Moving node DOA estimation method based on hello packet and inertial navigation

Abstract

The present invention discloses a method for estimating the DOA of moving nodes based on hello packets and inertial navigation. By using the information contained in the hello packets used for neighbor discovery and the inertial navigation system carried by the nodes, the DOA estimation between moving nodes is converted into a positioning problem based on anchor nodes. This method can conveniently and accurately implement the DOA estimation of moving nodes under asynchronous conditions without infrastructure support. The method is a completely autonomous direction information algorithm. Compared with existing blind neighbor discovery schemes, the present invention implements the estimation of the direction of the radiation source through signaling interaction, can reduce the beam search space under the hardware conditions of simulated beamforming, and accelerate the neighbor discovery process.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless self-organizing network communication, and in particular relates to a moving node DOA estimation method based on hello packets and inertial navigation. Background Art

[0002] In drone self-organizing networks, directional beams offer greater directivity and security than omnidirectional beams, enabling longer transmission distances and higher communication rates. However, self-organizing networks using directional antennas face greater difficulties in neighbor discovery. The strong directionality of the beams significantly reduces the probability of nodes discovering each other, significantly increasing the time required for initial networking. The decentralized nature and node mobility of self-organizing networks require efficient and reliable neighbor discovery mechanisms to ensure the establishment and stable operation of communication links.

[0003] Neighbor discovery techniques are categorized into two types: informed discovery and blind discovery, depending on whether a node has prior information about its neighbors when performing neighbor discovery. Blind discovery algorithms address the problem of node discovery in scenarios where prior information about neighbors is lacking and clock and operation synchronization is unavailable. They require random or sequential spatial scanning to identify possible neighbors. Informed discovery algorithms often assume that nodes have already acquired information about neighbor nodes, such as their relative sectors, directions, number, location coordinates, and distances, through public control channels or other technical means, thereby simplifying the search space. Simulation studies have confirmed that neighbor discovery based on prior information about neighbor locations can effectively improve discovery timeliness.

[0004] At the same time, precise positioning technology is at the core of mobile ad hoc networks and collaborative ad hoc networks. Existing positioning technologies can be categorized as GPS-based, anchor-based, and proximity-based. Different types of ad hoc networks and applications require different positioning accuracies, and therefore require different positioning technologies.

[0005] In MANET, in most cases, receiving the geographic coordinates of the mobile terminal through GPS technology can meet the requirements; when GPS is unavailable, anchor node-based positioning technology or proximity-based positioning technology can be used.

[0006] In VANET, GPS positioning technology can provide sufficient accuracy for route guidance functions; for collaborative safety applications such as collision avoidance, which require higher accuracy, assisted GPS (AGPS) or differential GPS (DGPS) technology can be used to perform distance correction using some type of ground reference station with an accuracy of about 10 cm.

[0007] In FANET, the high mobility of flying nodes requires more frequent and more accurate location information. In this case, each drone must be equipped with a GPS and an inertial measurement unit (IMU) to provide its location information to other drones at any time.

[0008] Among anchor node-based positioning algorithms, the CHAN algorithm is a classic wireless positioning algorithm that utilizes TDOA information. The terminal initiates a positioning request and uses uplink physical channels to carry measurement information, which is sent to the base station on the access network. This measurement information carries characteristics of the location to be located, such as signal strength related to distance or time offset related to signal propagation delay. TDOA differential calculations are performed by the base station or core network equipment, and the coordinate position is calculated based on the measurement signals reported by the terminal. The CHAN algorithm is a non-recursive solution to a system of hyperbolic equations. When the measurement error follows an ideal Gaussian distribution, it achieves high positioning accuracy and minimizes computational complexity. In the case of three-dimensional positioning, it can be divided into two types: only four observation base stations participate in positioning, and more than four observation base stations participate in positioning. When the number of base stations is greater than four, the number of nonlinear equations obtained based on the TDOA values ​​is greater than the number of unknown variables. The weighted least squares (WLS) method can fully utilize the redundant data to obtain a better MS position estimate. In this case, the initial nonlinear TDOA equations are first converted into a linear system, and then WLS is used to obtain an initial solution. The second WLS estimation is then performed using the estimated coordinates obtained in the first estimation and known constraints such as additional variables to obtain improved estimated coordinates.

