A cross-layer resource allocation method for reliable communication of unmanned aerial vehicle formation

Abstract

The present application relates to the field of wireless communication network, and specifically relates to a cross-layer resource allocation method for reliable communication of unmanned aerial vehicle formation, comprising: a long machine node acquires link state information and transmission demand information of each wingman node; the long machine node determines the optimal time slot length and time slot allocation for the wingman node according to the received link state information, with the optimization target of maximizing the throughput of the unmanned aerial vehicle formation system; the long machine node calculates the transmission priority of each wingman node according to the received transmission demand information and issues the time slot allocation result, and each wingman node performs data transmission in turn according to the transmission priority ranking. The present application can effectively prevent the single collision characteristic of the wireless channel, avoid the collision or conflict of multiple data packets in the same time transmission, and ensure the reliability of data transmission between the long machine and the wingman in the unmanned aerial vehicle formation.

Description

Technical Field

[0001] This invention relates to the wireless communication network domain, and more specifically to a cross-layer resource allocation method for reliable communication of UAV swarms. Background Technology

[0002] With the rapid development of science and technology and the economy, drone swarms have experienced explosive growth in scale and complexity, posing significant challenges to the reliable operation of swarm systems. During missions, drone swarms exchange data such as mission planning, flight status, and intelligence information through information interaction. Therefore, smooth information transmission and efficient information flow are the core support for drone swarms and crucial for their successful operation. Communication is the foundation for information interaction and collaboration between nodes within a drone swarm, requiring reliable data exchange within the network. However, due to the single collision domain characteristic of wireless channels, drone swarm networks are typically sensitive to collisions. When multiple drones operate simultaneously in a swarm, sharing limited wireless time slots, collisions or conflicts are prone to occur, potentially leading to data loss, retransmission, corruption, or even system crashes, thus reducing data reliability.

[0003] For data reliability in UAV swarms, the key focus is on multi-access scheduling strategies. Currently, hybrid CSMA / TDMA scheduling is gaining attention. It combines CSMA and TDMA, fully utilizing the advantages of both mechanisms while mitigating their weaknesses, thus significantly enhancing network communication performance. However, hybrid scheduling mechanisms that adaptively switch between CSMA and TDMA based on network load levels cannot accurately distinguish network load levels. Furthermore, within the fixed time slots allocated by TDMA, when a node has no data service needs or has many idle time slots, other nodes with higher data service demands can still utilize these slots via CSMA, but this still cannot avoid the shortcomings of the CSMA mechanism. Therefore, most hybrid scheduling mechanisms rely on strict and fixed time slot allocation, are insensitive to changes in network load and nodes, and cannot cope with UAV swarm combat scenarios.

[0004] Currently, the default data packet size in the CSMA / TDMA hybrid scheduling mechanism is the same, and the time slot length is fixed. However, in reality, the data packet sizes are inconsistent. Therefore, reliable data packet transmission in UAV formations when the data packet lengths are inconsistent is a problem that urgently needs to be solved in this field. Summary of the Invention

[0005] To address the issue of reliable data transmission in UAV formations when data packet lengths are inconsistent, this invention proposes a cross-layer resource allocation method for reliable communication in UAV formations, specifically including the following steps:

[0006] The lead node obtains the link status information and transmission requirement information of each wingman node;

[0007] The lead node determines the optimal time slot length and time slot allocation for the wingman nodes based on the received link status information, with the optimization goal of maximizing the throughput of the drone formation system.

[0008] The lead node calculates the transmission priority of each wingman node based on the received transmission request information and sends out the time slot allocation results. Each wingman node then transmits data in sequence according to the transmission priority.

