A method and device for dynamically distributing data link data based on data content

Abstract

The present invention relates to the field of information technology, and provides a method and device for dynamically distributing data in a data link based on data content. The purpose of the present invention is to solve the problem of how to dynamically and adaptively distribute data when the data distribution destination is unclear under the data link architecture. The main scheme is used for a data platform that caches its own status information and the status information of neighboring platforms. The data platform parses the acquired original data packet, matches the distribution rules, and checks whether the parsed data packet includes all the attributes required for distribution reasoning; if some attributes are not included, a widened attribute field is added to the end of the parsed data packet; the relative address of the distribution destination is inferred based on the attributes required for distribution reasoning in the data packet according to the distribution rules; the distribution destination is converted into an absolute address, and the data packet is restored to the absolute address of the original data packet distributed to the distribution destination.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and provides a data content-based data link data dynamic distribution method and device. Background Art

[0002] Facing the needs of multi-platform autonomous collaborative operations, collaborative platforms need to perform accurate and real-time autonomous collaboration in a fast, complex and ever-changing operating environment. However, in the actual operating network environment, due to the limitations of perception and communication capabilities, collaborative platforms can only obtain observation information of part of the environment, and cannot obtain global information on the situation to achieve comprehensive and effective collaboration. Therefore, efficient data distribution between collaborative platforms has become a key technology to improve the efficiency of collaborative operations. In existing research, the collaborative data distribution relationship between platforms is usually pre-bound before collaboration, and it is difficult for data distribution services to be dynamically adjusted according to the task and status of the operation during the collaborative operation, resulting in the following problems:

[0003] 1. When the target platform for data distribution is unclear, relying on excessive binding of data distribution relationships to achieve data coverage will lead to redundant data transmission, data overload and data confusion, resulting in network congestion, insufficient stability and high latency;

[0004] 2. When the task or status changes, the target platform to which a certain data needs to be distributed may also change. For scenarios with little change, you can consider pre-binding the data distribution relationship after the change, but for scenarios with high changes, it is impossible to enumerate the changes, in which case this data distribution method will not meet the requirements.

[0005] 3. Data distribution relies on bound data distribution relationships that do not involve data content and cannot filter invalid data. Invalid data may be distributed, leading to erroneous collaborative decisions.

[0006] Therefore, such a data distribution model that is highly dependent on static distribution relationships is bound to be unsuitable for the future open, complex, and highly variable clustered intelligent operation model.

[0007] For targeted data distribution, when a piece of data needs to be distributed;

[0008] (1) If the data distribution destination is certain, then the data distribution relationship can be established in advance and the data can be distributed based on the established data distribution relationship. However, this method is not applicable when the data distribution destination is unclear.

[0009] (2) There is currently no method to achieve dynamic data distribution based on rule reasoning based on data content.

[0010] Based on this, we can analyze that in future collaborative operations, each platform should have the following application characteristics:

[0011] In the process of executing tasks, each platform needs to have the ability to adaptively and dynamically distribute data in response to dynamic changes in tasks and status. Different data distribution relationships should be formed based on the content of the data packet, situation status information, and data distribution requirements to provide the most timely task-related information to each platform.

[0012] Since traditional Internet communication is only responsible for the transmission of data packets, the content of data packets is transparent between the protocol layers, while the data link system is responsible for information transmission and processing, and the content of data packets on the data link is opaque between the protocol layers, the data elements can be parsed through the data link data element dictionary to obtain the data content, so that under the data link architecture, distribution reasoning based on data content can be realized, data distribution relations can be dynamically formed, and data packets can be directed to the platform that needs to receive the data packet, solving the above problems. Therefore, the present invention designs a data link data dynamic distribution method based on data content under the data link architecture, and uses the specific content of the data packet and the data in the resource pool of this platform as the judgment basis for dynamic data distribution based on distribution rule reasoning.

