A cooperative task allocation and air-ground channel access optimization method for reconnaissance scenarios

Abstract

The application discloses a kind of reconnaissance scene-oriented cooperative task allocation and air-ground channel access optimization method, step 1: determine task allocation model, and determine the task demand and user's own ability in model;Step 2: determine data transmission model;Step 3: determine the calculation method of task actual utility, and optimization is carried out;Step 4: two-stage optimization framework is used to solve optimization problem, for the first stage, determine the clustering strategy of user;Step 5: in the second stage, determine channel exploration strategy, and channel exploration strategy is converted into data receiving problem.The application can effectively shorten channel access time and significantly improve system actual total utility while meeting task capability demand.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, and in particular relates to a method for collaborative task allocation and air-to-ground channel access optimization for reconnaissance scenarios. Background Technology

[0002] In emergency reconnaissance scenarios, effective coordination among ground users is crucial for protecting life and property. During mission execution, key issues such as task allocation and channel access are typically involved. Regarding task allocation, different ground users possess varying capabilities, while reconnaissance missions often demand different capabilities. Inappropriate task allocation can easily lead to insufficient capability matching between ground users and the mission, affecting mission effectiveness and even causing mission failure if basic capability requirements are not met. Regarding channel access, considering the complexity of reconnaissance missions, ground users typically need to receive mission-related data in clusters before mission execution to improve effectiveness. However, in disaster areas, the status information of air-to-ground communication channels is often difficult to obtain accurately beforehand. If the command center were to comprehensively explore all available channels and precisely allocate channels to each cluster head before mission execution, it would incur significant channel exploration overhead, impacting the overall efficiency of the reconnaissance mission.

[0003] Existing research has explored the issues of task allocation and channel access to some extent. Regarding task allocation, related studies mainly model and solve from the perspective of capability coordination or resource optimization, but they do not adequately consider the comprehensive requirements of multiple capabilities needed by a task, making it difficult to simultaneously meet the requirements of a task for multiple capability combinations. Regarding channel access, existing research mostly allocates channels based on statistical channel state information, while research on channel access strategies under unknown channel states is relatively limited, and there is a lack of integration between cluster data requirements. Summary of the Invention

[0004] The purpose of this invention is to address the problems of insufficient matching between task requirements and user capabilities in unknown environments and low efficiency of channel access decision-making, and to propose a collaborative task allocation and air-to-ground channel access optimization method for reconnaissance scenarios.

[0005] To achieve the objectives of this invention, this invention provides a method for cooperative task allocation and air-to-ground channel access optimization in reconnaissance scenarios, the method comprising:

[0006] Step 1: Determine the task allocation model, and determine the task requirements and user capabilities within the model;

[0007] Ground users and reconnaissance missions are randomly distributed throughout the network, represented as follows: and Users need to participate in the reconnaissance mission, and users participating in the same mission should cooperate in clusters to enhance the execution effect.

[0008] The collaborative relationships between users are influenced by both task requirements and the users' own capabilities; among which, task requirements are reflected in the comprehensive demands on multiple capabilities; let the set of all capability requirement types be . For any task Capability requirements are represented as vectors. ;

[0009] In order to portray different abilities in the mission The importance of introducing capability weights ;when When, it indicates a task. No ability required Meanwhile, for any task The ability weights satisfy ;Task The set of capability requirements is represented as ;The task The total value is expressed as Among them, the k-th type of ability in the task The corresponding value is defined as:

[0010] ;

[0011] Regarding user capabilities, it is assumed that users dispatched by the command center possess the ability to complete all tasks; for any user... Mastery ability is represented as a vector. ,in Indicates user Regarding ability The degree of mastery;

[0012] The degree of mastery is characterized by probability; the stronger a user's mastery of a certain ability, the higher the probability of independently completing the corresponding ability in a task; for a given task... When multiple users are capable of fulfilling a certain capability requirement, the overall probability of fulfilling that capability will be further increased; In the mission The probability of completion is expressed as:

[0013] ;

[0014] in, Indicates support for the task The user set;

[0015] Considering that task utility depends on the value of each ability and the probability of satisfying the corresponding ability, completing the task... The returns are expressed as:

[0016] ;

[0017] in, The indicator function is used to ensure that all required capabilities for the task are met.