[0009] Among existing positioning technologies, both GPS-based and anchor-node-based positioning technologies rely on comprehensive and effective infrastructure deployment and service coverage, making them impractical in challenging scenarios. Proximity-based positioning relies on known network topology information and is therefore inapplicable during the initial access phase. If, without infrastructure support and asynchronously, a mechanism for estimating neighbor positions or directions could be employed through information exchange, effectively narrowing the beam search space and even confirming the direction of the radiating source for informed discovery, the efficiency of neighbor discovery could be significantly improved. Therefore, studying autonomous neighbor position or direction estimation in self-organizing networks with directional beam-switching antennas is of practical significance. Summary of the Invention

[0010] To this end, the present invention provides a moving node DOA estimation method based on hello packets and inertial navigation. By using the information contained in the hello packets used for neighbor discovery and the inertial navigation system carried by the nodes, the DOA estimation between moving nodes is converted into a positioning problem based on anchor nodes, thereby being able to realize the DOA estimation of moving nodes more conveniently and accurately under conditions without infrastructure support and asynchronously.

[0011] The present invention provides a method for estimating DOA of a moving node based on a hello packet and inertial navigation, comprising the following steps:

[0012] Step 1: The receiving node obtains time information by the interval between receiving hello packets n times, and obtains its own relative motion state information through the inertial navigation system on board, and then knows its own displacement in the north-east coordinate system during the process of receiving hello packets n times in a row, where n ≥ 4;

[0013] The receiving node estimates the displacement of the transmitting node in the north-east coordinate system based on the movement speed information and reception time interval of the transmitting node contained in each received hello packet, thereby obtaining the relative position relationship between the receiving node and the transmitting node each time the hello packet is received;

[0014] Construct a relative coordinate system with the coordinates of the receiving node's first hello packet received as the origin and the transmitting node's coordinates as the fixed point. Then, obtain the coordinates of the receiving node's position each time it receives a hello packet from the second time onwards.

[0015] Step 2: The hello packet sent by the transmitting node contains a pseudo-random spreading sequence. The receiving node that receives the hello packet captures and synchronizes the pseudo-random spreading code and calculates the difference in transmission time between the two received hello packets based on the change in synchronization phase, i.e., the TDOA information.

[0016] Step 3: Based on the location coordinates and TDOA information of each hello packet received by the receiving node, the narrow beam position estimation of the transmitting node is obtained.

[0017] Furthermore, step 3 is implemented by the CHAN algorithm, and the specific operations are as follows:

[0018] The position coordinates of the receiving node receiving the hello packet n times are taken as the observed base station coordinates, and the position coordinates of the transmitting node are taken as the coordinates of the mobile station to be measured. Through the two-step least squares estimation of the CHAN algorithm, 8 symmetrical transmission node position coordinate estimates are obtained;

[0019] Use the transmitter's height information to remove four of the coordinate estimates, then substitute the other four possible solutions into the relative coordinate system. Calculate the distance between the receiving node's location coordinates after receiving the hello packet n times and the transmitter. Compare this with the actual observed distance data and take the solution with the smallest distance difference as the transmitter's location coordinates. Calculate the azimuth and elevation angles of the transmitter's latest received location coordinates relative to the receiver in the relative coordinate system, and then obtain the transmitter's north-east coordinate system direction relative to the receiver.

[0020] By transforming the coordinates of the north-east ground coordinate system and the body coordinate system, the narrow beam position estimation of the receiving node to the transmitting node can be obtained.