[0009] This invention addresses the CSMA / TDMA hybrid scheduling mechanism in UAV formations. It considers the reliable transmission of data packets in UAV formations when packet lengths are inconsistent. During the CSMA phase, it improves the backoff mechanism based on the physical layer MCS mechanism, reducing the collision probability to some extent. Furthermore, during the time slot allocation phase, it maximizes the throughput of the formation system by rationally designing and allocating time slot resources for the CSMA / TDMA hybrid scheduling mechanism, avoiding data collisions, improving the overall data throughput of the system, and simultaneously enhancing the transmission efficiency of the formation system, thus ensuring reliable data transmission within the formation system. Attached Figure Description

[0010] Figure 1 This is an application scenario diagram of the accurate, closed-loop cross-layer resource allocation method for reliable communication in UAV formations according to the present invention;

[0011] Figure 2 This is a diagram of the time slot request frame structure constructed by a wingman node with transmission needs during the CSMA phase in this invention.

[0012] Figure 3 This is a diagram of the time slot allocation frame structure constructed by the primary node during the time slot allocation phase of this invention.

[0013] Figure 4 This is a flowchart of a cross-layer resource allocation method for reliable communication in UAV formations, as described in an embodiment of the present invention.

[0014] Figure 5 This is a schematic diagram illustrating the principle of data transmission between the lead node and the wingman node within a superframe in this invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] This invention proposes a cross-layer resource allocation method for reliable communication in UAV swarms, specifically including the following steps:

[0017] The lead node obtains the link status information and transmission requirement information of each wingman node;

[0018] The lead node determines the optimal time slot length and time slot allocation for the wingman nodes based on the received link status information, with the optimization goal of maximizing the throughput of the drone formation system.

[0019] The lead node calculates the transmission priority of each wingman node based on the received transmission request information and sends out the time slot allocation results. Each wingman node then transmits data in sequence according to the transmission priority.

[0020] Figure 1 Specific application scenarios of the present invention are given, in Figure 1 In this scenario, a drone formation consists of a lead drone node and multiple wingman nodes. The lead drone node forwards commands from the control center to each wingman node. The wingman nodes perform attitude changes according to the commands, and the wingman nodes transmit the data obtained from reconnaissance and surveillance back to the command center through the lead drone node. In this scenario, it is necessary to ensure the security and reliability of the transmission between the wingman nodes and the lead drone node, while also ensuring the transmission efficiency between the lead drone node and the wingman nodes.

[0021] like Figure 5 In this invention, the wireless channel superframe is divided into a CSMA phase, a time slot allocation phase, and a TDMA phase, wherein:

[0022] During the CSMA phase, each wingman node that successfully joins the network counts the link status and transmission requirements with the lead node and constructs a time slot request frame. During the CSMA channel access phase, in order to avoid unfair competition caused by the randomness of the backoff time of the traditional binary backoff mechanism, wingman nodes with transmission requirements use a backoff mechanism improved based on the physical layer MCS mechanism to take turns accessing the channel and initiate time slot requests to the lead node.

[0023] During the time slot allocation phase, the lead node extracts the link status and transmission requirements of each wingman node from the received time slot request frame, and determines the time slot and transmission priority allocated to each wingman node in the TDMA phase accordingly. The lead node constructs a time slot allocation frame and broadcasts the time slot allocation result to each wingman node.

[0024] During the TDMA phase, each wingman node accesses the channel sequentially according to the time slot allocation results in the time slot allocation frame to perform data transmission.

[0025] This embodiment proposes a specific implementation process for a cross-layer resource allocation method for reliable communication in UAV swarms, such as... Figure 4 This includes the following steps:

[0026] 101. During the CSMA phase, each wingman node that successfully joins the network counts the link status and transmission requirements with the lead node, and constructs a time slot request frame.

[0027] The wireless channel superframe is divided into three phases: CSMA, time slot allocation, and TDMA. During the CSMA phase, each successfully joined wingman node statistically analyzes the link status and transmission requirements with the lead node, and uses this information to construct a time slot request frame. This process includes the following steps:

[0028] 1) Wingman Node Link Status Statistics

[0029] The link status of wingman node i includes transmission rate, received signal strength, and transmission distance, where:

[0030] The transmission rate R of wingman node i i The physical layer MCS mechanism selects an appropriate rate index value (MCS) based on the wireless link quality of wingman node i. i And index to obtain the transmission rate R i ;