[0018] ;

[0019] Additionally, considering that users capable of fulfilling a certain task are not located in the corresponding task area, the movement cost for users to move to the task area needs to be considered; assuming users... Need to move to task The central point, representing the user's mobility cost as ,in This represents the cost per meter a user moves. Indicates user With the task The distance between them.

[0020] Step 2: Determine the data transmission model.

[0021] To ensure that each cluster can acquire the data needed to perform its mission within a limited time, the command center allocates one drone and a certain number of available channels to each cluster. The drone set is represented as follows: The set of available channel resources for a cluster is represented as Each cluster selects a suitable channel from its assigned set of available channels for data reception; to prevent interference between clusters during data reception, there is no overlap between the sets of available channels assigned to different clusters.

[0022] Data requirements are related to the capability requirements of the task: for any task , if ability Data requirements are expressed as Then execute the task. The total data requirement of the cluster is expressed as To avoid duplicate downloads, each cluster's cluster head downloads data from the drone and forwards it to users within the cluster. The set of cluster heads is represented as... Because the members within a cluster are close to each other, the time it takes for the cluster head to receive data from the drone is much longer than the time it takes for users within the cluster to forward data to each other.

[0023] Therefore, this invention focuses on analyzing the data transmission problem between UAVs and cluster heads.

[0024] Specifically, for each cluster, block fading is used to model the channel, that is, in The channel state remains unchanged within a certain time; the time for each cluster head to probe one channel is... The cluster head first probes the channel to determine a suitable receiving channel. Once the channel is determined, the remaining time is used for data reception. It is assumed that the channels in the available channel set are independent and identically distributed. Considering the complexity of the environment in the reconnaissance scenario, Rayleigh fading is used to characterize the channel properties. Arbitrary Channel Channel gain The probability density function is expressed as:

[0025] ;

[0026] in, Reflects the average channel gain, and has , The path loss coefficient is estimated through a rough pre-deployment survey conducted before task allocation.

[0027] If a channel is used When receiving data, the receiving rate is expressed as:

[0028] ;

[0029] in, For channel bandwidth, For the drone's transmission power, It is additive white Gaussian noise.

[0030] Step 3: Determine and optimize the method for calculating the actual utility of the task;

[0031] To accomplish these tasks more effectively, on the one hand, it is essential to ensure that the basic capability requirements of all tasks are met, and that the match between user capabilities and task requirements is as close as possible. On the other hand, since each cluster head is unaware of the channel quality within the available channel set beforehand, it first determines its channel exploration strategy and then selects the optimal channel from the explored channels for data reception. Therefore, the task... The actual utility is calculated as follows:

[0032] ;

[0033] in, Indicates task The number of channel explorations corresponding to the cluster. Indicates that it has been explored The optimal channel among all channels This indicates the degree of impact of data reception on utility; considering that the tasks supported by each user are determined by their own clustering strategy, and the channels accessed by each cluster head are determined by the channel exploration strategy, the two strategies are represented as follows: and ;

[0034] By optimizing the clustering strategy for each user and the channel exploration strategy for each cluster head, the actual total utility is maximized; the optimization problem is expressed as:

[0035] ;

[0036] Among these constraints, the channel exploration time of the cluster head is ensured to be less than the block fading duration, thus reserving a certain amount of time for data reception.

[0037] Step 4: Use a two-stage optimization framework to solve the optimization problem. For the first stage, determine the user's clustering strategy.

[0038] The clustering strategy is modeled as a transferable utility alliance-forming game, and the game tuple is represented as follows: ;in, For users The game-theoretic utility; For a clustered set, denoted as ; Given the network topology; assume any cluster Able to meet the task Based on the basic capability requirements, the benefits that users within a cluster can obtain after completing the corresponding task are:

[0039] ;

[0040] In addition, cluster The movement cost incurred by an internal user to complete a task is represented as follows:

[0041] ;

[0042] cluster The expected utility obtained by internal users is:

[0043] ;

[0044] In the mission During execution, cluster Each user plays an important role, and each user is considered to have equal value.