[0013] Prior art

[0014] Since the collaborative platform will generate a large amount of data during operation, it is impossible to synchronize the data between each platform to all other platforms directly once it is modified due to the limitations of network bandwidth and the characteristics of the constantly changing network topology in actual scenarios. In addition, this practice of synchronizing data immediately after data updates will require each platform to consume a large amount of computing resources to synchronize the data of each platform, thereby increasing the computing pressure of the platform. Therefore, a static directional data distribution method is provided to distribute data across platforms based on the subscription / notification relationship between preset platform resources. The platform subscribes to this resource based on the preset subscription relationship with the neighboring platform resource. When this resource generates new data, the neighboring platform notifies the source platform of the data.

[0015] Disadvantages: If the target platform for data distribution is unclear, in order to ensure data coverage, a subscription / notification relationship will be established between all possible target platforms and this resource on this platform. As a result, when this resource generates new data, it may be sent to a platform that does not actually need the data, causing data redundancy, confusion, or even overload, leading to network congestion and high latency.

[0016] If the target platform for data distribution changes when the task or status changes, for simple scenarios with few changes, you can also consider the changes and establish a subscription / notification relationship in advance, resulting in a small amount of data distribution redundancy when no changes occur. However, for complex scenarios with highly changing operating conditions, this data distribution model that relies on static distribution relationships will be difficult to adapt.

[0017] Data distribution depends on the preset subscription / notification relationship. If a resource generates new data and the data is invalid, it will also be notified to the corresponding platform according to the preset subscription / notification relationship, which may cause collaborative errors.

[0018] Prior art 1: A data distribution method, device, equipment and storage medium

[0019] The method parses the target data based on the target parsing rules to form a target classification label value of the target data, that is, the data type of the target data, and then determines the target distribution address corresponding to the target classification label value based on the correspondence between the pre-stored target classification label value and the pre-stored forwarding rules, and finally distributes the target data based on the target distribution address.

[0020] shortcoming:

[0021] The forwarding rules only include the destination type (message queue, database, application), IP, port, etc. corresponding to the predetermined target classification label value. Relying only on static relationship distribution, it is difficult to make dynamic adjustments based on tasks and status, and cannot meet the needs of dynamic changes in data distribution destinations.

[0022] Prior art 2: Method and device for distributing aircraft downlink data link messages

[0023] The method determines the transmission channel of the received data link telegram, obtains the message addressing rule information corresponding to the transmission channel, then extracts the characteristic parameters of the data link telegram, compares the characteristic parameters with the message addressing rules included in the message addressing rule information, and when the operation result is a match, sends the data link telegram through the transmission channel corresponding to the message addressing rule.

[0024] shortcoming:

[0025] According to the known needs of the airline, various message addressing rules are preset to obtain data link telegrams, and only rely on static relationship distribution. When the needs are not clear, it is impossible to preset message addressing rules to obtain data link telegram information.

[0026] In summary, the prior art all relies on static relationship distribution. When the data distribution destination changes dynamically, the static relationship will not be able to be distributed normally, and is not applicable when the data distribution destination is not clear. Summary of the invention

[0027] The purpose of the present invention is to solve the problem of how to dynamically and adaptively distribute data when the data distribution destination is unclear under the data link architecture.

[0028] In order to achieve the above purpose, the present invention adopts the following technical means:

[0029] A data link data dynamic distribution method based on data content includes a data platform for receiving or distributing data link data packets under a data link architecture, the data platform caches its own state information and state information of neighboring platforms, the data platform includes a data distribution module, and dynamically and adaptively distributes the received data. The specific implementation includes the following steps:

[0030] Step 1: Initialize the distribution rule library. Each platform loads the distribution rules into its own data distribution module. The distribution rules include the attribute information required for reasoning.

