A method for evaluating the anti-interference performance of UAV ad hoc networks

Abstract

This invention belongs to the field of UAV communication network technology, specifically relating to a method for evaluating the anti-interference performance of UAV ad hoc networks. It includes the following steps: scenario initialization and network initialization; transmitter link construction, supporting dynamic network entry and exit of interfered nodes, and converting digital signals into radio frequency analog signals; channel transmission link construction, simulating controllable interference applied to the wireless transmission path through a jammer; signal reception link construction, used to receive radio frequency signals and complete the analog-to-digital domain conversion, outputting baseband binary data packets through demodulation and synchronization processing; data acquisition and processing, obtaining the control group experimental dataset, encapsulating the evaluation model using the Naive Bayes algorithm, and performing reliability verification. This invention can quantify the anti-interference performance of UAV ad hoc networks, providing a comparative basis for anti-interference performance.

Description

Technical Field

[0001] This invention belongs to the field of UAV communication network technology and provides a method for evaluating the anti-interference performance of UAV self-organizing networks. Background Technology

[0002] With the rapid development of UAV technology, UAV swarm collaborative operations have been widely applied in many fields such as geographic surveying, emergency communication, environmental inspection, and regional reconnaissance. Ad Hoc networks (Ad Hoc) serve as the core communication architecture for UAV swarms, possessing characteristics such as decentralized, distributed deployment, dynamic node networking, and multi-hop relay transmission, enabling them to adapt to operational scenarios involving high-speed movement and dynamic topology changes in UAV swarms. However, in complex electromagnetic environments, Ad Hoc networks are highly susceptible to interference from co-channel, adjacent-channel, and malicious electromagnetic interference, leading to communication interruptions, data packet loss, and topology breaks, directly causing the failure of swarm collaborative operations. Therefore, anti-interference performance is a core performance indicator for Ad Hoc networks, and accurate, comprehensive, and quantifiable anti-interference performance evaluation methods are a crucial prerequisite for the research, development, solution verification, and engineering implementation of anti-interference technologies for Ad Hoc networks.

[0003] Currently, existing technologies primarily focus on the point-to-point single-link communication quality between UAVs and ground stations, and between UAVs themselves. By collecting link-layer metrics such as bit error rate, packet loss rate, outage probability, and interference-to-signal ratio, and combining theoretical analysis, software simulation, and field measurements, quantitative characterization of the anti-interference capability of a single communication link is achieved. This type of method is technologically mature and has been widely applied in single-point UAV communication systems. Building upon this foundation, some research attempts to extend to network-level evaluation. By introducing network-layer metrics such as network topology connectivity and routing convergence time, the overall anti-interference capability of UAV self-organizing networks is indirectly reflected, providing preliminary technical ideas for network-level evaluation.

[0004] In summary, existing evaluation methods typically assume that UAV node locations are fixed and network topology remains unchanged, which fails to reflect the impact of dynamic topology changes on communication link quality and overall network anti-interference capability. Furthermore, they are mostly customized designs targeting single interference patterns such as single-tone interference, specific node numbers, and specific protocol architectures, and cannot simulate the dynamic network operation process of UAVs during actual flight and in complex electromagnetic environments, resulting in significant deviations between evaluation results and actual operating conditions. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method for evaluating the anti-interference performance of UAV self-organizing networks, which solves the problem in the prior art that it is impossible to quantitatively compare the anti-interference performance of UAV self-organizing networks.

[0006] The technical solution of the present invention includes: Several test groups are constructed according to preset parameter combinations to obtain the bit error rate, interruption probability and effective coordinate observation data of the initial baseband binary data packets of each node in the UAV ad hoc network in each test group after interference is applied. The preset parameters include multiple interference patterns, multiple interference parameters, multiple channel fading patterns, multiple modulation methods and different numbers of UAVs. The baseband binary data packets include MAC header, sequence number, IP header, UDP header and measured data coordinates.

[0007] The initial baseband binary data packets in each test group are converted into radio frequency (RF) analog signals and transmitted. Controllable electromagnetic interference with interference patterns and parameters corresponding to these interference patterns is applied to the RF analog signals, and channel fading processing is performed to obtain RF signals that have undergone interference and channel fading processing. The interference parameters include interference-to-signal ratio, interference frequency, and interference bandwidth. The RF signals that have undergone interference and channel fading processing are received, demodulated, and synchronized to restore the interfered baseband binary data packets. The bit error rate, outage probability, and effective coordinate observation data are calculated based on the interfered baseband binary data packets and the initial baseband binary data packets.

[0008] The fusion error of the network is calculated based on the bit error rate, number of UAVs and effective coordinate observation data of each test group. A mapping dataset from interference parameters to fusion error is established. The interference parameters of the scenario to be evaluated are matched with the mapping dataset to obtain the fusion error of the corresponding UAV self-organizing network. The anti-interference performance of the UAV self-organizing network is calculated based on the fusion error, thereby completing the evaluation.

[0009] Furthermore, the interference patterns include single-tone interference, broadband noise interference, noise FM interference, noise AM interference, and no added interference. Channel fading patterns include free-space path loss, shadowing fading, frequency-selective fading, and flat fading. Modulation methods include QPSK, 16QAM, and 32QAM.

[0010] Furthermore, the effective coordinate observation data are the target coordinate observation values ​​parsed from the restored baseband binary data packets, and the coordinate observation data whose deviation from the target's true coordinates is within the preset measurement error standard deviation range.

[0011] Furthermore, the fusion error is calculated from the measurement error corresponding to the effective coordinate observation data and the number of networked UAVs.

