Method, device, equipment and storage medium for reliable transmission of situation graph based on coding cache

Abstract

The present invention provides a method, device, equipment and storage medium for reliable transmission of a situation map based on coding cache. The method comprises: collecting disaster situation map data based on the scheduling instructions of a perception drone at each regional unit in the disaster-stricken area in the current time slot; encoding the collected disaster situation map data into coding segments based on the coding cache parameters of the perception drone in the current time slot; determining a target relay drone matched by the perception drone in the current time slot and the transmission bandwidth of the target relay drone; so that the target relay drone sends the coding segment corresponding to the perception drone to a ground station via the transmission bandwidth, and all the coding segments are merged by the ground station to generate a disaster situation map of the disaster-stricken area.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method, device, equipment and storage medium for reliable transmission of a situation map based on coding cache. Background Art

[0002] The frequent occurrence of natural disasters and production accidents often results in significant casualties. Emergency rescue environments are characterized by high population density, diverse disaster types, and complex geographical environments. A key step in the emergency rescue process is identifying and continuously updating disaster situation information to provide command centers with real-time updates on the latest on-site conditions. Drones, with their flexible capabilities, can collect disaster situation map data in real time and upload it to ground stations via wireless networks for aggregation, reconstructing, and dynamically updating the disaster situation map.

[0003] However, due to the dynamic nature of disaster situations at emergency sites, high service demands, and unstable communication links, transmitting disaster situation maps over wireless networks is challenging. On the one hand, the mismatch between heavy communication load and limited wireless resources hinders efficient content transmission. While the disaster situation map contains a wealth of dynamic contextual information that can effectively guide emergency rescue operations, it also consumes significant communication bandwidth. On the other hand, drones must constantly move to update the latest disaster situation map data, but the limited contact time between the drone and the ground station prevents the transmission of all content. Summary of the Invention

[0004] The present invention provides a method, device, equipment and storage medium for reliable transmission of a situation map based on a coding cache, so as to solve the defects of the reliable transmission method of a situation map based on a coding cache in the prior art.

[0005] The present invention provides a method for reliable transmission of a situation map based on a coding cache, comprising the following steps:

[0006] Collect disaster situation map data based on the dispatch instructions of the perception drones at each regional unit in the disaster-stricken area in the current time slot, wherein the disaster-stricken area is divided into multiple regional units, and at least one perception drone is configured at each regional unit;

[0007] Encoding the collected disaster situation map data into coded segments based on the coding cache parameters of the perception drone in the current time slot;

[0008] Determine the target relay drone matched by the perception drone in the current time slot and the transmission bandwidth of the target relay drone; enable the target relay drone to send the coded segment corresponding to the perception drone to the ground station via the transmission bandwidth, and combine all the coded segments to generate a disaster situation map of the disaster-stricken area through the ground station;

[0009] The scheduling instruction, coding cache parameters, and transmission bandwidth of the target relay UAV in the current time slot are obtained by:

[0010] With the goal of maximizing the probability of successful recovery of the disaster situation map data at the regional unit in the current time slot, constructing an objective function under constraints;

[0011] The perception drone is regarded as an intelligent agent, and a deep reinforcement learning model corresponding to the objective function is constructed, wherein the state of the intelligent agent in the deep reinforcement learning model is determined based on the probability of success of the disaster situation map data corresponding to the perception drone being recovered by the ground station and the size of the coding segment; the action of the intelligent agent is determined based on the scheduling instruction decision of the perception drone, the transmission bandwidth allocation decision of the target relay drone, and the coding cache parameter adjustment of the perception drone; and the reward of the intelligent agent is determined based on the probability of successful recovery of the coding segment corresponding to the perception drone in the current time slot;

[0012] Based on the deep reinforcement learning model, the scheduling instructions, coding cache parameters and transmission bandwidth of the target relay drone in the current time slot are determined.

[0013] According to a reliable transmission method of a situation map based on code cache provided by the present invention, the probability of successful recovery of the code segment corresponding to the sensing drone in the current time slot is obtained by:

[0014] Determine a first success probability of transmitting the coded segment corresponding to the sensing UAV from the target relay UAV to the ground station in the current time slot;

[0015] Determining a second success probability that the disaster situation map data corresponding to the perception drone is restored by the ground station in the current time slot;

[0016] Based on the first success probability, the second success probability and the scheduling indication of the perception drone in the current time slot, the success recovery probability of the coding segment corresponding to the perception drone in the current time slot is determined.

[0017] According to a reliable transmission method of a situation map based on coding cache provided by the present invention, the target relay UAV includes a first target relay UAV and a second target relay UAV, and determining a first success probability of transmitting the coding segment corresponding to the perception UAV from the target relay UAV to the ground station in the current time slot includes:

[0018] determining a first signal-to-noise ratio at the ground station when a communication mode between the first target relay UAV, the second target relay UAV, and the ground station in a current time slot is a symbiotic communication mode, wherein in the symbiotic communication mode, the first target relay UAV establishes communication with the ground station through the second target relay UAV;

[0019] Determining a first transmission rate corresponding to the encoded segment corresponding to the sensing drone based on a first signal-to-noise ratio at the ground station;

[0020] Based on the size of the coding segment corresponding to the perception drone, the first transmission rate, the transmission bandwidth of the second target relay drone, and the communication time between the second target relay drone and the ground station, determine the first success probability of transmitting the coding segment corresponding to the perception drone from the second target relay drone to the ground station in the current time slot.

[0021] According to a method for reliable transmission of a situation map based on coding buffer provided by the present invention, determining a first signal-to-noise ratio at the ground station includes:

[0022] Based on the backscatter coefficient between the first target relay drone and the second target relay drone, the drone transmit power, the channel gain between the first target relay drone and the second target relay drone, and the channel gain between the second target relay drone and the ground station, a first signal-to-noise ratio at the ground station is determined.