[0031] Received signal strength RSSI of wingman node i i When wingman node i receives a beacon frame from the lead aircraft node, the wingman node's wireless network card converts the received radio wave signal into an electrical signal and calculates the received signal strength RSSI through the signal processing circuit. i value;

[0032] Transmission distance d of wingman node i i The wireless path between wingman node i and the lead node is considered a free-space path. The transmission distance is estimated using a free-space path loss model. The estimation process includes:

[0033] The free space path loss model is as follows:

[0034]

[0035] Where PL is the path loss, P tx d is the transmission power, f is the transmission distance, f is the signal frequency, and c is the speed of light;

[0036] The distance between wingman i and lead aircraft is calculated using path loss:

[0037]

[0038] 2) Statistics on transmission requirements of wingman nodes

[0039] The transmission requirements of wingman node i include queue length Q. iData priority weights and the distribution of data packet lengths in the cache queue, where:

[0040] Data priority weight ω i Represented as:

[0041]

[0042] Where k represents the priority value of the data, the smaller the value, the higher the priority, and vice versa; This represents the percentage of data packets with priority k in the cache queue of wingman node i.

[0043] In drone formations, the packet lengths of various data packets differ. Control signal packets are typically shorter, while sensor data packets are usually longer. Therefore, the packet lengths do not follow a simple exponential distribution, but can be estimated as a superposition of multiple exponential distributions. Treating the packet length distribution as an Irish distribution, its probability distribution function is:

[0044]

[0045] Where M is the shape parameter, representing the number of exponential stages involved in the distribution; λ is the rate parameter, equivalent to the rate parameter in the exponential distribution. i Let represent the rate parameter of the i-th wingman node; p represents the arrival probability of the data packet, which follows a Bernoulli process; and x represents the packet length.

[0046] 3) Construct a time slot request frame

[0047] Wingman nodes report time slot requests to the lead node by constructing a time slot request frame at the MAC layer. This frame only operates at the MAC layer. After receiving this frame, the lead node extracts the key information of each wingman node and then discards the frame. Wingman nodes with transmission needs fill the payload portion of this frame with link state information and service transmission requirement information when constructing it. The structure of the constructed time slot request frame is as follows: Figure 2The frame structure used in this embodiment is the 802.11 frame structure, which includes Frame Control, Duration ID, address fields (containing up to 4 address fields, namely Address1, Address2, Address3, Address4), Sequence Control, Frame Body, and Frame Check Sequence (FCS). In this embodiment, the Frame Body carries the link status information and transmission requirement information of the wingman node, including the MCS index value (MCS), Received Signal Strength Indicator (RSSI), transmission distance (d), queue length (Q), data priority weight (w), and parameters of the probability distribution function (M / λ / p).

[0048] 102. Wingman nodes with transmission needs use a backoff mechanism based on the physical layer MCS mechanism to take turns accessing the channel and send time slot request frames to the lead node.

[0049] In 802.11 wireless networks, the MCS index value reflects the link quality between network devices. A higher MCS index value indicates better link quality between nodes, which can support high data transmission rates; when the link quality is poor, in order to ensure transmission success rate, it is necessary to reduce the data transmission rate and select a lower MCS index value.

[0050] During the CSMA channel access phase, wingman nodes with transmission needs take turns accessing the channel and initiating time slot requests to the lead node. To avoid unfair competition caused by the randomness of traditional binary backoff time, an improved backoff mechanism based on the physical layer MCS mechanism is used.

[0051] The MCS index value reflects the link quality between network devices. Utilizing the physical layer MCS mechanism, different backoff window intervals are set for different MCS index values, meaning different backoff intervals are established for links with different quality levels. Each wingman node selects its own backoff count within the corresponding backoff interval based on its chosen MCS index value according to its link quality, thus reducing the probability of collisions to some extent. As an optional implementation scheme, the backoff interval is divided as follows:

[0052] Table 1. Division of Retreat Zones

[0053]

[0054]

[0055] 103. During the time slot allocation phase, the lead node allocates time slots and establishes transmission priorities based on the transmission demand information and link status information of each wingman node in the time slot request frame.