[0012] Furthermore, demodulation and synchronization processing includes: quadrature demodulation, low-pass filtering, LS channel estimation, linear interpolation, zero-forcing equalization, symbol decision, deinterleaving, and Viterbi decoding.

[0013] Furthermore, each group of experiments underwent at least 100 Monte Carlo cycles.

[0014] Furthermore, the criteria for judging the anti-interference performance R of UAV self-organizing networks are as follows: when At that time, the self-organizing network of drones has strong anti-interference performance.

[0015] when At that time, the anti-interference performance of the UAV self-organizing network was in the middle stage.

[0016] when At that time, the anti-interference performance of the drone self-organizing network was weak.

[0017] when At that time, the self-organizing network of drones does not have anti-interference capabilities.

[0018] Furthermore, the method for converting baseband binary data packets into radio frequency signals is as follows: performing convolutional coding, interleaving, modulation, OFDM symbol mapping, pilot insertion, IFFT transform, and cyclic prefix addition on the baseband binary data in sequence.

[0019] Furthermore, before converting the baseband binary data packets of the UAV self-organizing network nodes into radio frequency analog signals and transmitting them, scenario initialization and network initialization are performed to build the UAV self-organizing network communication architecture.

[0020] Scene initialization includes: constructing an autonomous electromagnetic clean reconnaissance scene of preset size, deploying programmable signal jammers, ground data transmission stations, switches and control consoles, and reserving a dedicated downlink communication frequency band between the ground base station and the UAV; Network initialization includes: deployment of TTNT data link protocol stacks for each node, pre-configuration of interface parameters at each layer, node ID allocation, IP address planning, TDMA time slot allocation, and distributed cluster head election.

[0021] Furthermore, during the process of converting baseband binary data packets of UAV ad hoc network nodes into radio frequency analog signals and transmitting them, dynamic network entry and exit of interfered nodes can be achieved. Dynamic network entry and exit include node network entry identity verification, normal network exit request response, abnormal network exit timeout monitoring, and dynamic scheduling, reclamation, and reallocation of time slot resources and IP addresses of the corresponding nodes.

[0022] The technical solution provided by this invention has the following advantages compared with the prior art: This invention, through an end-to-end closed-loop processing flow—constructing a process from baseband binary data packets to radio frequency analog signals for each node in a UAV ad hoc network, simulating controllable electromagnetic interference and channel fading characteristics with preset interference parameters, and demodulating and synchronizing the receiver's radio frequency signals while restoring the baseband data packets—firstly overcomes the limitation of existing technologies that only focus on point-to-point single-link anti-interference performance evaluation. It achieves overall coverage of the communication links of all nodes in a distributed UAV ad hoc network, fully characterizing the transmission impact of local link interference on the overall network communication capability. This fundamentally solves the core defect of existing technologies that cannot complete network-level overall anti-interference performance evaluation. Simultaneously, this invention constructs a control experiment using multiple sets of preset parameters, collecting network bit error rate, outage probability, and effective coordinate observation data via Monte Carlo cyclic sampling and calculating fusion error. This establishes a mapping dataset from interference parameters to fusion error, directly quantifying and binding the interference damage of the underlying communication links to the core service performance of the network. This avoids the problems of existing evaluation system indicators being disconnected from actual services and having limited engineering application value, making the evaluation... The results accurately reflect the actual impact of interference on the core operational capabilities of UAV ad hoc networks. Through comparative experiments covering multiple parameters, it adapts to network operation scenarios with different node sizes, interference environments, and channel conditions, overcoming the limitations of existing technologies based on static fixed topology evaluation. This ensures a high degree of consistency between the evaluation results and the actual complex operating conditions of UAV ad hoc networks. Ultimately, by matching the interference parameters of the scenario under evaluation with the mapping dataset, this invention can quickly obtain the fusion error of the corresponding network and complete the quantitative judgment of anti-interference performance. It forms a standardized and reusable quantitative evaluation model for the network-level anti-interference performance of UAV ad hoc networks. This model enables unified horizontal comparison of anti-interference performance under different network schemes and interference scenarios, as well as rapid prediction of anti-interference performance under unknown interference scenarios. It fills the gap in existing technologies lacking a standardized network-level anti-interference quantitative evaluation system adapted to the dynamic characteristics of UAV ad hoc networks, providing a unified, accurate, and quantifiable evaluation basis and judgment standard for the research, development, scheme verification, and engineering implementation of UAV ad hoc network anti-interference technology.

[0023] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1This is a framework diagram of a method for evaluating the anti-interference performance of an unmanned aerial vehicle (UAV) self-organizing network, provided in an embodiment of the present invention.

[0026] Figure 2 This is a schematic diagram of a method for evaluating the anti-interference performance of an unmanned aerial vehicle (UAV) self-organizing network, provided in an embodiment of the present invention.

[0027] Figure 3 A flowchart of a method for evaluating the anti-interference performance of an unmanned aerial vehicle (UAV) self-organizing network, provided as an embodiment of the present invention.

[0028] Figure 4 This is a scene diagram of an anti-interference experiment for a drone self-organizing network provided in an embodiment of the present invention.

[0029] Figure 5 This is a TDMA frame structure diagram provided in an embodiment of the present invention.

[0030] Figure 6 This is a flowchart of cluster head election provided as an embodiment of the present invention.

[0031] Figure 7 This is a flowchart of a node joining the network provided in an embodiment of the present invention.

[0032] Figure 8 This is a flowchart of a node decommissioning process provided in an embodiment of the present invention.

[0033] Figure 9 This is a schematic diagram of a signal transmission module provided in an embodiment of the present invention.