[0023] According to a reliable transmission method of a situation map based on coding cache provided by the present invention, the target relay UAV includes a third target relay UAV, and determining a first success probability of transmitting the coding segment corresponding to the perception UAV from the target relay UAV to the ground station in the current time slot includes:

[0024] When the communication mode between the third target relay UAV and the ground station in the current time slot is an active transmission communication mode, determining a second signal-to-noise ratio at the ground station;

[0025] Determining a second transmission rate corresponding to the encoded segment corresponding to the sensing drone based on a second signal-to-noise ratio at the ground station;

[0026] Based on the size of the coding segment corresponding to the perception drone, the second transmission rate, the transmission bandwidth of the third target relay drone and the communication time between the third target relay drone and the ground station, determine the first success probability of transmitting the coding segment corresponding to the perception drone from the third target relay drone to the ground station in the current time slot.

[0027] According to a method for reliable transmission of a situation map based on coding cache provided by the present invention, the constraints include:

[0028] The number of the coded segments at the regional unit received by the ground station in the current time slot meets the preset number requirement;

[0029] The scheduling indication of the sensing drone in the current time slot is scheduling or not scheduling;

[0030] The total transmission bandwidth of the target relay drone matched by the sensing drone in the current time slot is not greater than the total bandwidth threshold;

[0031] In the current time slot, the signal-to-noise ratio between the target relay drone matched by the perception drone and the ground station is not less than a signal-to-noise ratio threshold.

[0032] The present invention also provides a reliable transmission device for a situation map based on a coding cache, the device comprising:

[0033] A first coding cache-based situation map reliable transmission module is configured to collect disaster situation map data based on the scheduling instructions of the perception drones at each regional unit in the disaster-stricken area in the current time slot, wherein the disaster-stricken area is divided into a plurality of regional units, and at least one perception drone is configured at each regional unit;

[0034] A second coding cache-based situation map reliable transmission module is used to encode the collected disaster situation map data into coding segments based on the coding cache parameters of the perception drone in the current time slot;

[0035] The third coding cache-based situation map reliable transmission module is used to determine the target relay drone matched by the perception drone in the current time slot and the transmission bandwidth of the target relay drone; so that the target relay drone sends the coding segment corresponding to the perception drone to the ground station through the transmission bandwidth, and the ground station merges all the coding segments to generate a disaster situation map of the disaster-stricken area;

[0036] The scheduling instruction, coding cache parameters, and transmission bandwidth of the target relay UAV in the current time slot are obtained by:

[0037] With the goal of maximizing the probability of successful recovery of the disaster situation map data at the regional unit in the current time slot, constructing an objective function under constraints;

[0038] The perception drone is regarded as an intelligent agent, and a deep reinforcement learning model corresponding to the objective function is constructed, wherein the state of the intelligent agent in the deep reinforcement learning model is determined based on the probability of success of the disaster situation map data corresponding to the perception drone being recovered by the ground station and the size of the coding segment; the action of the intelligent agent is determined based on the scheduling instruction decision of the perception drone, the transmission bandwidth allocation decision of the target relay drone, and the coding cache parameter adjustment of the perception drone; and the reward of the intelligent agent is determined based on the probability of successful recovery of the coding segment corresponding to the perception drone in the current time slot;

[0039] Based on the deep reinforcement learning model, the scheduling instructions, coding cache parameters and transmission bandwidth of the target relay drone in the current time slot are determined.

[0040] The present invention also provides an electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for reliably transmitting a situation diagram based on coding cache as described above is implemented.

[0041] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for reliably transmitting a situation diagram based on coding cache as described above is implemented.

[0042] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned methods for reliable transmission of situation maps based on coding cache.

[0043] The present invention provides a reliable transmission method, apparatus, device, and storage medium for disaster situation maps based on code caching. By utilizing drones to cache coded segments of disaster situation map data, this method enables the coordinated upload of disaster situation map data, improving transmission reliability. Specifically, a corresponding objective function is established by coupling the probability of successful recovery of the disaster situation map in the affected area with reliable transmission. A multi-agent dynamic decision-making algorithm based on deep reinforcement learning is employed to adapt to map update needs in real time. By coordinating the scheduling instructions of perception drones, the allocation of bandwidth resources to relay drones, and the adjustment of code caching parameters, dynamic resource optimization is achieved, thereby increasing the effective recovery area of ​​the disaster situation map. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 This is one of the application scenario diagrams of the reliable transmission system of the situation map based on the coding cache provided by the present invention;

[0046] Figure 2 This is the second schematic diagram of an application scenario of the reliable transmission system of the situation map based on coding cache provided by the present invention;

[0047] Figure 3 It is a flow chart of a reliable transmission method of a situation diagram based on coding cache provided by the present invention;

[0048] Figure 4 This is a schematic diagram of the scenario of collaborative multi-agent reinforcement learning provided by the present invention;

[0049] Figure 5 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0050] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0051] In order to better understand the embodiments of the present invention, the relevant contents involved in the embodiments of the present invention are first explained accordingly:

[0052] like Figure 1 The figure shows one of the scenario diagrams of the reliable transmission system of the situation diagram based on the coding cache provided by the present invention. Figure 1 In the present invention, sensing UAVs (SU) randomly dispersed in the disaster-stricken area use equipped cameras to perceive the disaster situation map data of the surrounding scene, and transmit the disaster situation map data to the ground station for disaster situation map construction through relay UAVs (RU), such as ground vehicles (GV). It should be noted here that in the following embodiments, the ground vehicle GV is used as the ground station in this embodiment.

[0053] The relay UAV RU is equipped with backscatter devices (BDs), which can act as relay nodes to cache coded segments and achieve reliable transmission through symbiotic communication mode and active transmission mode. The index of the sensing UAV SU is defined as , the set of perception drones is ,in Indicates the Here, the disaster-stricken area is mapped to a rasterized geospatial index structure and divided into Each sensing UAV SU has a certain situation awareness range, record the The sensing range of a sensing drone is ,in It is a region set with region unit index as element, defined as Indicates the A regional unit that senses the geographical coverage of the drone.

[0054] In this embodiment, the maximum distance separable code is selected for encoding. The disaster situation map of the disaster-stricken area can be expressed as , ,in Indicates the The disaster situation map data collected by the sensing drone is recorded in size During the construction of the disaster situation map, the coded segments are stored on the nearby relay drones RU. The size of each coded segment is ,in Indicates the restoration of The number of coding segments required for the disaster situation map data of a perception drone is determined. Different coding cache parameters can be adapted to different scenarios.

