rest as slave drones;

[0019] Step 2: The master machine sequentially uses different channels to send test data packets of known size to each slave machine. Each slave machine measures the transmission rate of each channel through the wireless communication test equipment and returns it to the master machine.

[0020] Transmission rate = packet size\transmission time;

[0021] Step 3: The master machine takes the average value of all transmission rates corresponding to each channel as the transmission rate corresponding to the channel, and sorts the corresponding channels from large to small according to the transmission rate for use in priority judgment.

[0022] Step 4: Initially set the "absolute priority" of the drone group, from low to high, to 1, 2, 3...n gears, corresponding to channels with low to high transmission rates;

[0023] The priority levels of the drone swarm are divided according to the importance of the transmitted data, with n being the highest level and allocating the channel with the highest current transmission rate.

[0024] Step 5: Multiple slaves send data to the same master at the same time. The transmitted data is automatically divided into corresponding gears according to the "absolute priority" set in advance, and is allocated to the corresponding transmission channel.

[0025] The data sent includes: control instructions for the drone’s flight, attitude, and speed; sensor data including images, videos, temperature, humidity, and air pressure; the drone’s current location, altitude, and speed; and information about the mission being performed;

[0026] Step 6: Calculate the conditional estimated delay of each channel corresponding to the transmission data, and set the time gap according to the delay;

[0027] The delay of drone communication consists of five aspects: communication link delay, data processing delay, signal propagation delay, network congestion and interference delay.

[0028] The calculation for signal propagation delay is as follows:

[0029] s ij is the distance between the i-th UAV and the j-th UAV, c air is the propagation speed of electromagnetic waves in air;

[0030] The calculations for communication link delay and data processing delay are as follows:

[0031] T is the time period measured before the UAV takes off; f(t) is the fitted straight line function of the predicted delay with respect to time; t0 is the starting measurement time.

[0032] The calculation of network congestion and interference delay is as follows:

[0033]

[0034] t future is the future time to be predicted, t lim The short time required to adjust the accuracy is t lim <<0.1s, Δt is a very short period of time, τ future is the final predicted delay at the future time.

[0035] Intermediate variables b() is the function curve obtained by approximate fitting;

[0036] So the estimated delay is calculated as: T total =T1+T2+T3;

[0037] Step 7: Determine whether the transmission time of each data exceeds the time gap; if so, the data is forcibly cut off, and the overdue data is assigned to a channel with a lower rate to continue transmission; if there is data being transmitted on the lower rate channel, the data is assigned to the channel used by the overdue data to continue transmission. Otherwise, normal transmission is OK.

[0038] The advantages of the present invention are:

[0039] 1) A drone flying network transmission method that predicts delay and selects channels, which is real-time and adaptive: the drone detects the transmission rate of different channels in real time, determines the expected delay according to the real-time environment, and dynamically adjusts the time interval and channel selection of data transmission, thereby ensuring the efficiency and stability of data transmission.

[0040] 2) A method for drone flying network transmission that predicts delays and selects channels, and a multi-level control mechanism: channels of corresponding levels are selected for different transmission data through priority judgment, and after the data transmission times out, the data will be allocated to the appropriate channel according to the rate, thereby maximizing the guarantee of smooth data transmission.

[0041] 3) A UAV flying network transmission method that predicts delays and performs channel sorting, with high resource utilization: through real-time detection and dynamic adjustment, it can effectively utilize available spectrum resources and improve the transmission efficiency and capacity of the entire network. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a flow chart of a method for transmitting through a UAV network by predicting the delay for channel sorting according to the present invention; DETAILED DESCRIPTION

[0043] In order to clearly and completely present the advantages, technical features and invention purposes of the present invention, the technical details of the present invention will be described in detail and completely in combination with the accompanying drawings and embodiments. It should be noted that the enumerated embodiments belong to a part of the present invention, but not the entire content of the invention. Other embodiments obtained by other persons in the field without making innovative breakthroughs are all within the scope of protection of the present invention.