[0034] Figure 10 This is a schematic diagram of a channel module provided in an embodiment of the present invention.

[0035] Figure 11 This is a schematic diagram of a signal receiving module provided in an embodiment of the present invention. Detailed Implementation

[0036] The following detailed description of a specific embodiment of the present invention is provided in conjunction with the accompanying drawings. However, it should be understood that the scope of protection of the present invention is not limited to the specific embodiment.

[0037] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the technical solution of this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0038] In the description of the embodiments of the present invention, unless otherwise stated, "a plurality of" means two or more.

[0039] like Figures 1 to 11 As shown, this invention provides a method for evaluating the anti-interference performance of unmanned aerial vehicle (UAV) self-organizing networks, comprising: Step S1: Scene initialization and network initialization, specifically including: Step S101: Model the electromagnetic transmission environment of the UAVS data link. This involves a reconnaissance autonomous scenario in an open, relatively clean electromagnetic environment with a length × width × height of at least 50 km × 50 km × 5 km. Figure 4 As shown.

[0040] Deploy a programmable signal jammer, which can set parameters such as jamming style, transmission power, center frequency, and bandwidth, to apply controlled electromagnetic interference to the UAV ad hoc network. A safe frequency band for communication between the UAV and the ground base station must be reserved.

[0041] Deploy switches to connect jammers, ground data radios, and control consoles via wired links to achieve wired networking and stable data transmission between ground devices.

[0042] The system deploys antennas and LR24-F ground data transmission stations. These stations connect to the switches and antennas, enabling bidirectional conversion between radio frequency and digital signals, as well as scheduling and forwarding of signals from multiple unmanned personnel. The control console serves as the ground control and data processing center, running the Mission Planner ground station software.

[0043] Step S102: Perform network initialization, including TTNT data link protocol stack deployment, interface parameter presetting at each layer, node ID allocation, IP planning, TDMA time slot allocation, and cluster head election, as detailed below: The application layer is the starting point of the data encapsulation process, and its core is to generate raw business data payloads adapted to the UAV networking scenario. This layer constructs differentiated data units according to networking requirements, such as HELLO packets for neighbor topology awareness (containing unique node identifiers, geographic coordinates, number and list of neighbor nodes) or business data packets for task interaction (containing target monitoring coordinates, data acquisition errors, task instructions, etc.). All data is output in a structured format, providing the basic payload for subsequent encapsulation layers.

[0044] The transport layer receives the structured payload output from the application layer and implements end-to-end communication control based on the IP protocol. Through port number management, data segmentation, sequence number control, checksum calculation, and timeout retransmission mechanism, it ensures the integrity, orderliness, and reliability of data transmission. Its core encapsulation is to add a UDP protocol header to the application layer payload to achieve port-level addressing and transmission control.

[0045] This layer first initializes the UDPHeader data structure. The specific structure of the UDP header is shown in Table 1. The source port is randomly assigned a temporary port number, and the destination port is fixed as the TTNT dedicated port 5000. The checksum field is initialized to 0 to adapt to the low-latency communication requirements of UAVs. Then, the UDP header is converted into a binary stream through serialization and concatenated with the application layer payload to form a complete UDP data packet, ensuring that the data can be identified by the target node's specified port during transmission.

[0046] Table 1: Detailed Structure of the UDP Header The network layer uniquely identifies each node in the network using IP addresses, providing end-to-end network transmission services for upper-layer applications. This layer encapsulates IP protocol headers based on UDP packets to achieve cross-node internet addressing and transmission range control.

[0047] This layer constructs an IPHeader data structure conforming to the IPv4 specification, as shown in Table 2. The protocol number identifies the UDP protocol, the Time-to-Live (TTL) limits the number of hops the packet can be forwarded, and both the source IP (the current node's private IP) and the destination IP (the receiving node's IP or broadcast address) are assigned values. After serialization, the IP header is concatenated with the UDP packet to form an IP datagram, ensuring that data can be routed and forwarded across nodes within the UAV network.

[0048] Table 2: IPHeader Data Structure Simultaneously, a lightweight dynamic routing protocol is implemented. As shown in Table 3, the entire network topology and routing table information are synchronized through routing update messages, and the reachable paths and link costs of each node are maintained. The cluster head node periodically broadcasts this message to update the routing status of the entire network, ensuring that the UAV network can still complete efficient cross-node routing forwarding and topology adaptation when nodes dynamically join or leave the network.

[0049] Table 3: Routing Update Message Data Structure After network layer encapsulation, a sequence number addition step is required, which is a crucial step to ensure the orderliness and integrity of data transmission. The system assigns a unique, incrementing sequence number to each IP datagram. This sequence number is associated with and bound to the source node ID and destination node ID of the data packet and appended to the end of the IP datagram. This not only enables the receiving end to identify duplicate data packets and perform deduplication, but also allows the system to determine whether a data packet has been lost by checking for missing sequence numbers, providing a basis for subsequent retransmission mechanisms. This addresses the data packet out-of-order and loss problems that are prone to occur in UAV wireless communication.

[0050] The data link layer receives IP data packets with sequence numbers and includes a Media Access Control (MAC) sublayer and a Logical Link Control (LLC) sublayer. The MAC sublayer uses a TDMA time slot scheduling mechanism to achieve multi-node coordinated communication and completes media sharing access through priority queue management, data frame encapsulation, and channel access control. The LLC sublayer is responsible for frame synchronization, error control, and flow control, providing the network layer with indiscriminate data frame transmission services. The core of this layer's encapsulation is assembling a MAC header into the IP datagram with a sequence number, enabling link-level addressing and transmission scheduling.