[0055] In this embodiment, the workflow of the reliable transmission system of the situation map based on the coding cache is as follows:

[0056] 1. Map collection (Phase ①): The perception UAV SU collects disaster situation map data according to the dispatch instruction variables.

[0057] 2. Encoding Caching (Phase 2): The sensing UAV (SU) encodes the acquired disaster situation map data based on optimized encoding caching parameters and shares the encoded segments with surrounding relay UAVs (RU). The optimal relay UAV selection and bandwidth allocation are then performed based on the transmission content and channel quality.

[0058] 3. Coded segment sharing (Phase ③): Multiple relay UAVs RU with cached encoded segments work together to transmit the encoded segments to the ground vehicle GV via the wireless network.

[0059] 4. Map construction (Phase ④): The ground vehicle GV combines these coded fragments to restore a complete disaster situation map of the affected area.

[0060] It should be noted that due to the differences in communication quality between links, the reasonable selection of relay drones and the distribution of coded segments of disaster situation map data are the key to the successful recovery of the disaster situation map. The set of relay drones for the coded segments of disaster situation map data collected by the sensing drone is ;

[0061] in, Indicates the Due to the instability of the communication link, not all relay drones in the area can successfully send the coded fragments of the disaster situation map data. In this embodiment, it is defined as the signal-to-noise ratio between the relay drone RU and the ground vehicle GV is greater than the minimum signal-to-noise ratio requirement. , it can be used as a candidate relay drone, then the set of candidate relay drones is expressed as:

[0062] (1)

[0063] in Relay drone To ground vehicles GV Transmission The signal-to-noise ratio of the coded segment of the disaster situation map data collected by the sensing drone, is the minimum signal-to-noise ratio threshold.

[0064] Assumptions There are The relay drone can successfully transmit the The coded fragments of the disaster situation map data collected by the sensing drones are possible combinations, so the number of possible choices for all candidate relay drones is defined as:

[0065] (2)

[0066] in , .at the same time, The complement of can be expressed as:

[0067] (3)

[0068] Therefore, for any possible , a total of The probability that a link can be successfully transmitted can be expressed as .

[0069] when When, The disaster situation map data collected by the first sensing drone can be restored. Regional units collected by sensing drones After the disaster situation map data is transmitted to the ground vehicle GV, the success probability of being recovered by the ground vehicle GV is expressed as:

[0070] (4)

[0071] In this embodiment, to maximize , select the one with the largest hit rate candidate relay drones to collaborate in transmitting the Coded fragment of disaster situation map data collected by a sensing drone.

[0072] In this embodiment, a binary variable is also defined Indicates the Whether the candidate relay UAV transmits the first The coded fragment of the disaster situation map data collected by the sensing drone, among which, .

[0073] Therefore, assisting The set of target candidate relay drones for the encoded segments of the disaster situation map data collected by the sensing drone is:

[0074] (5)

[0075] in, .

[0076] In addition, in this embodiment, orthogonal frequency division multiple access technology is used as the access scheme for different communication links. Since there is a line-of-sight component between the UAV and the ground station, in this embodiment, it is assumed that the air-to-ground (A2G) channel obeys the Rician channel fading model. Disaster situation map data collected by a sensing drone The relay drone collection is ;

[0077] in, Indicates the A relay drone. There are two communication modes for transmitting the encoded fragments of disaster situation map data collected by the perception drone from the cache relay drone to the ground vehicle GV: symbiotic communication mode and active transmission mode.

[0078] refer to Figure 2 As shown, in the symbiotic communication mode, the relay drone via relay drones The first The coded segments of disaster situation map data collected by the sensing drone are transmitted to the ground vehicle GV.

[0079] Specifically, in the relay drone The backscattered signal at can be expressed as:

[0080] (6)

[0081] in, represents the backscattering coefficient, and ; Indicates the UAV transmission power; Relay drone With relay drone The channel gain between Relay drone The transmission signal; It's a relay drone. and relay drones The channel gain between is the large-scale average channel gain, is the small-scale fading coefficient, It's a relay drone. and relay drones Therefore, the receiving signal of the ground vehicle GV can be expressed as:

[0082] (7)

[0083] in, It's a relay drone. and the channel gain between the ground vehicle GV, where is the large-scale average channel gain, It's a relay drone. and the distance between the ground vehicle GV. In addition, represents additive white Gaussian noise, Represents the noise power.

[0084] Here, by substituting formula (6) into formula (7), we can obtain:

[0085] (8)

[0086] Therefore, in this embodiment, in the symbiotic communication mode, the signal-to-noise ratio at the ground vehicle GV can be expressed as:

[0087] (9)

[0088] Therefore, the transmission rate can be expressed as:

[0089] (10)

[0090] Rule No. The encoded fragments of the disaster situation map data collected by the sensing drone are sent by the relay drone. The success probability of transmission to the ground vehicle GV can be expressed as follows:

[0091] (11)

[0092] in, It's a relay drone. Contact time with the ground vehicle GV, Indicates the number of relay drones assigned to the relay transmission bandwidth.

[0093] Continue to refer Figure 2 As shown, in the active transmission communication mode, the relay drone The first The coded segments of disaster situation map data collected by the sensing drone are directly transmitted to the ground vehicle GV.

[0094] Specifically, the signal received at the ground vehicle GV can be expressed as:

[0095] (12)

[0096] in, is the UAV transmission power, It's a relay drone. The transmission signal, It's a relay drone. and the channel gain between the ground vehicle GV. In addition, represents additive white Gaussian noise, Represents the noise power.

[0097] Relay drone The transmission rate between the ground vehicle GV can be expressed as:

[0098] (13)

[0099] in, Relay drone and the signal-to-noise ratio between the ground vehicle GV.

[0100] In active transmission communication mode, the relay drone The first The success probability of directly transmitting the coded fragments of the disaster situation map data collected by the sensing UAV to the ground vehicle GV can be expressed as:

[0101] (14)

[0102] in It's a relay drone. Contact time with the ground vehicle GV, Indicates the number of relay drones assigned to the relay transmission bandwidth.