[0044] The present invention discloses a method for unmanned aerial vehicle (UAV) flying network transmission with estimated delay and channel sorting, comprising: first, the UAV receives the same data packet at a specified frequency, detects the transmission rates of different channels in real time and sorts the channels in descending order according to the transmission rate; then, when receiving multiple data packets, the UAV selects channels of corresponding levels for different transmission data by priority judgment; the estimated delay is determined according to the real-time environment such as the used channel, so as to define a time gap for the data transmission of the UAV, if the transmission time exceeds the time gap, it will be forcibly cut off, and the timed data transmission will be allocated to a channel with a lower rate to continue transmission, through real-time detection and dynamic adjustment, thereby avoiding channel conflicts, being able to effectively utilize available spectrum resources, and improving the transmission efficiency and capacity of the entire network.

[0045] like Figure 1 As shown, the specific steps are as follows:

[0046] Step 1: For the network composed of drone swarms, select the drone located at the center of the network as the master drone, and the rest as slave drones;

[0047] Step 2: The master machine sequentially uses different channels to send fixed-size test data packets to each slave machine. Each slave machine measures the transmission rate of each channel through the wireless communication test equipment and returns it to the master machine.

[0048] Transmission rate = packet size\transmission time;

[0049] Step 3: The master machine takes the average value of all transmission rates corresponding to each channel as the transmission rate corresponding to the channel, and sorts the corresponding channels from large to small according to the transmission rate for use in priority judgment.

[0050] Step 4: Initially set the "absolute priority" of the drone group, from low to high, to 1, 2, 3...n gears, corresponding to channels with low to high transmission rates;

[0051] The priority levels of the drone swarm are divided according to the importance of the transmitted data, with n being the highest level and allocating the channel with the highest current transmission rate.

[0052] Step 5: Multiple slaves send data to the same master at the same time. The transmitted data is automatically divided into corresponding gears according to the "absolute priority" set in advance, and is allocated to the corresponding transmission channel.

[0053] The data sent includes: control instructions for the flight, attitude and speed of the drone; sensor data including images, videos, temperature, humidity and air pressure; the current position, altitude and speed of the drone; and information about the mission being performed; for example: sensor data - from various sensors on the drone, such as cameras, radars, meteorological sensors, etc., are assigned to smaller gears corresponding to smaller "absolute priorities". Remote control instructions - used to control the flight, attitude, speed and other operations of the drone, are assigned to larger gears. When the drone fleet is working, the data transmitted by the drone will be automatically divided into the corresponding gears according to the "absolute priority" gears set in advance.

[0054] Step 6: Calculate the conditional estimated delay of each channel corresponding to the transmission data, and set the time gap according to the delay;

[0055] The delay of drone communication consists of five aspects: communication link delay, data processing delay, signal propagation delay, network congestion and interference delay.

[0056] The calculation for signal propagation delay is as follows:

[0057] s ij is the distance between the i-th UAV and the j-th UAV, c air is the propagation speed of electromagnetic waves in air;

[0058] The calculations for communication link delay and data processing delay are as follows:

[0059] T is the time period measured before the UAV takes off; f(t) is the fitted straight line function of the predicted delay with respect to time; t0 is the starting measurement time.

[0060] The calculation of network congestion and interference delay is as follows:

[0061]

[0062] t futureis the future time to be predicted, t lim The short time required to adjust the accuracy is t lim <<0.1s, Δt is a very short period of time, τ future is the final predicted delay at the future time.

[0063] Intermediate variables b() is the function curve obtained by approximate fitting;

[0064] So the estimated delay is calculated as: T total =T1+T2+T3;

[0065] Step 7: Determine whether the transmission time of each data exceeds the time gap; if so, the data is forcibly cut off, and the overdue data is assigned to a channel with a lower rate to continue transmission; if there is data being transmitted on the lower rate channel, the data is assigned to the channel used by the overdue data to continue transmission. Otherwise, normal transmission is OK.