[0056] The lead node receives time slot request frames and extracts the link status information and transmission requirement information of each wingman node. The packet length distribution of each wingman node is as follows:

[0057]

[0058] Therefore, the average packet length of each wingman node is For the arrival rate, T S The duration of a time slot.

[0059] The lead node uses the extracted link state information and transmission demand information to determine the time slot length T for the TDMA stage. S The number of time slots allocated to each wingman node and its transmission priority. Specifically, this includes the following steps:

[0060] 1) Time slot allocation: Determine the time slot length T S and the number of time slots β allocated to the wingman node i .

[0061] The length of a single time slot in the TDMA phase is T. S Then the total number of time slots A in the TDMA stage is:

[0062]

[0063] Where T represents the duration of a superframe, T CSMA T represents the duration of the CSMA phase in a superframe. TSAP Indicates the duration of the time slot allocation phase. This indicates rounding up to the nearest integer.

[0064] The lead node allocates the number of time slots for the TDMA phase based on the amount of data that each wingman node needs to transmit. The number of time slots β required for each wingman node to successfully transmit is determined. i for:

[0065]

[0066] Among them, Sum i This indicates the total amount of data to be transmitted in the buffer queue of wingman node i, which has transmission requirements:

[0067]

[0068] Where, δ iLet represent the packet loss rate of the link between wingman node i and the lead aircraft node. If the bit error rate of the link is τ, the transmission of each data bit between wingman node i and the lead aircraft node follows a binomial distribution. Therefore, the packet loss rate δ of the link between wingman node i and the lead node is... i Represented as:

[0069]

[0070] In UAV formation combat scenarios, to ensure the reliability of wireless communication, it is possible to maximize the utilization of data link layer time slots, i.e., to seek the optimal time slot length to maximize data transmission efficiency, at which point the entire formation system can achieve its maximum throughput. Therefore, an optimization function is constructed with maximizing the throughput of the entire formation system as the optimization objective:

[0071]

[0072] stC1:T=T CSMA +T TSAP +T TDMA

[0073]

[0074] C3:β aid ≥1 aid∈[1,N]

[0075] C4:μ i ≤R i i∈[1,N]

[0076] Where C1 represents the superframe length constraint, and T represents the length of a superframe; C2 represents the time slot number constraint in the TDMA stage; C3 represents the transmission fairness constraint, that is, each wingman node with transmission needs is allocated at least one time slot; C4 represents the congestion avoidance constraint, that is, the average arrival rate of each wingman node's buffer queue is not greater than its transmission rate to ensure the normal transmission of data in the node's buffer queue:

[0077]

[0078] Right now: This allows us to obtain the smallest slot value based on the link status information and transmission demand information of the wingman node i.

[0079] The Lagrange relaxation method can transform a complex optimization problem into a more easily solvable subproblem. By introducing Lagrange multipliers, constraints can be relaxed, decomposing the original problem into a manageable single-variable problem. Typically, an iterative approach is used; by adjusting the Lagrange multipliers, the constraints of the optimization problem can be flexibly controlled, thereby finding an approximation of the optimal solution. Therefore, in UAV formation, the Lagrange relaxation method can be used to select the optimal time slot length T. S To maximize the throughput of the entire formation while satisfying the data queue transmission requirements of each wingman node and the constraints of time slot resources, a Lagrange multiplier ζ is introduced to transform the constraints into penalty terms. The Lagrange function is then constructed:

[0080]

[0081] in: σ∈[1,N].

[0082] The Lagrange multiplier ζ is adjusted iteratively to gradually approximate the actual total time slot limit, thereby maximizing the Lagrange function; the iterative process of the Lagrange multiplier is as follows:

[0083]

[0084] Where, ζ (k) Let represent the Lagrange multiplier at the k-th iteration; ε represents the step size of the iteration.