[0051] This layer initializes the MACHeader data structure, as shown in Table 4. The priority field embeds the source node ID for differentiated scheduling. After serialization to generate the MAC header, it is concatenated with the IP data packet containing the sequence number. A CRC checksum is added to the end of the data packet, ultimately forming a complete MAC frame that can be transmitted via the wireless channel. At this point, the encapsulation process from the application layer to the data link layer is complete. The data packet possesses all the capabilities for addressing, routing, fault tolerance, and link transmission, and can be sent to the target node via the UAV's wireless communication module.

[0052] Table 4: Initializing the MACHeader Data Structure The physical layer is the final carrier of wireless data transmission. It is responsible for modulating and demodulating the signal, coding and decoding the channel, and simulating interference of the complete MAC frame output from the data link layer. It realizes the complete physical link simulation from baseband processing to radio frequency conversion. Through waveform generation, spectrum analysis, and bit error rate calculation, it provides reliable bit stream transmission services to the upper layer. Finally, it sends the encapsulated data packet to the target node through the UAV wireless communication module. The physical layer encapsulation content is shown in Table 5.

[0053] Table 5: Physical Layer Encapsulation Contents TDMA frame structure, such as Figure 5 As shown, the broadcast HELLO message subframe is used by each UAV node in the network to periodically broadcast neighbor discovery and basic network information. The event contention subframe is used for new nodes to initiate network access verification and for online nodes to initiate normal network exit requests. The allocation / response subframe is exclusively used by the cluster head to respond to and confirm access requests within contention slots, while simultaneously completing node IP address allocation, slot resource scheduling, and the distribution of network-wide synchronization information. The data transmission subframe is used by nodes that have completed access to complete the orderly transmission of service data packets according to the dedicated slots allocated by the cluster head; its time domain length is... Transmission efficiency and changes in actual nodes within the network need to be considered comprehensively.

[0054] Cluster leader election, such as Figure 6 As shown, since nodes cannot simultaneously receive and transmit signals, to avoid transmission conflicts and improve network efficiency, each node needs to conduct distributed cluster head election based on its own connectivity, the number of neighboring nodes, and channel quality. When the network is initially powered on and no cluster head exists, nodes sense each other by broadcasting HELLO messages. If no response from a cluster head is received within m periods, the first round of initial cluster head election is automatically triggered, prioritizing nodes with high connectivity, wide coverage, and stable channels as cluster heads to support subsequent centralized network management.

[0055] If a cluster head exits abnormally during communication and cluster members do not receive control information from the cluster head within a preset period, a cluster head re-election request will be broadcast in the next period. Once all cluster members have completed the broadcast of the re-election request, the network will start a new round of cluster head election process to maintain network continuity and robustness.

[0056] After the cluster head election is completed, the cluster head node periodically broadcasts its own status and control information, undertaking core network management responsibilities, including: collecting the operating status information of all nodes in the current network, IP address allocation, and time slot resource scheduling. Simultaneously, the cluster head is responsible for node network entry authentication, normal network exit information verification, and the detection and handling of abnormal node offline events.

[0057] Step S2: Transmitter link modeling, supporting dynamic network entry and exit of interfered nodes, converting digital signals into radio frequency analog signals, completing signal conditioning and wireless transmission, including networking module and signal transmission module.

[0058] Step S201: Model the network module, including rules for node entry and exit from the network.

[0059] Node entry rules, such as Figure 7 As shown, when an external node enters the network broadcast range, it listens for HELLO messages sent by internal nodes to obtain basic information such as network identifier and frame structure. After successfully parsing the HELLO message, the node initiates a network entry request broadcast in the event contention subframe. If the node does not receive a verification response from the cluster head within the allocation / response time slot or the verification fails, it automatically listens for the HELLO message in the next cycle until the network entry request interaction and authentication are completed. After successful verification, the cluster head allocates IP addresses and schedules time slot resources for the new node based on the current number of network nodes and channel resource occupancy. After completing the network entry process, the node can enter the service data transmission stage.

[0060] Node decommissioning rules, such as Figure 8 As shown, in normal network decommissioning mode, the decommissioning node sends a decommissioning request to the cluster head in the event contention subframe and waits to receive a response. After the decommissioning request is verified by the cluster head, the node actively disconnects from the network. The cluster head reclaims the time slot resources and IP address occupied by the node, re-completes the network resource scheduling, and broadcasts the updated allocation scheme. If the node does not receive a confirmation response from the cluster head in the allocation / response time slot or the verification fails, it will re-initiate the decommissioning request in the next cycle. Abnormal network decommissioning adopts a timeout monitoring mechanism. The cluster head periodically monitors the time slot sending behavior of all nodes in the network. If... If no transmission behavior of the target node is detected within a certain period, it is determined that the node has abnormally left the network. The cluster head directly reclaims the network resources and IP address it occupies, completes the reallocation of time slots and IP addresses, and broadcasts them.

[0061] Step S202, Modeling the signal transmission module, such as Figure 9 As shown in the diagram, the meaning of each symbol and the specific process are as follows: Set symbol rate and sampling frequency Then the symbol period Sampling period Number of sampling points per symbol The network module transmits... Bit binary data packet Channel coding is performed using convolutional coding with a constraint length of 7 and a code rate of [missing information]. The generator polynomial is and The shift register is initially set to all zeros. The encoded bitstream is obtained by performing a linear XOR combination based on the generator polynomial. , length is .

[0062] Let the seed of the random number generator be . Based on seeds Generate a length of A random permutation sequence, input bit stream Rearrange the bits according to the permutation sequence to obtain the interleaved bit stream. , length is .