[0103] Based on this, this embodiment proposes a new metric to evaluate the effective recovery area of ​​the disaster situation map, namely, the regional unit The disaster situation map data is In this embodiment, assuming that the observation period is divided into T time slots, the time slot set is used To express. Definition For regional units The disaster situation map data is The probability of successful recovery of a time slot is expressed as:

[0104] (15)

[0105] in, . Indicates the first A sensing unit for the area collected by the drone The probability of success of transmitting the encoded fragment of the disaster situation map data at the jth relay UAV to the ground vehicle GV is .

[0106] Here, the binary indicator variable , indicating whether to schedule the Perception drone Carry out disaster situation map data collection tasks. Indicates that it is dispatched to perform the disaster situation map data task. Indicates that the task of collecting disaster situation map data has not been scheduled.

[0107] In this embodiment, in order to improve the rescue efficiency, it is ensured that at least one sensing drone executes the regional unit. The disaster situation map data collection task can restore the regional unit Therefore, the regional unit The effective recovery area of ​​the disaster situation map data can be expressed as: ,in Represents regional unit area.

[0108] Furthermore, in this embodiment, based on the above, in order for the ground vehicle GV to obtain disaster situation map data from a specific area, it is necessary to The size of the code segment transmitted by each relay UAV RU is recorded as , various transmission schemes have a great impact on the success of transmission, and the transmission scheme itself is constrained by network capacity limitations. Due to the limited contact time between the relay UAV RU and the ground vehicle GV, transmitting a large amount of content from the same node may lead to link interruption and transmission failure. This is because the transmission time exceeds the contact time. When using the coding cache mode, reducing the content fragments of each cache increases the possibility of successful transmission on a single link. However, indiscriminately dividing the entire file into many small fragments may reduce the recovery speed of the entire file. Therefore, in order to meet the transmission requirements of disaster situations of different granularity, different coding caches are implemented in this embodiment according to the specific granularity requirements of the detailed situation information.

[0109] In one example, based on the characteristics of disaster situation map data, disaster situation map data collection can be divided into two layers: a dynamic fine-grained layer (high-definition situation information for key disaster areas) and a static coarse-grained layer (coarse-grained global situation information overview). To address this characteristic, this embodiment designs a layered coding cache framework for multi-granularity information transmission. This framework includes a highly reliable coding cache layer and a non-coding cache layer. Specifically, during a rescue operation, when global information is needed, transmitting a global overview to meet the operational needs of the command is sufficient. Global information requires a general overview of a specific area and is characterized by coarse collection granularity and small file size. In this case, a non-coding cache strategy can be effectively used. However, for specific rescue missions in a designated area, a more detailed and nuanced understanding of the on-site situation is essential for effective rescue operations. Detailed disaster situation map data requires fine-grained map collection, resulting in larger collection files. For large file sizes, a highly reliable coding cache strategy can be employed for transmission within given contact time constraints. This involves selecting different coding cache parameters for the transmission of coded segments to improve reliability.

[0110] In another example, the collection of disaster situation map data can be divided into three layers according to the characteristics of update frequency: high-frequency update layer: this layer contains dynamic information that needs to be updated frequently, including rescue vehicle locations, personnel information and other real-time data; medium-frequency update layer: this layer contains information such as obstacle locations, status, weather conditions and other data that need to be updated within a few minutes; low-frequency update layer: this layer consists of general features such as road closures, terrain and other static information, which only needs to be updated once every few hours or longer.

[0111] Figure 3 This is a flow chart of a reliable transmission method of a situation diagram based on coding cache provided by the present invention, such as Figure 3 As shown, the method includes the following:

[0112] Step 310: Collect disaster situation map data based on the dispatch instructions of the sensing drones at each regional unit in the disaster-stricken area in the current time slot, wherein the disaster-stricken area is divided into a plurality of regional units, and at least one sensing drone is deployed at each regional unit;

[0113] Step 320: Encode the collected disaster situation map data into code segments based on the code cache parameters of the perception drone in the current time slot;

[0114] Step 330: Determine the target relay drone and the transmission bandwidth of the target relay drone that the perception drone matches in the current time slot; enable the target relay drone to send the coded segment corresponding to the perception drone to the ground station via the transmission bandwidth, and combine all the coded segments to generate a disaster situation map of the disaster-stricken area through the ground station;

[0115] The scheduling instruction, coding cache parameters, and transmission bandwidth of the target relay UAV in the current time slot are obtained by:

[0116] With the goal of maximizing the probability of successful recovery of the disaster situation map data at the regional unit in the current time slot, constructing an objective function under preset constraints;

[0117] The perception drone is regarded as an intelligent agent, and a deep reinforcement learning model corresponding to the objective function is constructed, wherein the state of the intelligent agent in the deep reinforcement learning model is determined based on the probability of success of the disaster situation map data corresponding to the perception drone being recovered by the ground station and the size of the coding segment; the action of the intelligent agent is determined based on the scheduling instruction decision of the perception drone, the transmission bandwidth allocation decision of the target relay drone, and the coding cache parameter adjustment of the perception drone; and the reward of the intelligent agent is determined based on the probability of successful recovery of the coding segment corresponding to the perception drone in the current time slot;

[0118] Based on the deep reinforcement learning model, the scheduling instructions, coding cache parameters and transmission bandwidth of the target relay drone in the current time slot are determined.

[0119] Here, as mentioned above, the scheduling instructions of the perception drone in the current time slot are divided into two situations, namely being scheduled or not scheduled. When being scheduled, the perception drone will collect disaster situation map data for the area unit it is responsible for sensing, and then use the corresponding encoding cache parameters to encode the disaster situation map data into several coded segments and distribute them to the matching target relay drones. Here, the specific screening process of the target relay drone is through the above formulas (1)-(6), which will not be described in detail here; then the target relay drone sends the coded segments to the ground station with its configured transmission bandwidth, and the ground station merges all the coded segments to generate a disaster situation map of the disaster area.