[0045] Therefore, by distributing the expected utility equally among all users, the expected utility function for each user within the cluster is obtained as follows:

[0046] ;

[0047] During the process of forming clusters for ground users, a cooperation criterion is adopted to optimize the clustering;

[0048] For any user as well as If the user I prefer to join the cluster It should satisfy the following relationship:

[0049] ;

[0050] If user In cluster In the middle, the utility function Represented as:

[0051] ;

[0052] in For users The clustering strategy, namely , To exclude users Clustering strategies for other users; by adjusting the clustering strategy for each user, the expected total utility of all users is continuously improved, eventually resulting in a stable consortium partition;

[0053] Algorithmically, considering that the basic capabilities of each task need to be met, a preprocessing algorithm is determined to obtain the initial cluster structure; then, the cluster structure is adjusted through a coalition formation algorithm.

[0054] The algorithm steps for cluster preprocessing are as follows:

[0055] Step A1: Initialize all users into a free user set;

[0056] Step A2: Based on task value Determine the processing order from high to low;

[0057] Step A3: Extract tasks in sequence to determine the current unmet capability requirements;

[0058] Step A4: Select a user from the free users who best matches the task's ability requirements and add them to the corresponding task. If multiple users have the same matching degree, select the user closest to the task.

[0059] Step A5: Remove the user performing the task from the free user set;

[0060] Step A6: Repeat steps A2 to A5 until the basic capability requirements of all tasks are met, and finally output the current clustering structure. ;

[0061] The alliance formation algorithm is as follows:

[0062] Step B1: Input the current clustering structure ;

[0063] Step B2: Randomly generate the processing order, processing users who have not joined the cluster first, and then processing users who have joined the cluster;

[0064] Step B3: Extract users according to the processing order Try adding other clusters and calculate the clustering utility. If adding other clusters improves the user's game utility, then change the user's clustering strategy and update the clustering structure.

[0065] Step B4: Repeat steps B2 to B3 until a stable alliance partition is obtained.

[0066] Step 5: In the second stage, the channel exploration strategy is determined, which is then transformed into a data reception problem.

[0067] Considering that the available channel sets do not overlap and the channel resources are homogeneous across different clusters, a channel exploration strategy is selected for one cluster and then extended to other clusters. In this stage, each cluster needs to acquire as much data as possible, and the goal of channel exploration is to select a suitable channel from the explored channels to receive data. Therefore, the channel exploration strategy is transformed into a data reception problem, and the data reception rate of the cluster head is expressed as:

[0068] ;

[0069] in, For explored The maximum channel gain in the channel is calculated as follows: The optimization goal at this stage is to maximize the data reception rate of each cluster.

[0070] To reduce computational complexity, a 1-SLA-based channel exploration algorithm is determined. During channel exploration, the channel exploration stopping time is determined by comparing the data reception rate with a pre-calculated threshold value, thus achieving a better trade-off between channel exploration and data reception. The derivation of the threshold value is as follows:

[0071] ;

[0072] in, , , ; It is an exponential integral function, and its specific expression is: ;

[0073] The channel exploration algorithm based on 1-SLA is as follows:

[0074] Step C1: Elect a cluster head for each cluster And set the number of detections. ;

[0075] Step C2: Each cluster head Determine the maximum channel gain in the set of detected channels. and calculate ;

[0076] Step C3: Determine if the condition is met. or If satisfied, output ,otherwise ;

[0077] Step C44: Repeat steps B6 to B7 until all clusters are determined. .

[0078] The significant advancement of this invention compared to existing technologies lies in:

[0079] This invention addresses the problem of efficient task allocation and rapid channel access in reconnaissance scenarios. It constructs a two-stage joint optimization framework for task allocation and channel access in emergency reconnaissance scenarios. First, task allocation is transformed into a user clustering problem and modeled as a cooperative game with transferable utilities. Then, the channel access problem is modeled as an optimal stopping model. Based on this, a user clustering and channel access optimization algorithm is proposed, and the proposed algorithm is verified to effectively shorten channel access time and significantly improve the actual total utility of the system while meeting task capability requirements.