[0066] Example:

[0067] 1. First, when the drone swarm is working, it conducts a channel rate detection test every 30 seconds; the selected master uses different channels to send known fixed-size test data packets to the remaining slaves in turn, and uses wireless communication test equipment such as Iperf to measure the transmission rates of different channels and return them to the master. The master takes the average of all transmission rates corresponding to each channel as the transmission rate corresponding to the channel, and sorts the corresponding channels by transmission rate for priority determination.

[0068] For example, if the master uses a 1.3Hz channel to send data to all slaves, the average speed of the master sending data to different slaves is taken as the average rate of the 1.3Hz channel. Then, the master uses a 1.4Hz channel to send data to all slaves, and the average speed of the master sending data to different slaves is taken as the average rate of the 1.4Hz channel. Finally, the average rates of all channels are calculated and sorted.

[0069] 2. When the main machine receives data from multiple slave machines, it can set the corresponding channels from low to high to 1, 2, 3...n gears of "absolute priority" according to the transmission rate from small to large; when the slave machine sends data, it will automatically match the corresponding gear and transmission channel according to the "absolute priority";

[0070] Alternatively, the transmission channel is selected by judging "two priorities"; the "two priorities" include "preemption priority" and "response priority".

[0071] Set the "preemption priority" level to a0, a1, a2...an (set the size of n according to actual needs), an is the highest priority, corresponding to the transmission channel with the highest allocation rate, which means that the channels in use at a0, a1, a2...a(n-1) can be preempted. The preempted data transmission will be postponed in turn, and the channel with a lower rate will be allocated. When the drone group is working, the data is allocated to the gear according to the urgency and importance, and the more important the data transmission is, the higher the gear is.

[0072] Set the "response priority" level to b0, b1, b2...bm (set the size of m according to actual needs), bm is the highest priority, corresponding to the transmission channel with the highest allocation rate. Unlike "preemption priority", "response priority" cannot interrupt data transmission, and is mainly used to deal with situations where multiple data are transmitted at the same time. The higher the gear of "response priority", the faster the transmission channel can be allocated. When the drone swarm is working, the data is allocated to gears according to the degree of urgency and importance, and the more important the data transmission is, the higher the gear is.

[0073] If five slave machines transmit data to the master machine at the same time, the data S with the highest "response priority" will be selected, and the transmission channel with the highest allocation rate will be selected. If data T is already being transmitted in the channel, the "preemption priority" will directly interrupt the original data T transmission and data S will be transmitted; at the same time, data T will be allocated a channel with a lower rate for transmission.

[0074] 3. After the gears are allocated, the conditional estimated delay of each channel corresponding to the transmitted data is calculated, and the time gap is set according to the size of the delay;

[0075] The specific components of each delay are shown in the following table:

[0076]

[0077]

[0078] The estimated delay is calculated as: T total =T1+T2+T3;

[0079] The first category: the exact value can be calculated by mathematical knowledge

[0080] (1) Delay type: signal propagation delay

[0081] (2) Calculation method:

[0082] Based on known physical laws, the propagation speed of electromagnetic waves in the medium is calculated and then the calculation is performed.

[0083] For example, if the distance between drone No. i and drone No. j is sij , the propagation speed of electromagnetic waves in air c air ; Then, the calculation formula for the delay is:

[0084]

[0085] The second category - can be approximately measured by experiment:

[0086] (1) Delay type:

[0087] Including sensor delay, transmission data protocol processing time, data encoding, communication connection handshake delay, communication connection protocol processing delay

[0088] (2) Calculation method:

[0089] Since the above delays are all determined by the drone hardware, a period of measurement is performed before takeoff, i.e. prediction. After obtaining the latest data, the average value is obtained as the calculation value in the next environment;

[0090] Taking sensor delay as an example, before takeoff, measure the sensor sensitivity data N times (N is extremely large, the larger the better) within T seconds (T is approximately 1 second).