[0120] In this embodiment, the probability of successful recovery of the disaster situation map data at the regional unit at the current time slot t is maximized. For any time slot t, define For the A sensing drone scheduling indicator variable; For the The transmission bandwidth allocation decision variables of the target relay UAV matched by the sensing UAV; For the The adjustment variables of the encoding cache parameters of the sensing drone. Therefore, the objective function can be expressed as:

[0121] (16)

[0122] Among them, the constraints of this objective function include:

[0123] , the number of coded segments at the regional unit c received by the ground station at the current time slot t meets the preset number requirement , that is, the ground station needs to obtain the disaster situation map data of regional unit c Slice encoding fragment, here It is calculated by the above formula (5) and will not be detailed here.

[0124] , the scheduling instruction of the perceived UAV at the current time slot t is scheduling or not scheduling;

[0125] , the total transmission bandwidth of the coded segment at the transmission area unit c in the current time slot t is not greater than the total bandwidth threshold;

[0126] , the signal-to-noise ratio between the target relay UAV matched by the sensing UAV and the ground station at the current time slot t is not less than the signal-to-noise ratio threshold.

[0127] In this embodiment, a deep reinforcement learning model corresponding to the objective function is constructed to describe the collection and transmission process of disaster situation map data. Specifically, each perception drone SU can be regarded as an independent intelligent agent, and their common goal is to maximize the effective restoration area of ​​the disaster situation map. The key elements of the deep reinforcement learning model are defined as follows:

[0128] State: The state is represented by the local observation value of the agent. The local observation value of each agent is expressed as:

[0129] (17)

[0130] in, Indicates that in the current time slot t , represents the size of the encoding segment at the current time slot t. Therefore, a set of local observations obtained by all agents can be expressed as:

[0131] (18)

[0132] Action: In the current time slot , the decision variables of each agent include: , , , then the action of each agent can be expressed as:

[0133] (19)

[0134] Then the joint action of all agents can be expressed as:

[0135] (20)

[0136] award:

[0137] (twenty one)

[0138] in, is a hyperparameter that controls the size of the reward.

[0139] Agent Network: The agent network is the network in which each agent is used to fit function and a network that facilitates distributed execution. The network uses deep The structure of the network, including the evaluation network and the target network. The input of the agent network includes the current observation value and previous actions The output of the network is value .

[0140] Furthermore, in this embodiment, using The network is used to collaborate and learn actions. A hybrid network is used to expand and aggregate all the outputs of the local deep Q network. Figure 4 As shown, the weight 、 and deviation 、 It is dynamically generated according to the current environment state. Therefore, the global action value It can be expressed as:

[0141] (twenty two)

[0142] Where, Represents a hybrid network, characterizing the mapping function from local action value to global action value, that is, the mapping function defined by the hybrid network structure. In addition, the function The two layers are connected and nonlinearity is introduced into the hybrid network so that it can fit various mapping functions. In this embodiment, ReLu is used as the nonlinear function.

[0143] For each agent , the agent network represents its individual value function In order to extract a distributed policy that is completely consistent with the centralized policy, the QMIX network needs to learn a joint action-value function To ensure consistency, you need to ensure that The global argmax performed on yields the same result as the function for each individual value A collection of single argmax operations performed:

[0144] (twenty three)

[0145] Where, It is a history of action observation. It is a joint action. The absolute activation function is used to ensure that the weight of the mixed Q value is non-negative, so that the global action value always follows the local action value of each agent. Monotonically increasing, that is:

[0146] (twenty four)

[0147] Among them, the global parameters Including hybrid network parameters , also including the local Q-network parameters of each agent.

[0148] During training, the global parameters are updated by minimizing the following loss :

[0149] (25)

[0150] in, is the batch size of transforms sampled from the replay buffer. is the temporal difference target:

[0151] (26)

[0152] in, Indicates the target network.

[0153] The present invention provides a reliable transmission method for disaster situation maps based on coding cache. By utilizing drones to cache coded segments of disaster situation map data, this method enables coordinated upload of disaster situation map data, improving transmission reliability. Specifically, a corresponding objective function is established by coupling the probability of successful recovery of the disaster situation map in the affected area with reliable transmission. A multi-agent dynamic decision-making algorithm based on deep reinforcement learning is employed to adapt to map update needs in real time. By coordinating the scheduling instructions of perception drones, the allocation of bandwidth resources to relay drones, and the adjustment of coding cache parameters, dynamic resource optimization is achieved, thereby increasing the effective recovery area of ​​the disaster situation map.

[0154] In some embodiments, the probability of successful recovery of the coded segment corresponding to the sensing drone in the current time slot is obtained by:

[0155] Determine a first success probability of transmitting the coded segment corresponding to the sensing UAV from the target relay UAV to the ground station in the current time slot;

[0156] Determining a second success probability that the disaster situation map data corresponding to the perception drone is restored by the ground station in the current time slot;

[0157] Based on the first success probability, the second success probability and the scheduling indication of the perception drone in the current time slot, the success recovery probability of the coding segment corresponding to the perception drone in the current time slot is determined.

[0158] Specifically, the first A sensing unit for the area collected by the drone The probability of successful recovery of the encoded fragment of the disaster situation map data at Refer to Formula (15) above. Here, the first success probability at the current time slot t can be calculated using Formula (11) or Formula (14), and the second success probability at the current time slot t can be calculated using Formula (4). Details will not be repeated here.

[0159] Specifically, the target relay drone includes a first target relay drone and a second target relay drone, and determining a first success probability of transmitting the coded segment corresponding to the sensing drone from the target relay drone to the ground station in the current time slot includes:

[0160] determining a first signal-to-noise ratio at the ground station when a communication mode between the first target relay UAV, the second target relay UAV, and the ground station in a current time slot is a symbiotic communication mode, wherein in the symbiotic communication mode, the first target relay UAV establishes communication with the ground station through the second target relay UAV;

[0161] Determining a first transmission rate corresponding to the encoded segment corresponding to the sensing drone based on a first signal-to-noise ratio at the ground station;

[0162] Based on the size of the coding segment corresponding to the perception drone, the first transmission rate, the transmission bandwidth of the second target relay drone, and the communication time between the second target relay drone and the ground station, determine the first success probability of transmitting the coding segment corresponding to the perception drone from the second target relay drone to the ground station in the current time slot.