[0063] After intertwining Mapped to The bipolar sequence, split into odd and even positions. Luhe Each path consists of 2 bits forming a QPSK symbol.

[0064] Will Power normalization of the two bit paths yields a discrete QPSK symbol sequence. .Pick Length is Pilot symbol count: ,in, For pilot spacing, For the floor function, the total number of signs is: Randomly generate normalized QPSK pilot sequences, selecting the real and imaginary parts independently. Ensure that the pilot signal power is consistent with the data symbol power. Construct a system initialized with all zeros. Complex matrices, where, This is the number of FFT points. The data symbols and pilot symbols are arranged according to the pilot interval. By inserting the aforementioned complex matrix, we obtain the OFDM frequency domain symbol matrix after inserting the pilots. For the OFDM frequency domain symbol matrix Performing an inverse fast Fourier transform (IFFT) along the subcarrier dimension yields... OFDM time domain matrix The time-domain matrix of the IFFT output. Extract its last The line is used as a loop prefix and concatenated to At the top, we obtain the OFDM time-domain matrix after adding the cyclic prefix. .

[0065] OFDM matrix after adding cyclic prefix Reconstruct the signal into a one-dimensional time-domain transmission signal in column-major order, separate the real and imaginary parts of this one-dimensional time-domain transmission channel, and multiply the real part by... imaginary part multiplied by ,Will Roadbed band signal modulated to carrier frequency Above, a radio frequency transmission signal is generated. .

[0066] Step S3: Channel transmission link modeling. Simulate the application of controllable interference to the wireless transmission path through an jammer, and construct a channel module, such as... Figure 10 As shown in the diagram, the meaning of each symbol and the specific process are as follows: The air-to-ground wireless transmission channel in dynamic UAV scenarios includes interference patterns and channel models. The interference patterns include single-tone interference, broadband noise interference, noise frequency modulation interference, noise amplitude modulation interference, and no interference. The channel models include free space path loss model, shadowing fading, frequency selective fading, and flat fading.

[0067] Radio frequency signals After passing through the wireless channel and interference, the radio frequency signal received by the receiver is : in, For the total received signal at the receiving end, The effective signal components after channel evolution. Background Gaussian white noise, The active interference signal component is represented by t, which is time.

[0068] The theoretical model for single-tone interference patterns is as follows: in, For interference power, The linear conversion coefficients for the interference-to-information ratio are... To interfere with the center frequency, It is a random phase.

[0069] The theoretical model for broadband noise interference patterns is as follows: in, It is a zero-mean Gaussian white noise sequence, and the other parameters are defined in the same way as the single-tone interference pattern.

[0070] The theoretical model for the frequency modulation interference pattern is as follows: in, For frequency modulation slope, The integral term of the noise signal. It is a random phase.

[0071] The theoretical model for amplitude-modulated noise interference patterns is as follows: in, For carrier amplitude, This is a baseband noise signal. This is the phase jitter term; the other parameters are defined in the same way as in the aforementioned model.

[0072] The channel impulse response formula for the free-space path loss model is as follows: in, It is amplitude attenuation. It is path loss (dB). It is the Dirac function. It is a time delay variable. It is the propagation delay. It is distance (meters). It's the speed of light. It is the carrier wavelength.

[0073] The channel impulse response formula for the shadow fading pattern is: in, For random occurrences of shadow fading, The standard deviation of the shadow (dB) reflects the strength of the occlusion fluctuations.

[0074] The channel impulse response formula for frequency-selective fading is: in, For the number of multipaths, For the first Complex gain of the stripe diameter, It is the first Phase shift introduced by multipath, where The time delay is for each path, and the time delay spread is not negligible.

[0075] The channel impulse response formula for a flat fading pattern is: in, For random amplitudes that follow a Rayleigh distribution, For uniformly distributed random phases, This represents the average propagation delay.

[0076] Step S4: Model the signal receiving link, corresponding to the signal receiving module, such as... Figure 11 As shown, this is used to receive radio frequency signals and complete the analog-to-digital conversion. It outputs baseband binary data packets through demodulation and synchronization processing. The meaning of each symbol and the specific process are as follows: Generate and transmit carrier frequency Synchronized local The carrier signal will receive the radio frequency signal. Multiply each of these components by the carrier wave to complete quadrature demodulation.

[0077] A 6th-order Butterworth low-pass filter with a cutoff frequency of 320MHz was used to perform zero-phase filtering on the demodulated signal to remove the high-frequency components at twice the carrier frequency during the down-conversion process. The baseband signal was extracted, combined into a complex baseband signal, and multiplied by 2 for amplitude compensation to offset the amplitude attenuation during the down-conversion process, thus obtaining the complex baseband signal. .

[0078] Will According to the transmitter Reshape into a two-dimensional matrix in column-major order, and remove its leading edge. Okay, we get the time-domain matrix after removing the cyclic prefix. For the time-domain matrix after removing the cyclic prefix... Perform an FFT along the subcarrier dimension to recover the time-domain signal into a frequency-domain signal. Based on the pilot spacing at the transmitting end Determine the pilot symbol positions on the time axis and extract valid positions not exceeding the number of pilot symbols. Initialize a complex zero matrix with the same dimension as the received frequency domain matrix to store the channel estimation results for all subcarriers and all symbols. Extract the received frequency domain symbol corresponding to the pilot position for each subcarrier, divide it by the known transmitted pilot sequence to achieve LS channel estimation, and obtain the complex channel frequency domain response.