[0080] To more clearly illustrate the functional characteristics and structural parameters of the present invention, further explanation is provided below in conjunction with the accompanying drawings and specific embodiments. Attached Figure Description

[0081] Figure 1 This is a schematic diagram of task allocation and channel access optimization in a reconnaissance scenario provided in this application;

[0082] Figure 2 This is a convergence curve diagram of the cluster formation process under different numbers of users and tasks provided in this application;

[0083] Figure 3 This is a diagram showing the stopping detection results under different numbers of channels provided in this application;

[0084] Figure 4 This is a comparison chart of the utility gained from executing tasks with different numbers of tasks provided in this application. Detailed Implementation

[0085] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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.

[0086] Figure 1 This is a schematic diagram illustrating task allocation and channel access optimization in a reconnaissance scenario according to the present invention. First, ground users determine their tasks based on their capabilities and mission requirements, and move from their current location to the mission area. Then, to further improve mission execution effectiveness, relevant data needs to be collected before execution. Specifically, the command center dispatches drones to provide data to the clusters, and each cluster elects a cluster head to download data from the drones and forward it to users within the cluster.

[0087] Figure 2 The convergence curves of the clustering process in Phase 1 are shown under different numbers of ground users and tasks. When the number of tasks remains constant, increasing the number of ground users leads to more exploration and adjustments to determine the final clustering strategy, thus increasing the number of iterations. Conversely, when the number of ground users remains constant, increasing the number of tasks has little impact on the number of iterations. This indicates that increasing the number of tasks mainly increases the number of clusters without significantly affecting the exploration process of the clustering strategy.

[0088] Figure 3 The characteristics of the average number of channel stops in the second phase are illustrated. Considering that each task independently explores the transmission channels in this phase, the average number of channel stops for each task is calculated for greater generality. The figure shows that the number of channel stops decreases for each group as the probing time increases. This is because, under higher probing overhead, channels tend to stop probing earlier and begin data transmission. Furthermore, when the probing time is short, the average number of channel stops increases with the number of channels; however, when the probing time is long, the average number of stops tends to stabilize after the number of channels reaches a certain level. This is because, under high probing overhead, the expected gain from continuing channel exploration is less than the data reception loss caused by occupying transmission time. Therefore, each cluster tends to start transmission after exploring a fixed number of channels to maintain more data transmission.

[0089] Figure 4This study demonstrates the impact of different task numbers on the total actual utility of task completion. In the first stage, as the number of tasks increases, the clustering utility based on the coalition-forming game theory algorithm gradually improves. This is because, on the one hand, increasing the number of tasks directly increases the expected total utility; on the other hand, although increasing the number of tasks may lead to a decrease in the number of members in each cluster, the coalition-forming algorithm can better combine user capabilities and task requirements, thus resulting in an overall upward trend in utility. In contrast, the k-means algorithm clusters based on distance, which fails to fully consider the capabilities and task requirements of ground users. Therefore, a decrease in the number of cluster members leads to a decrease in the utility of completing tasks, or even the inability to complete some tasks. Thus, when the number of tasks exceeds 4, the utility tends to stabilize or even decrease slightly. In the second stage, the more tasks there are, the fewer channels each task is allocated. For the optimal stopping and random probing methods, transmission begins after probing a fixed number of channels, so changes in the number of channels have a relatively small impact on these two methods. However, for the full probing method, a decrease in the number of channels means that more time can be devoted to data transmission, so this method experiences the largest increase in utility when the number of tasks increases.

[0090] Based on the sequential relationship between task allocation and channel access, this invention proposes a two-stage optimization framework. In the first stage, the task allocation problem is transformed into a ground user clustering problem. A coalition formation game is introduced to model the user clustering process, and user clustering optimization is achieved based on coalition criteria. In the second stage, considering the unknown channel state in air-to-ground communication, optimal stopping theory is introduced to model the channel access process. By optimizing the timing of channel detection, a suitable channel is selected for access from the set of detected channels. Based on this two-stage optimization framework, this invention proposes an optimization algorithm for user clustering and channel access, providing support for subsequent reconnaissance missions.