[0163] refer to Figure 2 As shown, in the symbiotic communication mode, the relay drone (i.e. the first target relay drone) through the relay drone (i.e. the second target relay drone) will The encoded fragments of the disaster situation map data collected by the sensing drone are transmitted to the ground vehicle GV. In this communication mode, the first success probability at the current time slot t can be calculated by the formulas (6)-(11) described above, which will not be detailed here.

[0164] Specifically, determining a first signal-to-noise ratio at the ground station includes:

[0165] Based on the backscatter coefficient between the first target relay drone and the second target relay drone, the drone transmit power, the channel gain between the first target relay drone and the second target relay drone, and the channel gain between the second target relay drone and the ground station, a first signal-to-noise ratio at the ground station is determined.

[0166] Specifically, as described above, in the symbiotic communication mode, the first signal-to-noise ratio at the ground vehicle GV can be expressed as:

[0167] ;

[0168] Here, represents the backscatter coefficient between the first target relay UAV and the second target relay UAV, Indicates the UAV transmission power, represents the noise power, represents the channel gain between the first target relay UAV and the second target relay UAV, Represents the channel gain between the second target relay UAV and the ground station.

[0169] In some embodiments, the target relay UAV includes a third target relay UAV, and determining a first success probability of transmitting the coded segment corresponding to the sensing UAV from the target relay UAV to the ground station in the current time slot includes:

[0170] When the communication mode between the third target relay UAV and the ground station in the current time slot is an active transmission communication mode, determining a second signal-to-noise ratio at the ground station;

[0171] Determining a second transmission rate corresponding to the encoded segment corresponding to the sensing drone based on a second signal-to-noise ratio at the ground station;

[0172] Based on the size of the coding segment corresponding to the perception drone, the second transmission rate, the transmission bandwidth of the third target relay drone and the communication time between the third target relay drone and the ground station, determine the first success probability of transmitting the coding segment corresponding to the perception drone from the third target relay drone to the ground station in the current time slot.

[0173] As mentioned above, in active transmission communication mode, the relay drone (i.e. the third target relay drone) will The encoded fragments of the disaster situation map data collected by the sensing drone are directly transmitted to the ground vehicle GV. The calculation formula for the first success probability can refer to Formula (12)-Formula (14) above. The calculation formula for the second transmission rate can refer to Formula (13) above, which will not be detailed here.

[0174] The following describes the method and device for reliable transmission of a situation diagram based on coding cache provided by the present invention. The method and device for reliable transmission of a situation diagram based on coding cache described below and the method for reliable transmission of a situation diagram based on coding cache described above can refer to each other.

[0175] In this embodiment, the method and apparatus for reliable transmission of a situation map based on coding cache includes: a first reliable transmission module for a situation map based on coding cache, a second reliable transmission module for a situation map based on coding cache, and a third reliable transmission module for a situation map based on coding cache.

[0176] A first coding cache-based situation map reliable transmission module is configured to collect disaster situation map data based on the scheduling instructions of the perception drones at each regional unit in the disaster-stricken area in the current time slot, wherein the disaster-stricken area is divided into a plurality of regional units, and at least one perception drone is configured at each regional unit;

[0177] A second coding cache-based situation map reliable transmission module is used to encode the collected disaster situation map data into coding segments based on the coding cache parameters of the perception drone in the current time slot;

[0178] The third coding cache-based situation map reliable transmission module is used to determine the target relay drone matched by the perception drone in the current time slot and the transmission bandwidth of the target relay drone; so that the target relay drone sends the coding segment corresponding to the perception drone to the ground station through the transmission bandwidth, and the ground station merges all the coding segments to generate a disaster situation map of the disaster-stricken area;

[0179] The scheduling instruction, coding cache parameters, and transmission bandwidth of the target relay UAV in the current time slot are obtained by:

[0180] With the goal of maximizing the probability of successful recovery of the disaster situation map data at the regional unit in the current time slot, constructing an objective function under constraints;

[0181] The perception drone is regarded as an intelligent agent, and a deep reinforcement learning model corresponding to the objective function is constructed, wherein the state of the intelligent agent in the deep reinforcement learning model is determined based on the probability of success of the disaster situation map data corresponding to the perception drone being recovered by the ground station and the size of the coding segment; the action of the intelligent agent is determined based on the scheduling instruction decision of the perception drone, the transmission bandwidth allocation decision of the target relay drone, and the coding cache parameter adjustment of the perception drone; and the reward of the intelligent agent is determined based on the probability of successful recovery of the coding segment corresponding to the perception drone in the current time slot;

[0182] Based on the deep reinforcement learning model, the scheduling instructions, coding cache parameters and transmission bandwidth of the target relay drone in the current time slot are determined.

[0183] The present invention provides a reliable transmission device for disaster situation maps based on code cache. By utilizing drones to cache coded segments of disaster situation map data, this device enables coordinated upload of disaster situation map data, improving transmission reliability. Specifically, a corresponding objective function is established by coupling the probability of successful recovery of the disaster situation map in the affected area with reliable transmission. A multi-agent dynamic decision-making algorithm based on deep reinforcement learning is employed to adapt to map update needs in real time. By coordinating the scheduling instructions of perception drones, the allocation of bandwidth resources to relay drones, and the adjustment of code cache parameters, dynamic resource optimization is achieved, thereby increasing the effective recovery area of ​​the disaster situation map.