[0085] The entire iterative process is as follows:

[0086] S1: Initially select the iteration step size ε and the time slot allocation β for each wingman node. i ×T S ;

[0087] S2: Calculate the Lagrangian function L(T) under the current time slot allocation. S ,ζ);

[0088] S3: Adjust the time slot allocation β for each wingman node. i ×T S Maximize the current Lagrange function;

[0089] S4: Update the Lagrange multipliers using the subgradient method and adjust the constraints;

[0090] S5: Repeat S2-S4 until the convergence condition is met.

[0091] After several iterations, the lead node will find an optimal time slot length T. s Time slot allocation β i To maximize the total system throughput while meeting the total time slot limit.

[0092] 2) Transmission priority

[0093] The transmission priority can be set by those skilled in the art according to the task requirements. For example, the transmission of service types with higher real-time requirements can be prioritized. Those skilled in the art can also set it according to the transmission quality of the link. For example, the transmission of nodes with better link quality can be prioritized.

[0094] As an optimal implementation, this embodiment considers the different data priority weights and link states of each wingman node with transmission needs, and assigns them different data transmission orders. This ensures that high-priority data transmission is guaranteed while prioritizing data transmission from wingman nodes with good link states, thereby maximizing the system's total throughput. Therefore, the transmission priority of each wingman node consists of two parts: service priority and link priority.

[0095] Business priority is the data priority weight ω in the wingman node's cache queue. i The normalization is expressed as:

[0096]

[0097] in, ω represents the service priority of wingman node i. i Let ω be the data priority weight of wingman node i, minω be the minimum data priority weight of wingman node i, and maxω be the maximum data priority weight of wingman node i.

[0098] Link priority is composed of the normalized MCS index value and the transmission distance weight, and is expressed as:

[0099]

[0100] in, For the link priority of wingman node i, MCS i Let be the MCS index value of wingman node i, maxMCS be the maximum value of the MCS index, and minMCS be the minimum value of the MCS index; d i Let be the distance between wingman node i and helm node i, and D be the maximum distance that wingman node i and helm node i can communicate over.

[0101] Therefore, the transmission priority of wingman node i is:

[0102]

[0103] The lead aircraft determines the transmission order of each wingman node by comparing the transmission priorities of each wingman node.

[0104] 104. The lead node constructs a time slot allocation frame and broadcasts the time slot allocation results to each wingman node.

[0105] The lead node constructs a time slot allocation frame and broadcasts the time slot allocation results to all wingman nodes; the constructed time slot allocation frame is as follows: Figure 3 As shown, the information carried by the frame body includes the TDMA begintime, time slot length, time slot number, and the length of a single time slot T for each wingman node in the TDMA phase. S The allocation table (Ts) and the TDMA endtime, where the length of a single time slot T for each wingman node during the TDMA phase are... S The allocation table looks up the corresponding allocation information based on the unique identifier (AID) of the wingman node. The unique identifier of the wingman node is followed by the slot number, start slot, and end slot of the wingman node.

[0106] During the TDMA phase, each wingman node accesses the channel sequentially according to the time slot allocation results in the time slot allocation frame to perform data transmission.

[0107] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A cross-layer resource allocation method for reliable communication in UAV swarms, characterized in that, Specifically, it includes the following parts: The lead node obtains the link status information and transmission requirement information of each wingman node; Based on the received link status information, the lead node determines the optimal time slot length and allocation for the wingman nodes with the goal of maximizing the throughput of the UAV swarm system. Within the current superframe, an optimization function is constructed with the goal of maximizing the throughput of the UAV swarm system to select the optimal time slot length and allocation for the current UAV swarm system. The optimization function is expressed as follows: Optimization function: Constraint C1: T = T CSMA +T TSAP +T TDMA C3:β aid ≥1aid∈[1,N] C4:m i ≤R i i∈[1,N] Where H represents the system throughput of the drone formation, and N is the number of wingman nodes in the drone formation; β i The number of time slots allocated for successful transmission by wingman node i; T S The optimal time slot length for the TDMA stage; R i Let be the transmission rate of wingman node i; T represents the duration of a superframe. CSMA T is the duration of the CSMA phase in a superframe. TSAP To represent the duration of the time slot allocation phase, T TDMA β represents the duration of the TDMA phase in the superframe; A represents the total number of time slots in the TDMA phase; β represents the duration of the TDMA phase in the superframe. aid The number of time slots allocated to wingman nodes with transmission needs; μ i The average arrival rate of data packets for wingman node i; The lead node calculates the transmission priority of each wingman node based on the received transmission request information and sends out the time slot allocation results. Each wingman node then transmits data in sequence according to the transmission priority.