[0079] Zero-forcing equalization is performed on the carrier signal based on the obtained complex channel frequency domain response to eliminate inter-symbol interference. An equalizer coefficient matrix is ​​constructed to compensate for channel frequency domain distortion, resulting in a frequency domain symbol matrix after distortion removal. The core parameters of the pilot insertion at the transmitter are then reused. , By reverse-engineering the data position, the original data sequence can be extracted. Using 0 as the decision threshold, the complex symbols are respectively... The real and imaginary parts are compared using binary rules, with positive values ​​being assigned 1 and non-positive values ​​being assigned 0, resulting in... The two bit sequences are combined and reassembled to obtain a complete binary bit stream. , length is .

[0080] Based on the same random seed as the transmitter. The generated length is The random permutation sequence should be completely consistent with the permutation sequence during interleaving at the transmitter.

[0081] Construct the inverse permutation of the permutation sequence, and Rearrange the bits according to the inverse permutation sequence to obtain the original deinterleaved bitstream. .

[0082] Generator polynomial based on transmitter convolutional coding , With a constraint length of 7, construct a convolutional code network diagram and calculate... The cumulative path length is used to backtrack based on the backtracking depth to obtain the most probable information bit path. The last 6 bits of the decoding result are removed to obtain the final received bits. .

[0083] Step S5 involves modeling the data processing and evaluation modules, collecting and processing data, executing a Monte Carlo loop to form multiple experimental control groups, recording underlying indicators such as bit error rate and outage probability, obtaining a mapping dataset from interference parameters to fusion error, and then basing the results on the fusion error. The anti-interference performance is graded, and an evaluation model is encapsulated using the Naive Bayes algorithm, followed by model reliability verification.

[0084] Step S501: The data processing module models the data packets, unpacks them, and cleans them. The specific process is as follows: The received data packets are unpacked, and the headers and tails of the corresponding layers are removed layer by layer from bottom to top to restore the target coordinate observation values.

[0085] The data obtained within each communication time slot period is filtered, and the following judgment thresholds are set: in, The target coordinate observation values ​​obtained from unpacking, For the true coordinates of the target, This represents the standard deviation of the measurement error, which follows a normal distribution, when the UAV acquires the target coordinates. Only when all three of the above inequalities hold true is the set of coordinate data included in the valid dataset of the same period.

[0086] Step S502: Evaluation module modeling, build an evaluation model for the anti-interference performance of UAV ad hoc networks, based on fusion error. The anti-interference performance is graded and encapsulated into an evaluation model using the Naive Bayes algorithm.

[0087] The number of drones launched is shelf, The true coordinates of the target.

[0088] In an ideal scenario with no interference and a zero bit error rate, the amount of data in a single data transmission is the same as the number of drones launched. The drones within the network receive the first... The target coordinates broadcast by the drone The three-dimensional spatial coordinate measurement errors respectively satisfy: in, It is the first drones Coordinate measurement error, It is the first drone Coordinate measurement error, It is the first drone Coordinate measurement error, for The standard deviation of the normal distribution it follows for The standard deviation of the normal distribution it follows for The standard deviation of the normal distribution it follows, and , , and The following conditions must be met: but , , The value is their actual coordinates plus the corresponding measurement error: Let the predicted coordinates be Then the x-coordinate residual for: residual The standard deviation is: Similarly, we can obtain: The fusion error is then: The actual communication scenario is similar to the ideal situation described above, but there are two differences.

[0089] The first difference is the standard deviation of the coordinate error obtained within a period. It's no longer just simple measurement error; it also includes errors caused by transmission bit errors due to interference, channel fading, and noise during transmission. In the experiment, the measurement error was caused by hardware, while the bit errors generated during transmission were caused by channel noise and interference. These two are independent of each other. This represents the standard deviation of the coordinate error caused by the transmitted bit error within this period, because and If they are independent, they can be directly superimposed. , satisfy , It is the sum of two errors, so we don't need to discuss the two errors separately when processing the data; we can first make an overall assessment. ,and Given a known quantity, we can further obtain It is important to note that It cannot be directly measured; here it represents the relationship between channel noise, interference, and the final coordinate error. However, we can use... By reverse reasoning , The dimensions of the coordinates are the same, and the result it represents is independent of whether the coordinates are floating-point or integer types. Both floating-point and integer types can be written as bit streams, and they are essentially the same. It is independent of the number of bits in the coordinate data where errors occur, because the errors themselves are randomly generated on the corresponding bit stream. Furthermore, during encoding, the bit stream lengths of the three spatial coordinate components of the coordinate data are set to be the same, so their error probabilities are the same and are independent of the number of bits in the coordinate data where errors occur.

[0090] The second difference is In The issue is no longer the number of drones activated. Due to communication interruptions caused by interference, some activated drones are unable to successfully transmit the measured data to the cluster head. Therefore, the amount of data in each data transmission cycle is different. An effective node ratio coefficient can be set. This is used to represent the ratio of the number of nodes actually participating in the fusion to the number of nodes that started during the current cycle. The proportion. Here, a valid node refers to a node whose data packets can be received by the cluster head, and whose received, unpacked data meets the data filtering threshold mentioned in step S501. It cannot be obtained directly; it is affected by many factors such as the distribution of drones within the network, the power of interference, the location of the jammer, and weather conditions. It characterizes the relationship between communication interruptions and the amount of coordinate data that can participate in calculating the fusion error. It is not a simple ratio of the number of effective nodes, nor is it a comprehensive coefficient weighted by link quality for different nodes. It cannot be directly calculated; it only serves a representational purpose. However, we can roughly determine its value range because communication interruptions only occur during the process of other members transmitting data to the cluster head. The cluster head will transmit all data back to the ground base station through a secure frequency band. The data measured by the cluster head itself during this process is definitely valid. Therefore, the number of drones that can effectively transmit data must be greater than 1. ,Right now When all influencing factors are nearly ideal, the number of nodes that can effectively transmit data is the number of nodes under ideal conditions. ,Right now ,so .