[0021] Furthermore, step 3 is implemented by a neural network model; the sample data set used for training the neural network model is generated by simulation, and each sample data set is simulated in a relative coordinate system to generate the coordinates of the transmitting point and the coordinates of n receiving points R1, R2, ..., R n , where R1 is located at the origin of the relative coordinate system, and R2, ..., R n The distance difference with R1 is obtained by adding noise on this basis, specifically including n-1 receiving point coordinates and n-1 distance difference data, and the transmitting point is relative to the receiving point R n The azimuth and elevation angles are used as labels;

[0022] The sample data set is input into the neural network model for model training until the fitting effect of its output on the sample data set label converges or reaches the set maximum number of updates, thereby obtaining the narrow beam position estimation model of the transmitting node.

[0023] Furthermore, the neural network model adopts a three-layer feedforward neural network model, including an input layer, two hidden layers and an output layer.

[0024] Furthermore, the spreading code in step 2 uses a combination of long and short codes, where the long code period exceeds the time of the entire DOA estimation process, and the short code period is much smaller than the long code period.

[0025] The present invention also provides a method for finding neighbors in a UAV self-organizing network, which uses a three-step handshake process to establish neighbor relationships.

[0026] The first handshake step: The existing node uses a narrow beam to periodically send hello packets at various beam positions. The newly-joined node uses a widened beam to receive them. If the newly-joined node receives hello packets from the same existing node n times in a period of time, it extracts the relative receiving position information and TDOA information based on the above method to estimate the incoming direction of the beam from the existing node and the candidate beam positions aligned with the incoming direction of the beam from the existing node.

[0027] The second handshake step: The new node sends a hello-ACK packet at the candidate beam position estimated in the first step. The existing node receives the hello-ACK packet using a narrow beam. The hello-ACK packet contains the relative direction information between the new node and the existing node.

[0028] The third handshake step: The existing node successfully receives the hello-ACK packet, aligns the wave position of the new node according to the relative direction information in the hello-ACK packet, and sends an ACK message to it; if the new node successfully receives the ACK message, the third handshake step is successful and the neighbor relationship is established.

[0029] The present invention also protects a computer device comprising a memory, a processor, and a computer program stored in the memory and executed on the processor.

[0030] A computer program running on the processor, which implements the above-mentioned drone self-organizing network neighbor discovery method when the processor executes the computer program; and a computer-readable storage medium on which the computer program is stored, which is executed by the processor to implement each step of the above-mentioned drone self-organizing network neighbor discovery method.

[0031] The present invention obtains TDOA information through the information contained in the hello packet, the displacement estimation of the IMU and the spread spectrum code, thereby realizing neighbor direction information estimation. It is independent of GPS receivers, base stations and other infrastructure and is a completely autonomous direction information algorithm. Compared with existing blind neighbor discovery schemes, the present invention realizes the direction estimation of the radiation source through signaling interaction, can reduce the beam search space under the hardware conditions of simulated beamforming, and accelerate the neighbor discovery process. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a schematic diagram of the neighbor discovery process for drone ad hoc networks;

[0033] Figure 2 This is a schematic diagram of the relative position relationship between a newly connected node and existing nodes when the node receives hello packets multiple times in a row. DETAILED DESCRIPTION

[0034] The present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments. The embodiments of the present invention are provided for purposes of illustration and description and are not intended to be exhaustive or to limit the invention to the disclosed forms. Many modifications and variations will be apparent to those skilled in the art. The embodiments are chosen and described to better illustrate the principles of the invention and its practical application, and to enable those skilled in the art to understand the invention and design various embodiments with various modifications suitable for specific applications. Example 1

[0035] During the neighbor discovery process, there are two types of nodes in the drone self-organizing network: existing nodes and newly joined nodes. The network time frame is divided into two parts: the neighbor discovery phase and the directional communication phase. In the neighbor discovery phase, the existing nodes will periodically scan and send hello packets. After the newly joined nodes are launched, they need to actively listen to the channel to receive possible handshake signals. This embodiment uses a three-step handshake process to establish neighbor relationships, such as Figure 1 As shown, the ability of the phased array antenna to adjust the beam width can be fully utilized.

[0036] The first handshake step: The existing node uses a narrow beam to periodically send hello packets at various wave positions. The newly-joined node uses a widened beam to receive them. If the newly-joined node receives hello packets from the same existing node for n consecutive times (n ≥ 4) within a period of time, it extracts the relative position information and TDOA information, estimates the incoming wave direction of the moving node (that is, the existing node), and estimates the candidate wave positions aligned with the incoming wave direction of the existing node. At this point, the first handshake process is successfully completed.