[0184] Figure 5 An example of a physical structure diagram of an electronic device is shown below. Figure 5 As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530, and a communication bus 540, wherein the processor 510, the communication interface 520, and the memory 530 communicate with each other via the communication bus 540. The processor 510 may call the logic instructions in the memory 530 to execute a method for reliable transmission of a situation map based on a coding cache, the method including:

[0185] Collect disaster situation map data based on the dispatch instructions of the perception drones at each regional unit in the disaster-stricken area in the current time slot, wherein the disaster-stricken area is divided into multiple regional units, and at least one perception drone is configured at each regional unit;

[0186] Encoding the collected disaster situation map data into coded segments based on the coding cache parameters of the sensing drone in the current time slot;

[0187] Determine the target relay drone matched by the perception drone in the current time slot and the transmission bandwidth of the target relay drone; enable the target relay drone to send the coded segment corresponding to the perception drone to the ground station via the transmission bandwidth, and combine all the coded segments to generate a disaster situation map of the disaster-stricken area through the ground station;

[0188] The scheduling instruction, coding cache parameters, and transmission bandwidth of the target relay UAV in the current time slot are obtained by:

[0189] With the goal of maximizing the probability of successful recovery of the disaster situation map data at the regional unit in the current time slot, constructing an objective function under constraints;

[0190] The perception drone is regarded as an intelligent agent, and a deep reinforcement learning model corresponding to the objective function is constructed, wherein the state of the intelligent agent in the deep reinforcement learning model is determined based on the probability of success of the disaster situation map data corresponding to the perception drone being recovered by the ground station and the size of the coding segment; the action of the intelligent agent is determined based on the scheduling instruction decision of the perception drone, the transmission bandwidth allocation decision of the target relay drone, and the coding cache parameter adjustment of the perception drone; and the reward of the intelligent agent is determined based on the probability of successful recovery of the coding segment corresponding to the perception drone in the current time slot;

[0191] Based on the deep reinforcement learning model, the scheduling instructions, coding cache parameters and transmission bandwidth of the target relay drone in the current time slot are determined.

[0192] Furthermore, the logic instructions in the aforementioned memory 530 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0193] On the other hand, the present invention further provides a computer program product, comprising a computer program, which may be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the method for reliable transmission of a situation map based on a coding cache provided by the above methods, the method comprising:

[0194] Collect disaster situation map data based on the dispatch instructions of the perception drones at each regional unit in the disaster-stricken area in the current time slot, wherein the disaster-stricken area is divided into multiple regional units, and at least one perception drone is configured at each regional unit;

[0195] Encoding the collected disaster situation map data into coded segments based on the coding cache parameters of the perception drone in the current time slot;

[0196] Determine the target relay drone matched by the perception drone in the current time slot and the transmission bandwidth of the target relay drone; enable the target relay drone to send the coded segment corresponding to the perception drone to the ground station via the transmission bandwidth, and combine all the coded segments to generate a disaster situation map of the disaster-stricken area through the ground station;

[0197] The scheduling instruction, coding cache parameters, and transmission bandwidth of the target relay UAV in the current time slot are obtained by:

[0198] With the goal of maximizing the probability of successful recovery of the disaster situation map data at the regional unit in the current time slot, constructing an objective function under constraints;

[0199] The perception drone is regarded as an intelligent agent, and a deep reinforcement learning model corresponding to the objective function is constructed, wherein the state of the intelligent agent in the deep reinforcement learning model is determined based on the probability of success of the disaster situation map data corresponding to the perception drone being recovered by the ground station and the size of the coding segment; the action of the intelligent agent is determined based on the scheduling instruction decision of the perception drone, the transmission bandwidth allocation decision of the target relay drone, and the coding cache parameter adjustment of the perception drone; and the reward of the intelligent agent is determined based on the probability of successful recovery of the coding segment corresponding to the perception drone in the current time slot;

[0200] Based on the deep reinforcement learning model, the scheduling instructions, coding cache parameters and transmission bandwidth of the target relay drone in the current time slot are determined.

[0201] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for reliably transmitting a situation map based on a coding cache provided by the above methods is implemented. The method includes:

[0202] Collect disaster situation map data based on the dispatch instructions of the perception drones at each regional unit in the disaster-stricken area in the current time slot, wherein the disaster-stricken area is divided into multiple regional units, and at least one perception drone is configured at each regional unit;

[0203] Encoding the collected disaster situation map data into coded segments based on the coding cache parameters of the perception drone in the current time slot;

[0204] Determine the target relay drone matched by the perception drone in the current time slot and the transmission bandwidth of the target relay drone; enable the target relay drone to send the coded segment corresponding to the perception drone to the ground station via the transmission bandwidth, and combine all the coded segments to generate a disaster situation map of the disaster-stricken area through the ground station;

[0205] The scheduling instruction, coding cache parameters, and transmission bandwidth of the target relay UAV in the current time slot are obtained by:

[0206] With the goal of maximizing the probability of successful recovery of the disaster situation map data at the regional unit in the current time slot, constructing an objective function under constraints;

[0207] The perception drone is regarded as an intelligent agent, and a deep reinforcement learning model corresponding to the objective function is constructed, wherein the state of the intelligent agent in the deep reinforcement learning model is determined based on the probability of success of the disaster situation map data corresponding to the perception drone being recovered by the ground station and the size of the coding segment; the action of the intelligent agent is determined based on the scheduling instruction decision of the perception drone, the transmission bandwidth allocation decision of the target relay drone, and the coding cache parameter adjustment of the perception drone; and the reward of the intelligent agent is determined based on the probability of successful recovery of the coding segment corresponding to the perception drone in the current time slot;

[0208] Based on the deep reinforcement learning model, the scheduling instructions, coding cache parameters and transmission bandwidth of the target relay drone in the current time slot are determined.

[0209] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0210] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0211] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A reliable transmission method of a situation map based on coding cache, characterized in that: The method comprises: Collect disaster situation map data based on the dispatch instructions of the perception drones at each regional unit in the disaster-stricken area in the current time slot, wherein the disaster-stricken area is divided into multiple regional units, and at least one perception drone is configured at each regional unit; Encoding the collected disaster situation map data into coded segments based on the coding cache parameters of the sensing drone in the current time slot; Determine the target relay drone matched by the perception drone in the current time slot and the transmission bandwidth of the target relay drone; enable the target relay drone to send the coded segment corresponding to the perception drone to the ground station via the transmission bandwidth, and combine all the coded segments to generate a disaster situation map of the disaster-stricken area through the ground station; The scheduling instruction, coding cache parameters, and transmission bandwidth of the target relay UAV in the current time slot are obtained by: With the goal of maximizing the probability of successful recovery of the disaster situation map data at the regional unit in the current time slot, constructing an objective function under constraints; The perception drone is regarded as an intelligent agent, and a deep reinforcement learning model corresponding to the objective function is constructed, wherein the state of the intelligent agent in the deep reinforcement learning model is determined based on the probability of success of the disaster situation map data corresponding to the perception drone being recovered by the ground station and the size of the coding segment; the action of the intelligent agent is determined based on the scheduling instruction decision of the perception drone, the transmission bandwidth allocation decision of the target relay drone, and the coding cache parameter adjustment of the perception drone; and the reward of the intelligent agent is determined based on the probability of successful recovery of the coding segment corresponding to the perception drone in the current time slot; Based on the deep reinforcement learning model, the scheduling instructions, coding cache parameters and transmission bandwidth of the target relay drone in the current time slot are determined.