[0091] 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 method for collaborative task allocation and air-to-ground channel access optimization in reconnaissance scenarios, characterized in that, The methods include: Step 1: Determine the task allocation model, and determine the task requirements and user capabilities within the model; Step 2: Determine the data transmission model; Step 3: Determine and optimize the method for calculating the actual utility of the task; Step 4: Use a two-stage optimization framework to solve the optimization problem. For the first stage, determine the user's clustering strategy. Step 5: In the second stage, determine the channel exploration strategy and transform the channel exploration strategy into a data reception problem.

2. The method according to claim 1, characterized in that, Step 1: Determine the task allocation model, and define the task requirements and user capabilities within the model; including: Ground users and reconnaissance missions are randomly distributed throughout the network, represented as follows: and Users need to participate in the reconnaissance mission, and users participating in the same mission should cooperate in clusters to enhance the execution effect. The collaborative relationships between users are influenced by both task requirements and the users' own capabilities; among which, task requirements are reflected in the comprehensive demands on multiple capabilities; let the set of all capability requirement types be . For any task Capability requirements are represented as vectors. ; In order to portray different abilities in the mission The importance of introducing capability weights ;when When, it indicates a task. No ability required Meanwhile, for any task The ability weights satisfy ;Task The set of capability requirements is represented as ; the task The total value is expressed as Among them, the k-th type of ability in the task The corresponding value is defined as: ； Regarding user capabilities, it is assumed that users dispatched by the command center possess the ability to complete all tasks; for any user... Mastery ability is represented as a vector. ,in Indicates user Regarding ability The degree of mastery; The degree of mastery is characterized using probability; the stronger a user's mastery of a certain ability, the higher the probability of independently completing the corresponding ability in a task; for a given task... When multiple users are capable of fulfilling a certain capability requirement, the overall probability of fulfilling that capability will be further increased; In the mission The probability of completion is expressed as: ； in, Indicates support for the task The user set; Considering that task utility depends on the value of each ability and the probability of satisfying the corresponding ability, completing the task... The returns are expressed as: ； in, The indicator function is used to ensure that all required capabilities for the task are met. ； Additionally, considering that users capable of fulfilling a certain task are not located in the corresponding task area, the movement cost for users to move to the task area needs to be considered; assuming users... Need to move to task The central point, representing the user's mobility cost as ,in This represents the cost per meter a user moves. Indicates user With the task The distance between them.

3. The method according to claim 2, characterized in that, Step 2: Determine the data transmission model, including: To ensure that each cluster can acquire the data needed to perform its mission within a limited time, the command center allocates one drone and a certain number of available channels to each cluster. The drone set is represented as follows: The set of available channel resources for a cluster is represented as Each cluster selects a suitable channel from its assigned set of available channels for data reception; to prevent interference between clusters during data reception, there is no overlap between the sets of available channels assigned to different clusters. Data requirements are related to the capability requirements of the task: for any task , if ability Data requirements are expressed as Then execute the task. The total data requirement of the cluster is expressed as To avoid duplicate downloads, each cluster's cluster head downloads data from the drone and forwards it to users within the cluster. The set of cluster heads is represented as... Because the members within a cluster are close to each other, the time it takes for the cluster head to receive data from the drone is much longer than the time it takes for users within the cluster to forward data to each other. For each cluster, block fading is used to model the channel, that is, in The channel state remains unchanged within a certain time; the time for each cluster head to probe one channel is... The cluster head first probes the channel to determine a suitable receiving channel. Once the channel is determined, the remaining time is used for data reception. It is assumed that the channels in the available channel set are independent and identically distributed. Considering the complexity of the environment in the reconnaissance scenario, Rayleigh fading is used to characterize the channel properties. Arbitrary Channel Channel gain The probability density function is expressed as: ； in, Reflects the average channel gain, and has , The path loss coefficient is estimated through a rough pre-deployment survey conducted before task allocation. If a channel is used When receiving data, the receiving rate is expressed as: ； in, For channel bandwidth, For the drone's transmission power, It is additive white Gaussian noise.