[0037] Second handshake step: The new node sends a hello-ACK packet on the candidate beam position estimated in the first step. The existing node receives the hello-ACK packet using a narrow beam. The hello-ACK packet contains the relative direction information between the new node and the existing node.

[0038] The third handshake step: The node on the network successfully receives the hello-ACK packet, aligns the wave position of the new node according to the relative direction information in the hello-ACK packet, and sends an ACK message to it. If the new node successfully receives the ACK message, the third handshake step is successful and the neighbor relationship is established.

[0039] In the neighbor discovery process described above, existing nodes are transmitting nodes, and newly connected nodes are receiving nodes. Both are in motion. The following describes a method for estimating the DOA of moving nodes, using the neighbor discovery process as an example.

[0040] 1. Assume that during the neighbor discovery process, a network node receives five consecutive hello packets within a period of time. The node obtains time information based on the intervals between the five hello packet receptions and obtains its own relative motion state information through the inertial navigation system on board. This allows the node to determine its displacement in the north-east coordinate system during the five consecutive hello packet receptions.

[0041] At the same time, the newly connected node estimates the displacement of the node in the north-east coordinate system through the movement speed information and reception time interval of the node contained in each hello packet received, thereby obtaining the relative position relationship between the newly connected node and the transmitting node each time the hello packet is received.

[0042] Construct a relative coordinate system, with the coordinates of the location of the new network node receiving the hello packet for the first time as the origin of the relative coordinate system, and the location coordinates of the network node as the fixed point T, and then obtain the coordinates of the location of the new network node receiving the hello packet from the second time onwards, refer to Figure 2 shown. Figure 2 This represents the relative position of a newly joined node to an existing node when it receives five consecutive hello packets. D1 through D5 represent the distances between the newly joined node and the existing node. Although the existing node is constantly moving, it can be considered stationary in the relative motion relationship.

[0043] 2. The hello packet sent by the transmitting node contains a pseudo-random spreading sequence. The receiving node that receives the hello packet captures and synchronizes the pseudo-random spreading code and calculates the difference in transmission time of the two received hello packets based on the change in synchronization phase, that is, the TDOA information.

[0044] Ranging based on pseudo-random spread spectrum sequences and then achieving time synchronization through changes in the synchronization phase to obtain TDOA information is a prior art and will not be described in detail here. Assume that the phase capture error of the spread spectrum code is When using a 10.23MHz pseudo-random spread spectrum sequence, the TDOA error can be less than 0.1ns.

[0045] However, the periodicity of the pseudo-random sequence can cause phase ambiguity. In this embodiment, long and short codes are used simultaneously. The long code is used to eliminate the phase ambiguity problem, and its period exceeds the time of the entire DOA estimation process. The period of the short code is much smaller than that of the long code, and is used to quickly capture the long code phase each time a message is received.

[0046] Assume that the symbol rate of the short spreading code used is , the code element width is , the sequence length is , the sequence period is , the symbol rate of the long spreading code used is , the code element width is , the sequence length is , the sequence period is , the period of long code and short code satisfies Much smaller than The accuracy of TDOA is provided by the short code, and the long code assists the short code to eliminate phase ambiguity. ,Right now ,in And it is an integer.

[0047] The moment when the shift register that generates the spread spectrum code appears in the "all 1" state is called the phase reference of the spread spectrum code. The phase period of the spread spectrum code is equal to the sequence period. The phase information can be regarded as a timing unit. The transmitting node will send Msg1 multiple times. Assuming that the receiving node has passed time Msg1 is received twice, and the alignment phases of the short spreading codes during the two receptions are The chip time and chip time, the alignment phases of the long spreading codes are The chip time and chip time.