2. The method for reliable transmission of a situation map based on coding cache according to claim 1, characterized in that: The probability of successful recovery of the coded segment corresponding to the sensing drone in the current time slot is obtained by: Determine a first success probability of transmitting the coded segment corresponding to the sensing UAV from the target relay UAV to the ground station in the current time slot; Determining a second success probability that the disaster situation map data corresponding to the perception drone is restored by the ground station in the current time slot; Based on the first success probability, the second success probability and the scheduling indication of the perception drone in the current time slot, the success recovery probability of the coding segment corresponding to the perception drone in the current time slot is determined.

3. The method for reliable transmission of a situation map based on coding cache according to claim 2, characterized in that: The target relay UAV includes a first target relay UAV and a second target relay UAV, and determining a first success probability of transmitting the coded segment corresponding to the sensing UAV from the target relay UAV to the ground station in the current time slot includes: determining a first signal-to-noise ratio at the ground station when a communication mode between the first target relay UAV, the second target relay UAV, and the ground station in a current time slot is a symbiotic communication mode, wherein in the symbiotic communication mode, the first target relay UAV establishes communication with the ground station through the second target relay UAV; Determining a first transmission rate corresponding to the encoded segment corresponding to the sensing drone based on a first signal-to-noise ratio at the ground station; Based on the size of the coding segment corresponding to the perception drone, the first transmission rate, the transmission bandwidth of the second target relay drone, and the communication time between the second target relay drone and the ground station, determine the first success probability of transmitting the coding segment corresponding to the perception drone from the second target relay drone to the ground station in the current time slot.

4. The method for reliable transmission of a situation map based on coding cache according to claim 3, characterized in that: Determining a first signal-to-noise ratio at the ground station includes: Based on the backscatter coefficient between the first target relay drone and the second target relay drone, the drone transmit power, the channel gain between the first target relay drone and the second target relay drone, and the channel gain between the second target relay drone and the ground station, a first signal-to-noise ratio at the ground station is determined.

5. The method for reliable transmission of a situation map based on coding cache according to claim 2, characterized in that: The target relay UAV includes a third target relay UAV, and determining a first success probability of transmitting the coded segment corresponding to the sensing UAV from the target relay UAV to the ground station in the current time slot includes: When the communication mode between the third target relay UAV and the ground station in the current time slot is an active transmission communication mode, determining a second signal-to-noise ratio at the ground station; Determining a second transmission rate corresponding to the encoded segment corresponding to the sensing drone based on a second signal-to-noise ratio at the ground station; Based on the size of the coding segment corresponding to the perception drone, the second transmission rate, the transmission bandwidth of the third target relay drone and the communication time between the third target relay drone and the ground station, determine the first success probability of transmitting the coding segment corresponding to the perception drone from the third target relay drone to the ground station in the current time slot.

6. The method for reliable transmission of a situation map based on coding cache according to claim 1, characterized in that: The constraints include: The number of the coded segments at the regional unit received by the ground station in the current time slot meets the preset number requirement; The scheduling indication of the sensing drone in the current time slot is scheduling or not scheduling; The total transmission bandwidth of the target relay drone matched by the sensing drone in the current time slot is not greater than the total bandwidth threshold; In the current time slot, the signal-to-noise ratio between the target relay drone matched by the perception drone and the ground station is not less than a signal-to-noise ratio threshold.

7. A reliable transmission device for situation graph based on coding cache, characterized in that: The device comprises: A first coding cache-based situation map reliable transmission module is configured to collect disaster situation map data based on the scheduling instructions of the perception drones at each regional unit in the disaster-stricken area in the current time slot, wherein the disaster-stricken area is divided into a plurality of regional units, and at least one perception drone is configured at each regional unit; A second coding cache-based situation map reliable transmission module is used to encode the collected disaster situation map data into coding segments based on the coding cache parameters of the perception drone in the current time slot; The third coding cache-based situation map reliable transmission module is used to determine the target relay drone matched by the perception drone in the current time slot and the transmission bandwidth of the target relay drone; so that the target relay drone sends the coding segment corresponding to the perception drone to the ground station through the transmission bandwidth, and the ground station merges all the coding segments to generate a disaster situation map of the disaster-stricken area; The scheduling instruction, coding cache parameters, and transmission bandwidth of the target relay UAV in the current time slot are obtained by: With the goal of maximizing the probability of successful recovery of the disaster situation map data at the regional unit in the current time slot, constructing an objective function under constraints; The perception drone is regarded as an intelligent agent, and a deep reinforcement learning model corresponding to the objective function is constructed, wherein the state of the intelligent agent in the deep reinforcement learning model is determined based on the probability of success of the disaster situation map data corresponding to the perception drone being recovered by the ground station and the size of the coding segment; the action of the intelligent agent is determined based on the scheduling instruction decision of the perception drone, the transmission bandwidth allocation decision of the target relay drone, and the coding cache parameter adjustment of the perception drone; and the reward of the intelligent agent is determined based on the probability of successful recovery of the coding segment corresponding to the perception drone in the current time slot; Based on the deep reinforcement learning model, the scheduling instructions, coding cache parameters and transmission bandwidth of the target relay drone in the current time slot are determined.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method for reliable transmission of a situation map based on coding cache as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for reliable transmission of a situation map based on coding cache as claimed in any one of claims 1 to 6 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method for reliable transmission of a situation map based on coding cache as claimed in any one of claims 1 to 6 is implemented.

Images