[0031] Step 2: The data platform obtains the original data packet sent by the data generation end, and parses the data elements of the original data packet according to the data packet data structure and data element dictionary of the data link to obtain a parsed data packet;

[0032] Step 3: Assuming that the distribution rule of this data packet has been matched, check whether the parsed data packet includes all the attributes required for distribution reasoning. If all are included, proceed to step 6; if some attributes required for distribution reasoning are not included in this data packet, these attributes are called widened attributes, and proceed to the next step;

[0033] Step 4: widen the parsed data packet, that is, add a widened attribute field to the parsed data packet to obtain a widened data packet, and the widened part is called a widened field;

[0034] Step 5: Call the corresponding preprocessing function of each widening attribute to perform data preprocessing, obtain the widening attribute value, and assign it to the widening data field;

[0035] Step 5.1, for each widening attribute, obtain the data required to calculate the widening attribute from the local resource cache status information or the parsed data packet content;

[0036] Step 5.2: Perform mathematical operations on the acquired data required for calculating the widening attribute, obtain the widening attribute value and assign it to the widening field;

[0037] Step 6: Based on the attributes required for distribution reasoning in the widened data packet (if not widened, based on the parsed data packet), the relative address of the distribution destination is inferred according to the distribution rules.

[0038] Step 7: Convert the relative address of the distribution destination to the absolute address of the distribution destination.

[0039] Step 8: Restore the widened data packet (or the parsed data packet if it is not widened) to the original data packet and distribute it to the distribution destination absolute address obtained in step 7.

[0040] In the above technical solution, step 2 is specifically as follows:

[0041] The formatted messages transmitted in the data link use a certain bit sequence to represent the context content. The data link data element dictionary specifies the content, data item definition and usage scope of the data elements used in the data link, and uses the data field identifier (DFI) and data usage domain identifier (DUI) as keywords to identify the data element items, that is, the data element is uniquely identified by a pair of DFI and DUI.

[0042] Since the construction and identification of each message are completely dependent on the data structure and data element dictionary of the message, and the data packet data structure and data element dictionary are opaque, the parsing process is to first identify the original data packet message header to obtain the message word type, then identify the message data structure based on the message identification field in this message word type, and finally quickly index the data element dictionary based on the message data structure, the DFI and DUI of the data element, parse the data elements in the original data packet, and obtain the parsed data packet.

[0043] In the above technical solution, step 4 is specifically as follows:

[0044] According to the data type of the widened attribute, a field with the size of this data type is added to the end of the parsed data packet. Figure 2 , and name it so that the attribute value can be obtained in the form of "data packet ['attribute name']" in distribution rule reasoning.

[0045] In the above technical solution, step 5 is specifically as follows:

[0046] Corresponding preprocessing functions are preset for the attributes required in the distribution rules. When a widened field is added to the parsed data packet, the preprocessing function corresponding to the widened attribute is called to obtain the data required to calculate this attribute from the local resource cache status information or the content of the parsed data packet, perform mathematical operations on it, obtain the attribute value and assign it to the widened field.

[0047] In the above technical solution, the distribution rules in step 6 are as follows:

[0048] The distribution rule exists in the form of a script function, the parameter of which is the widened data packet (if not widened, it is the parsed data packet), and the inference result is the relative address of the distribution destination of this data packet;

[0049] Use "if", "and", and "then" logical keywords to describe distribution conditions;

[0050] Use the "send_to_local" action keyword to describe data forwarding to this platform, and the parameter Dst destination defaults to this platform;

[0051] Use the "send_to" action keyword to describe whether the data packet is sent to the local platform or the neighboring platform. If it is sent to the local platform, the parameters include Type destination type and Dst destination; if it is sent to the neighboring platform, the parameters include Dst destination.

[0052] In the above technical solution, step 7 is specifically as follows:

[0053] In order to increase the reusability of the distribution rules, the result of the distribution rule reasoning specified in the above step 6 is the relative address between the platforms in the task group. After the relative address of the data packet distribution destination is obtained through distribution reasoning, the absolute address of the data packet distribution destination is mapped according to the relationship between the platforms in the task group.