[0048] The period ambiguity of the short code is obtained from the phase difference of the long code , assuming that the two transmission delays are and , the time interval for receiving hello packets can be calculated as follows:

[0049] ,

[0050] In this way, the receiving node can obtain the difference in the distance between the sending and receiving nodes corresponding to the two transmissions from the synchronization phase and its own receiving time interval. If the code chip synchronization time error of the spread spectrum code is less than The error of the distance difference measurement value will be Within, is the propagation speed of electromagnetic waves.

[0051] 3. Based on the location coordinates and TDOA information of each hello packet received by the receiving node, the narrow beam position estimation of the transmitting node is obtained.

[0052] Step 3 can be implemented by the CHAN algorithm, which converts the DOA estimation between moving nodes into a positioning problem based on anchor nodes. The specific operations are as follows:

[0053] The coordinates of the receiving node receiving the hello packet five times are used as the observed base station coordinates, and the coordinates of the transmitting node are used as the coordinates of the mobile station to be measured. Through the two-step least squares estimation of the CHAN algorithm, 8 symmetrical estimates of the transmitting node's coordinates are obtained.

[0054] Using the transmitter's height information, remove four of the coordinate estimates. Substitute the remaining four possible solutions into the relative coordinate system and calculate the distances between the receiving node's location coordinates after receiving the hello packet five times and the transmitter. These are compared with the actual observed distance data, and the solution with the smallest distance difference is taken as the transmitter's location coordinates. This allows the calculation of the azimuth and elevation angles of the transmitter's latest received location coordinates relative to the receiver in the relative coordinate system, ultimately determining the transmitter's north-east coordinate system direction relative to the receiver. The solution with the smallest distance difference is compared with the observed distance difference data, using the distance differences between D2-D5 and D1. The coordinates corresponding to the set of distance differences with the smallest Euclidean distance are then taken as the unique solution.

[0055] By transforming the coordinates of the north-east ground coordinate system and the body coordinate system, the narrow beam position estimation of the receiving node to the transmitting node can be obtained.

[0056] Step 3 can also be implemented through a neural network model; the sample data set used for neural network model training is generated through simulation. Each sample data set is generated by simulating the coordinates of the transmitting point and the coordinates of five receiving points R1, R2, R3, R4, and R5 in a relative coordinate system, where R1 is located at the origin of the relative coordinate system. The distance difference between R2, R3, R4, and R5 and R1 is calculated, and noise is added to this. Specifically, it includes four receiving point coordinates and four distance difference data, and the azimuth and pitch angles of the transmitting point relative to the receiving point R5 are used as labels;

[0057] The sample data set is input into the neural network model for model training until the fitting effect of its output on the sample data set label converges or reaches the set maximum number of updates, thereby obtaining the narrow beam position estimation model of the transmitting node.

[0058] Since TDOA is essentially a problem of finding intersection points of multiple curves and the problem complexity is not high, the neural network model can adopt a three-layer feedforward neural network model, including an input layer, two hidden layers and an output layer. Example 2

[0059] A computer device includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the drone self-organizing network neighbor discovery method described in Example 1 is implemented. Example 3

[0060] A computer-readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the various steps of the drone self-organizing network neighbor discovery method described in Example 1.

[0061] Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field and related fields without making any creative work should fall within the scope of protection of the present invention.

Claims

1. A moving node DOA estimation method based on hello packets and inertial navigation, characterized in that: The following steps are involved: Step 1: The receiving node obtains time information by the interval between receiving hello packets n times, and obtains its own relative motion state information through the inertial navigation system on board, and then knows its own displacement in the north-east coordinate system during the process of receiving hello packets n times in a row, where n ≥ 4; The receiving node estimates the displacement of the transmitting node in the north-east coordinate system based on the movement speed information and reception time interval of the transmitting node contained in each received hello packet, thereby obtaining the relative position relationship between the receiving node and the transmitting node each time the hello packet is received; Construct a relative coordinate system with the coordinates of the receiving node's first hello packet received as the origin and the transmitting node's coordinates as the fixed point. Then, obtain the coordinates of the receiving node's position each time it receives a hello packet from the second time onwards. Step 2: The hello packet sent by the transmitting node contains a pseudo-random spreading sequence. The receiving node that receives the hello packet captures and synchronizes the pseudo-random spreading code and calculates the difference in transmission time between the two received hello packets based on the change in synchronization phase, i.e., the TDOA information. Step 3: Based on the location coordinates and TDOA information of each hello packet received by the receiving node, the narrow beam position estimation of the transmitting node is obtained.