[0091] Therefore, the fusion error in actual communication scenarios for: Why not Normalization is performed to obtain the anti-interference performance evaluation index. Fusion error The worst-case scenario is that only the cluster head data is correct, in which case... The data from the cluster head only contains measurement errors. The threshold for data filtering in step S501 is... ,So The maximum value is:

[0092] The maximum value of the fusion error is: but The larger the R value, the greater the fusion error. The larger the proportion, that is The larger the value, the weaker the anti-interference performance of the network; the smaller the R value, the lower the fusion error. The smaller the proportion, that is The smaller the value, the stronger the anti-interference performance of the network, and the anti-interference performance can be classified into levels: when At that time, the self-organizing network of drones has strong anti-interference performance.

[0093] when At that time, the anti-interference performance of the UAV self-organizing network was in the middle stage.

[0094] when At that time, the anti-interference performance of the drone self-organizing network was weak.

[0095] when At that time, the self-organizing network of drones does not have anti-interference capabilities.

[0096] According to the interference pattern (single-tone interference, broadband noise interference, noise FM interference, noise AM interference, and no interference), channel fading pattern (free space path loss, shadowing fading, frequency-selective fading, and flat fading), modulation method (QPSK, 16QAM, 32QAM), and number of UAVs (3, 5, 7), steps S1, S2, S3, S4, and S501 are repeated to form an experimental control group. Each group's results are run independently 100 times, and underlying indicators such as bit error rate and outage probability are recorded to obtain a mapping dataset from interference parameters to fusion error. Based on the anti-interference performance classification, an evaluation model is encapsulated using the Naive Bayes algorithm.

[0097] Step S503: Experimental model of UAV self-organizing network. Compare the experimental results with the simulation results to verify the reliability of the model.

[0098] Following steps S1, S2, S3, S4, and S501, 100 experimental samples are randomly generated. These 100 sets of data are then input into the UAV self-organizing network anti-interference performance evaluation model, allowing the model to automatically evaluate the anti-interference performance of the UAV network and compare it with the actual simulation results to verify the model's reliability.

[0099] It should be noted that any parts not disclosed or specifically described in this invention are existing technology or conventional configurations, and their specific structures and working principles will not be elaborated further. In this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0100] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. It can be applied to various fields suitable for the present invention. Other modifications can be readily implemented by those skilled in the art. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and examples shown and described herein.

Claims

1. A method for evaluating the anti-interference performance of unmanned aerial vehicle (UAV) self-organizing networks, characterized in that, include: Several test groups are constructed according to preset parameter combinations to obtain the bit error rate, interruption probability and effective coordinate observation data of the initial baseband binary data packets of each node in the UAV self-organizing network in each test group after interference is applied. The preset parameters include multiple interference patterns, multiple interference parameters, multiple channel fading patterns, multiple modulation methods and different numbers of UAVs. The baseband binary data packets include MAC header, sequence number, IP header, UDP header and measured data coordinates. The initial baseband binary data packets in each test group are converted into radio frequency (RF) analog signals and transmitted. Controllable electromagnetic interference with corresponding interference patterns and parameters is applied to the RF analog signals, and channel fading processing is performed. The RF signals after interference and channel fading processing are received. The process of RF signal transmission, interference, and reception also includes the networking process of the UAV ad hoc network under dynamic topology changes. The networking process includes distributed cluster head election and dynamic node entry and exit. The interference parameters include interference-to-signal ratio, interference frequency, and interference bandwidth. The RF signals after interference and channel fading processing are received, demodulated, and synchronized to restore the interfered baseband binary data packets. The bit error rate, outage probability, and effective coordinate observation data are calculated based on the interfered baseband binary data packets and the initial baseband binary data packets. The fusion error of the network is calculated based on the bit error rate, number of UAVs and effective coordinate observation data of each test group. A mapping dataset from interference parameters to fusion error is established. The interference parameters of the scenario to be evaluated are matched with the mapping dataset to obtain the fusion error of the corresponding UAV self-organizing network. The anti-interference performance of the UAV self-organizing network is calculated based on the fusion error, thereby completing the evaluation.

2. The method for evaluating the anti-interference performance of a UAV self-organizing network according to claim 1, characterized in that... : Interference patterns include single-tone interference, broadband noise interference, FM noise interference, AM noise interference, and no interference added; Channel fading patterns include free space path loss, shadowing fading, frequency-selective fading, and flat fading. The modulation method includes one of QPSK, 16QAM, or 32QAM.

3. The method for evaluating the anti-interference performance of a UAV self-organizing network according to claim 1, characterized in that... The effective coordinate observation data are the target coordinate observation values ​​parsed from the restored baseband binary data packets, and the deviation between the target's true coordinates and the actual coordinates is within the preset measurement error standard deviation range.