[0091] Step 1: Eliminate bad data

[0092] The data is preprocessed using the least squares method. Let t0 be the starting measurement time, T be the ending measurement time, and N represent dividing the time period [t0, t0+T] into N equal segments. Then the following matrix A can be obtained:

[0093]

[0094] The first column of the matrix A It is N equal parts of time within [t0, t0+T], representing the time of each measured moment, and the other column is all 1.

[0095] Assume τ i (i=1,2,......N) is the delay obtained by reducing the above N equal parts of time. The above delay is written into the column vector b:

[0096]

[0097] Each element represents the approximate delay obtained by measurement;

[0098] Let f(t) be the function of the predicted delay with respect to time, which is a fitted straight line, k is the slope of the fitted straight line, and m is the intercept of the fitted straight line. According to the least squares formula:

[0099]

[0100] f(t)=kt+m

[0101] Set each point The degree of deviation from the straight line f(t) is e i :

[0102]

[0103] Let α be the accuracy requirement, set to 0.1 seconds. If e i >α, remove the point and repeat the above process.

[0104] Step 2: Predict latency

[0105] When there are no points that can be removed, the final fitting line f(t) is obtained, and the final predicted delay is

[0106]

[0107] That is, the communication link delay and data processing delay are obtained.

[0108] Category 3 - Predictable Delay

[0109] (1) Delay type:

[0110] Network congestion, external interference

[0111] (2) Calculation method:

[0112] Measurements are made during the data transmission intervals to obtain multiple sets of data within a period of time, and the data are substituted into the GM (1,1) grey prediction model for prediction. Two parameters are set for each set of data, one is the measurement time, and the other is the measurement value; take external interference as an example. Within a period of time, the degree of external interference can be considered to vary very little, so differential equations can be used for prediction, in which the measurement time should ensure that the time interval of each time is the same.

[0113] First, measure and obtain multiple sets of data, assuming t i (i=1,2,......,N) is each measurement moment, τ i (i=1,2,......,N) is the delay measured at each measurement moment:

[0114]

[0115] The first column in matrix B is the i (i=1,2,......,N), the second column is the sum of the measured delays at each moment before the current moment i:

[0116]

[0117] Assume b i The second column of the B matrix at each moment t i The corresponding value, Assume (t i ,b i ) is the function curve approximately fitted by b(t);

[0118] Next, since the data is approximately approximate, the cumulative sequence will be approximately an exponential function (straight line). Therefore, consider using a first-order linear non-homogeneous differential equation to approximate the function. Let the function formed by b and t be b(t): Let α and u be the unknown parameters of the differential equation:

[0119]

[0120] Discretize the differential equation. Since the time interval of each measurement is the same, let the time interval of each measurement moment be diff:

[0121] diff=t i+1 -t i (2)

[0122] Discretize equation (1) to get the expression:

[0123]

[0124] Substituting (3) into (2), we get:

[0125]

[0126] Will Substituting in, we get:

[0127] The above expression is a definite linear relationship, so it can be solved by linear algebra. Substituting multiple sets of data, we can obtain the values ​​of the unknown coefficients α and u, and substitute them into equation (1) to obtain the function expression of the predicted delay: C is the constant to be determined. Substitute the initial solution (t1, τ1) again to obtain the final function expression:

[0128] Substitute the future time t to be predicted future You can make predictions:

[0129] Assume a very short time t lim (t lim <<0.1s, adjustable according to the accuracy requirements), let the derivative of b(t) be b'(t), let Δt be a very short period of time, let τ future For the predicted delay at a future time:

[0130]

[0131] t future is the future time to be predicted, t lim The short time required to adjust the accuracy is t lim <<0.1s, which is a very short period of time, τ future is the final predicted delay at the future moment; intermediate variable b() is the function curve obtained by approximate fitting;

[0132] 4. After obtaining the expected delay, set the time gap according to the delay.