4. The method according to claim 3, characterized in that, Step 3: Determine and optimize the method for calculating the actual utility of the task; including: Therefore, the task The actual utility is calculated as follows: ； in, Indicates task The number of channel explorations corresponding to the cluster. Indicates that it has been explored The optimal channel among all channels This indicates the degree of impact of data reception on utility; considering that the tasks supported by each user are determined by their own clustering strategy, and the channels accessed by each cluster head are determined by the channel exploration strategy, the two strategies are represented as follows: and ; By optimizing the clustering strategy for each user and the channel exploration strategy for each cluster head, the actual total utility is maximized; the optimization problem is expressed as: ； Among these constraints, the channel exploration time of the cluster head is ensured to be less than the block fading duration, thus reserving a certain amount of time for data reception.

5. The method according to claim 4, characterized in that, Step 4: Solve the optimization problem using a two-stage optimization framework. For the first stage, determine the user's clustering strategy, including: The clustering strategy is modeled as a transferable utility alliance-forming game, and the game tuple is represented as follows: ;in, For users The game-theoretic utility; For a clustered set, it is represented as ; Given the network topology; assume any cluster Able to meet the task Based on the basic capability requirements, the benefits that users within a cluster can obtain after completing the corresponding task are: ； In addition, cluster The movement cost incurred by an internal user to complete a task is represented as follows: ； cluster The expected utility obtained by internal users is: ； In the mission During execution, cluster Each user plays an important role, and each user is considered to have equal value. Therefore, by distributing the expected utility equally among all users, the expected utility function for each user within the cluster is obtained as follows: ； During the process of forming clusters for ground users, a cooperation criterion is adopted to optimize the clustering; For any user as well as If the user I prefer to join the cluster It should satisfy the following relationship: ； If user In cluster In the middle, the utility function Represented as: ； in For users The clustering strategy, namely , To exclude users Clustering strategies for other users; by adjusting the clustering strategy for each user, the expected total utility of all users is continuously improved, eventually resulting in a stable consortium partition; Algorithmically, considering that the basic capabilities of each task need to be met, a preprocessing algorithm is determined to obtain the initial cluster structure; then, the cluster structure is adjusted through a coalition formation algorithm. The algorithm steps for cluster preprocessing are as follows: Step A1: Initialize all users into a free user set; Step A2: Based on task value Determine the processing order from high to low; Step A3: Extract tasks in sequence to determine the current unmet capability requirements; Step A4: Select a user from the free users who best matches the task's ability requirements and add them to the corresponding task. If multiple users have the same matching degree, select the user closest to the task. Step A5: Remove the user performing the task from the set of free users; Step A6: Repeat steps A2 to A5 until the basic capability requirements of all tasks are met, and finally output the current clustering structure. ; The alliance formation algorithm is as follows: Step B1: Input the current clustering structure ; Step B2: Randomly generate the processing order, processing users who have not joined the cluster first, and then processing users who have joined the cluster; Step B3: Extract users according to the processing order Try adding other clusters and calculate the clustering utility. If adding other clusters improves the user's game utility, then change the user's clustering strategy and update the clustering structure. ; Step B4: Repeat steps B2 to B3 until a stable alliance partition is obtained.

6. The method according to claim 5, characterized in that, Step 5: In the second stage, the channel exploration strategy is determined, which is then transformed into a data reception problem, including: A channel exploration strategy is selected for one cluster and then extended to other clusters. In this phase, each cluster needs to acquire as much data as possible, and the goal of channel exploration is to select a suitable channel from the explored channels to receive data. Therefore, the channel exploration strategy is transformed into a data reception problem, and the data reception rate of the cluster head is expressed as: ； in, For explored The maximum channel gain in the channel is calculated as follows: The optimization goal at this stage is to maximize the data reception rate of each cluster. A 1-SLA-based channel exploration algorithm is determined; during channel exploration, the channel exploration stopping time is determined by comparing the data reception rate with a pre-calculated threshold value; the derivation of the threshold value is as follows: ； in, , , ; It is an exponential integral function, and its specific expression is: ; The channel exploration algorithm based on 1-SLA is as follows: Step C1: Elect a cluster head for each cluster And set the number of detections. ; Step C2: Each cluster head Determine the maximum channel gain in the set of detected channels. and calculate ; Step C3: Determine if the condition is met. or If satisfied, output ,otherwise ; Step C44: Repeat steps B6 to B7 until all clusters are determined. .

Images