4. The method for evaluating the anti-interference performance of a UAV self-organizing network according to claim 1, characterized in that... The demodulation and synchronization processing includes: Generate local in-phase and quadrature carriers synchronized with the transmitter carrier, and multiply the received radio frequency signal by the two carriers respectively to complete the downconversion, thereby obtaining the I / Q intermediate frequency intermediate signals; The intermediate frequency signals of the I / Q channels are filtered by a zero-phase low-pass filter to remove high-frequency components at twice the carrier frequency, thus obtaining the filtered I / Q baseband signals. The I / Q baseband signals are combined into a complex baseband signal and multiplied by 2 for amplitude compensation. The number of complete OFDM symbols in the received signal is calculated and the complex baseband signal is truncated to an integer multiple of the length of the complete symbol. The signal is then reshaped into an OFDM time-domain matrix with a cyclic prefix according to the column priority rule. The cyclic prefix in the OFDM time-domain matrix is ​​removed, and a Fast Fourier Transform is performed along the subcarrier dimension on the de-cyclic prefixed time-domain matrix to obtain the received frequency-domain matrix. The pilot symbol positions in the received frequency-domain matrix are determined according to the pilot insertion rules preset by the transmitter. The received frequency-domain value corresponding to the pilot position is extracted and divided by the known pilot symbol sequence from the transmitter to obtain the channel frequency-domain response at the pilot position. A minimum threshold is set, and channel frequency-domain responses with amplitudes less than the minimum threshold are replaced with the minimum threshold. For the channel frequency domain response at the pilot position of each subcarrier, linear interpolation is performed on the time axis to complete the channel frequency domain response at the non-pilot symbol positions, resulting in a complete channel frequency domain response matrix covering all subcarriers and all symbols; Construct a zero-forcing equalizer coefficient matrix with the reciprocals of the elements of the complete channel frequency domain response matrix as coefficients, and multiply the received frequency domain matrix element by element with the zero-forcing equalizer coefficient matrix to obtain the frequency domain symbol matrix after eliminating channel distortion. Based on the pilot insertion parameters of the transmitting end, the position of the data symbol in the frequency domain symbol matrix is ​​located in reverse and the original data symbol sequence is extracted; with 0 as the decision threshold, the real part and imaginary part of each complex symbol in the original data symbol sequence are binary decided respectively, and the interleaved binary bit stream is obtained by recombination; A random permutation sequence is generated based on a random seed identical to that of the transmitter. The inverse permutation of the random permutation sequence is constructed. The interleaved binary bit stream is rearranged according to the inverse permutation to obtain the deinterleaved convolutionally coded bit stream. Based on the constraint length, code rate, and generator polynomial of the transmitter convolutional coding, a convolutional code state network diagram is constructed. The deinterleaved convolutional coded bitstream is then subjected to Viterbi decoding. The last 6 bits of the decoding result are removed to obtain the original baseband binary data packet.

5. The method for evaluating the anti-interference performance of a UAV self-organizing network according to claim 1, characterized in that... Each group of experiments shall have at least 100 Monte Carlo cycles.

6. The method for evaluating the anti-interference performance of a UAV self-organizing network according to claim 1, characterized in that... The criterion for judging the anti-interference performance R of the UAV self-organizing network is as follows: when At that time, the self-organizing network of drones has strong anti-interference performance; when At that time, the anti-interference performance of the UAV self-organizing network was moderate. when At that time, the anti-interference performance of drone self-organizing networks was weak; when At that time, the self-organizing network of drones does not have anti-interference capabilities.

7. The method for evaluating the anti-interference performance of a UAV self-organizing network according to claim 1, characterized in that... The method for converting the baseband binary data packet into a radio frequency signal is as follows: The initial constraint length is 7, the code rate is 1 / 2, and the generator polynomial is... and The convolutional encoder inputs baseband binary data bit by bit and outputs encoded bits, and finally inputs 6 all-zero tail bits to reset the encoder. A random permutation sequence is generated based on a random seed consistent with the receiving end, and the convolutionally encoded bitstream is interleaved and rearranged according to the permutation sequence. The interleaved bitstream is grouped and mapped into complex symbols according to the modulation scheme and then normalized in power. The modulation scheme includes one of QPSK, 16QAM, and 32QAM. Construct a full 0 OFDM frequency domain matrix with dimensions equal to the number of points required for the Fast Fourier Transform multiplied by the total number of symbols, and then assign the normalized complex symbols to the effective data subcarrier positions in sequence. Based on the same random seed, pilot sequences with the same data symbol power are generated and uniformly inserted into the specified positions of the OFDM frequency domain matrix at preset pilot intervals. Perform an inverse fast Fourier transform along the subcarrier dimension on the OFDM frequency domain matrix after inserting pilots to obtain the OFDM time domain baseband signal matrix; The last few sampling points of each OFDM time-domain symbol are extracted as a cyclic prefix, concatenated to the beginning of the corresponding symbol, and then reshaped into a one-dimensional serial time-domain transmitted signal according to the column priority rule.

8. The method for evaluating the anti-interference performance of a UAV self-organizing network according to claim 1, characterized in that... Before converting the baseband binary data packets of the UAV self-organizing network nodes into radio frequency analog signals and transmitting them, scene initialization and network initialization are also performed to build the UAV self-organizing network communication architecture. The scenario initialization includes: constructing an autonomous electromagnetic clean reconnaissance scenario of a preset size, deploying a programmable signal jammer, a ground data transmission station, a switch and a control console, and reserving a dedicated downlink communication frequency band between the ground base station and the UAV; The network initialization includes: deployment of the TTNT data link protocol stack for each node, preset of interface parameters at each layer, node ID allocation, IP address planning, TDMA time slot allocation, and distributed cluster head election.

9. The method for evaluating the anti-interference performance of an unmanned aerial vehicle (UAV) self-organizing network according to claim 8, characterized in that... During the process of converting baseband binary data packets of UAV self-organizing network nodes into radio frequency analog signals and transmitting them, it is possible to dynamically enable and disable interference nodes. The dynamic network entry and dynamic network exit include node network entry identity verification, normal network exit application response, abnormal network exit timeout monitoring, and dynamic scheduling, reclamation and reallocation of the corresponding node's time slot resources and IP addresses.

Images