2. The method for estimating DOA of a moving node based on hello packets and inertial navigation according to claim 1, wherein: Step 3 is implemented through the CHAN algorithm. The specific operations are: The position coordinates of the receiving node receiving the hello packet n times are taken as the observed base station coordinates, and the position coordinates of the transmitting node are taken as the coordinates of the mobile station to be measured. Through the two-step least squares estimation of the CHAN algorithm, 8 symmetrical transmission node position coordinate estimates are obtained; Use the transmitter's height information to remove four of the coordinate estimates, then substitute the other four possible solutions into the relative coordinate system. Calculate the distance between the receiving node's location coordinates after receiving the hello packet n times and the transmitter. Compare this with the actual observed distance data and take the solution with the smallest distance difference as the transmitter's location coordinates. Calculate the azimuth and elevation angles of the transmitter's latest received location coordinates relative to the receiver in the relative coordinate system, and then obtain the transmitter's north-east coordinate system direction relative to the receiver. By transforming the coordinates of the north-east ground coordinate system and the body coordinate system, the narrow beam position estimation of the receiving node to the transmitting node can be obtained.

3. The method for estimating DOA of a moving node based on hello packets and inertial navigation according to claim 1, wherein: Step 3 is implemented through a neural network model; The sample data sets used for neural network model training are generated by simulation. Each sample data set generates the coordinates of the transmitting point and the coordinates of n receiving points R1, R2, ..., R in the relative coordinate system. n , where R1 is located at the origin of the relative coordinate system, and R2, ..., R n The distance difference with R1 is obtained by adding noise on this basis, specifically including n-1 receiving point coordinates and n-1 distance difference data, and the transmitting point is relative to the receiving point R n The azimuth and elevation angles are used as labels; The sample data set is input into the neural network model for model training until the fitting effect of its output on the sample data set label converges or reaches the set maximum number of updates, thereby obtaining the narrow beam position estimation model of the transmitting node.

4. The method for estimating DOA of a moving node based on hello packets and inertial navigation according to claim 3, wherein: The neural network model adopts a three-layer feedforward neural network model, including an input layer, two hidden layers and an output layer.

5. The method for estimating DOA of a moving node based on hello packets and inertial navigation according to claim 1, wherein: The spreading code in step 2 uses a combination of long and short codes. The long code period exceeds the time of the entire DOA estimation process, and the short code period is much smaller than the long code period.

6. A method for neighbor discovery in a drone self-organizing network, characterized in that: A three-step handshake process is used to establish neighbor relationships: The first handshake step: the network node uses a narrow beam to periodically send hello packets at each wave position, and the newly joined node uses a widened beam to receive them. If the hello packet from the same network node is received n times in a row within a period of time, the newly joined node extracts the relative receiving position information and TDOA information based on the moving node DOA estimation method described in any one of claims 1 to 5, estimates the direction of arrival of the network node, and estimates the candidate wave position aligned with the direction of arrival of the network node; The second handshake step: The new node sends a hello-ACK packet at the candidate beam position estimated in the first step. The existing node receives the hello-ACK packet using a narrow beam. The hello-ACK packet contains the relative direction information between the new node and the existing node. The third handshake step: The existing node successfully receives the hello-ACK packet, aligns the wave position of the new node according to the relative direction information in the hello-ACK packet, and sends an ACK message to it; if the new node successfully receives the ACK message, the third handshake step is successful and the neighbor relationship is established.

7. A computer device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the method for neighbor discovery in a drone self-organizing network according to claim 6 is implemented.

8. A computer-readable storage medium, characterized in that A computer program is stored thereon, and the computer program is executed by a processor to implement the various steps of the drone self-organizing network neighbor discovery method described in claim 6.

Images