[0133] It is stipulated that the transmission time of each data cannot exceed the time gap. If it exceeds, it will be forcibly cut off; the timed data transmission will be allocated to a channel with a lower rate to continue transmission.

[0134] The present invention sets two different priorities to sort the importance of information, which is different from the previous one. It rationally divides the data into "preemption priority" of urgent importance and "response priority" of general importance. The different priority settings provide more detailed division of data. The response priority ensures the normal transmission of current data and also ensures that important information can be delivered in time, while taking into account the security, stability, timeliness and effectiveness of communication.

[0135] The present invention predicts the communication delay through an algorithm, queues according to the predicted delay, and then selects the transmission channel with the best rate. Data with different delays can be reasonably allocated to channels, thereby improving communication efficiency. When calculating the delay, the possible delay is divided into signal propagation delay, data processing and transmission delay, network congestion and external interference delay, and accurate formula calculation, experimental approximate estimation measurement calculation and differential equation prediction calculation are performed respectively. The algorithm is innovated, and the possible ways of delay are more comprehensively considered, so as to better serve the channel allocation prediction and be closer to the complex situation in reality. If the data transmission times out, this solution can forcibly cut off the data and allocate the timed data transmission to a channel with a lower rate for continued transmission. This processing method can effectively deal with delays or abnormal situations that may occur during the transmission process, and ensure the completeness and stability of data transmission.

Claims

1. A method for transmitting data over a drone network by predicting the delay and performing channel sorting, characterized in that: The specific steps are as follows: Step 1: For the network composed of drone swarms, after selecting the master and slave machines, the master machine uses different channels to send test data packets to each slave machine in turn, measures the transmission rate of each channel, and sorts the corresponding channels from small to large according to the transmission rate; Step 2: Set the "two priorities" of the drone group to select the transmission channel; or, when multiple slaves send data to the same mother machine at the same time, the transmission data is automatically divided into corresponding gears according to the "absolute priority" set in advance, and allocated to the corresponding transmission channel; "Two priorities" include "preemption priority" and "response priority"; The "preemption priority" is set to a0, a1, a2...an from low to high, corresponding to the transmission channels with transmission rates arranged from small to large. The highest level an can preempt the channels in use at levels a0, a1, a2...a(n-1); the preempted data transmission is postponed in turn and allocated to a channel with a lower rate. Set the "Response Priority" level to b0, b1, b2...bm, which correspond to the transmission channels with the transmission rates sorted from small to large. Unlike the "Preemption Priority", the "Response Priority" cannot interrupt the data transmission. It is mainly used to deal with the situation where multiple data are transmitted at the same time. The higher the "Response Priority" level, the faster the transmission channel is allocated. "Absolute priority" is set to 1, 2, 3...n gears from low to high, corresponding to channels with low to high transmission rates respectively; the priority gears of the drone group are divided according to the importance of the transmitted data, n is the highest gear, and the channel with the highest current transmission rate is allocated; Specifically: When multiple slaves transmit data to the master at the same time, the data S with the highest "response priority" is selected, and the transmission channel with the highest allocation rate is selected. If data T is already being transmitted in the channel, the "preemption priority" directly interrupts the original data T transmission and transmits data S; at the same time, data T is allocated a channel with a lower rate for transmission; Step 3: Calculate the conditional estimated delay of each channel corresponding to the transmitted data, and set the time gap according to the delay; The delay of drone communication consists of five aspects: communication link delay, data processing delay, signal propagation delay, network congestion and interference delay; The calculation for signal propagation delay is as follows: s ij is the distance between the i-th UAV and the j-th UAV, c air is the propagation speed of electromagnetic waves in air; The calculations for communication link delay and data processing delay are as follows: T is the time period measured before the UAV takes off; f(t) is the fitted straight line function of the predicted delay with respect to time; t0 is the starting measurement time; The calculation of network congestion and interference delay is as follows: t future is the future time to be predicted, t lim The short time required to adjust the accuracy is t lim <<0.<<0.1s1s,τ future is the final predicted delay at a future time; Intermediate variables b() is the function curve obtained by approximate fitting; So the estimated delay is calculated as: T total =T1+T2+T3; Step 4: Determine whether the transmission time of each data exceeds the time gap; If so, the data is forcibly truncated, and the timed-out data is assigned to a channel with a lower rate for continued transmission; if there was data originally being transmitted on the lower rate channel, the data is assigned to the channel used by the timed-out data for continued transmission; otherwise, normal transmission is sufficient.