2. The cross-layer resource allocation method for reliable communication in UAV swarms according to claim 1, characterized in that, The link status information of wingman nodes with transmission needs includes transmission rate, received signal strength, and transmission distance; the transmission demand information includes queue length, data priority weight, and packet length distribution of buffer queue data packets.

3. The cross-layer resource allocation method for reliable communication in UAV swarms according to claim 1, characterized in that, When wingman nodes compete for the channel during the access phase, they take turns accessing the channel based on the physical layer MCS mechanism. Different backoff window intervals are set for different MCS index values. The smaller the MCS index value, the larger the minimum value of the backoff interval and the longer the backoff interval.

4. The cross-layer resource allocation method for reliable communication in UAV swarms according to claim 1, characterized in that, The number of time slots β required for wingman node i to successfully transmit i Represented as: Among them, Sum i δ represents the total amount of data to be transmitted in the buffer queue of wingman node i. i This represents the packet loss rate of the link between wingman node i and the lead node.

5. A cross-layer resource allocation method for reliable communication in UAV swarms according to claim 1, characterized in that, By introducing the Lagrange multiplier ζ, the constraints of the optimization function are transformed into penalty terms to construct the Lagrange function. The optimal time slot length T is then selected using the Lagrange relaxation method. s Time slot allocation β i Specifically, it includes the following steps: S1. Initialize the iteration step size of the Lagrange multipliers and the time slot allocation for each wingman node; S2. Calculate the Lagrangian function under the current time slot allocation of each wingman node; S3. Adjust the time slot allocation of each wingman node with the goal of maximizing the current Lagrange function; S4. Update the Lagrange multipliers using the subgradient method and adjust the constraints. S5. Determine if the current Lagrangian function meets the convergence condition. If it does, output the current optimal time slot length T. s Otherwise, return to step 102.

6. A cross-layer resource allocation method for reliable communication in UAV swarms according to claim 5, characterized in that, The Lagrange function is expressed as: Among them, L(T) S ,ζ) denote the Lagrange function; T Sσ The length of a single time slot in the TDMA phase, set according to the transmission requirements of the σ-th wingman node, is denoted as: Let R be the average packet length of the σ-th wingman node. σ Let σ be the transmission rate of the σ-th wingman node, p represent the arrival probability of the data packet, and ζ be the Lagrange multiplier.

7. A cross-layer resource allocation method for reliable communication in UAV swarms according to claim 5, characterized in that, The process of updating the Lagrange multipliers using the subgradient method is represented as follows: Where, ζ (k) Let represent the Lagrange multiplier in the k-th iteration, and ε represent the iteration step size of the Lagrange multiplier.

8. A cross-layer resource allocation method for reliable communication in UAV swarms according to claim 1, characterized in that, The transmission priority calculation process for wingman nodes includes: Among them, G i G represents the transmission priority of wingman node i. i The smaller the value, the higher the priority; ω represents the service priority of wingman node i. i Let ω be the data priority weight of wingman node i, minω be the minimum data priority weight of wingman node i, and maxω be the maximum data priority weight of wingman node i. For the link priority of wingman node i, MCS i Let be the MCS index value of wingman node i, maxMCS be the maximum value of the MCS index, and minMCS be the minimum value of the MCS index; d i Let be the distance between wingman node i and helm node i, and D be the maximum distance that wingman node i and helm node i can communicate over.

9. A cross-layer resource allocation method for reliable communication in UAV swarms according to claim 8, characterized in that, The data priority weight of wingman node i is represented as follows: Where k represents the priority value of the data, the smaller the value, the higher the priority; This represents the percentage of data packets with priority k in the cache queue of wingman node i.

Images