2. The method for transmitting through a UAV network by predicting the delay and performing channel sorting as claimed in claim 1, characterized in that: In the step 1, the drone located at the center of the network is selected as the master drone, and the rest are slave drones; The transmission rate of each channel measured by the master machine is as follows: First, different channels are used to send fixed-size test data packets to each slave in turn. Each slave measures the transmission rate of each channel through the wireless communication test equipment and returns it to the master. The master takes the average of all the transmission rates corresponding to each channel as the transmission rate corresponding to the channel. Transmission rate = packet size\transmission time.

3. The method for transmitting through a UAV network by predicting the delay and performing channel sorting as claimed in claim 1, characterized in that: In step 2, the data sent includes: control instructions for the flight, attitude and speed of the drone; sensor data including images, videos, temperature, humidity and air pressure; the current position, altitude and speed of the drone; and information about the mission being performed.

4. The method for transmitting through a UAV network by predicting the delay and performing channel sorting as claimed in claim 1, characterized in that: In step 3, the calculation process for the communication link delay and the data processing delay T2 is: First, before takeoff, measure data N times within T seconds to remove bad data; The data is preprocessed using the least squares method. Let t0 be the starting measurement time, T be the ending measurement time, and N be the time period [t0, t0+T] divided into N equal segments. The matrix A is obtained: The first column of the matrix A It is N equal parts of time in [t0, t0+T], representing the time of each moment of measurement; Assume τ i (i=1,2,......N) is the delay obtained by reducing the above N equal parts of time. The above delay is written into the column vector b: Each element represents the approximate delay obtained by measurement; Let f(t) be the function of the predicted delay with respect to time, which is a fitted straight line, k is the slope of the fitted straight line, and m is the intercept of the fitted straight line. According to the least squares formula: f(t)=kt+m Set each point The degree of deviation from the straight line f(t) is e i : Let α be the accuracy requirement. If e i >α, then remove the point and repeat the above process until there are no more points to remove, and get the final fitting straight line f(t), and get the final predicted delay 5. The method for transmitting over a UAV network by predicting delay and performing channel sorting as claimed in claim 1, characterized in that: In step 3, the calculation process for network congestion and interference delay T3 is: First, measure and obtain multiple sets of data. Suppose t i (i=1,2,......,N) is each measurement moment, τ i (i=1,2,......,N) is the delay measured at each measurement moment: The first column in matrix B is the i (i=1,2,......,N), the second column is the sum of the measured delays at each moment before the current moment i: Assume b i The second column of the B matrix at each moment t i The corresponding value, Assume (t i ,b i ) is the function curve approximately fitted by b(t); Next, we use a first-order linear nonhomogeneous differential equation to approximate the function. Let the function formed by b and t be b(t): Let α and u be the unknown parameters of the differential equation: Discretize the differential equation. Since the time interval of each measurement is the same, let the time interval of each measurement moment be diff: diff=t i+1 -t i (2) Discretize equation (1) to get the expression: Substituting (3) into (2), we get: Will Substituting in, we get: Solve it through linear algebra, substitute multiple sets of data, obtain the values ​​of the unknown coefficients α and u, substitute them into formula (1) to solve, and obtain the function expression of the predicted delay: C is a constant to be determined; Substitute the initial solution (t1,τ1) again to obtain the final function expression: Finally, substitute the future time t to be predicted future You can make predictions.

Images