[0054] For example: if the distribution rule conditions are met, it will be distributed to the maser platform. Here, the relative address of the data packet distribution destination is master. In different collaborative groups, different platforms serve as master platforms. In task group A, "Platform A" serves as the master platform, and in task group B, "Platform B" serves as the master. Therefore, in task group A, the relative address of the master platform will be mapped to the absolute address of "Platform A", and in task group B, it will be mapped to the absolute address of "Platform B".

[0055] In the above technical solution, step 8 is specifically as follows:

[0056] For the widened data packet, the widened field is released to obtain the parsed data packet. According to the data link data structure and the data element dictionary, the parsed data packet is reversely parsed to obtain the original data packet, and distributed to the distribution destination absolute address obtained in step 7.

[0057] The present invention also provides a data link data dynamic distribution device based on data content, including a data platform for receiving or distributing data packets under a data link architecture, the data platform caches its own state information and state information of neighboring platforms, and a data distribution module of the data platform dynamically and adaptively distributes the received data, characterized in that the dynamic adaptive distribution is implemented by the following modules:

[0058] Distribution rule base initialization module: Distribution rule base initialization, each data platform loads distribution rules into its own data distribution module, the distribution rules include attribute information required for reasoning;

[0059] The original data packet parsing module and the data platform obtain the original data packet sent by the data generation end, parse the data elements of the original data packet according to the data packet data structure and data element dictionary of the data link, and obtain the parsed data packet;

[0060] The judgment module checks whether the parsed data packet contains all the attributes required for distribution reasoning if the distribution rule of the data packet has been matched. If all the attributes are included, the reasoning module is executed. If some attributes required for distribution reasoning are not included in the data packet, these attributes are called widening attributes, and the widening module is executed.

[0061] The widening module widens the parsed data packet, that is, adds a widening attribute field to the parsed data packet to obtain a widened data packet. The widened part is called a widening field. The preprocessing function corresponding to each widening attribute is called to preprocess the data to obtain the widening attribute value and assign it to the widening data field.

[0062] The reasoning module, if not widened, then based on the parsed data packet, otherwise based on the attributes required for distribution reasoning in the widened data packet, infers the relative address of the distribution destination according to the distribution rules;

[0063] An address conversion module converts the relative address of the distribution destination into the absolute address of the distribution destination;

[0064] The data packet restoration module, if not widened, restores the parsed data packet to the original data packet and distributes it to the distribution destination absolute address obtained in the address translation module; otherwise, the widened data packet is restored to the original data packet and distributed to the distribution destination absolute address obtained in the address translation module.

[0065] The present invention also provides a storage medium. When a processor executes a program in the storage medium, the method for dynamically distributing data link data based on data content is implemented.

[0066] Because the present invention adopts the above technical means, it has the following beneficial effects:

[0067] 1. Establish a data link data dynamic adaptive distribution module based on data content, parse the data link data content, check whether the data packet includes all the attributes required in the distribution rule reasoning, if some required attributes are not included, then add the required attribute fields not included to the parsed data packet, and pre-process the data in the local resource cache or in the data packet to obtain the added attribute field value, obtain the widened data packet, and obtain the relative address of the distribution destination based on the widened data packet according to the distribution rule reasoning, and then convert it into an absolute address, and finally restore the data packet and distribute it to the absolute address. When the data directional distribution destination changes dynamically or is unclear under the data link architecture, dynamic on-demand data distribution based on data content not only solves the problem of network congestion and high latency caused by excessive binding of data distribution relationships due to bandwidth scarcity under the data link architecture, but also solves the problem of data distribution in complex and highly changing scenarios due to the inability to enumerate the changing scenarios, resulting in the inability to pre-bind static distribution relationships.

[0068] 2. The distribution rules are formulated based on the task data requirements of the data receiving end. When the data receiving end has new task data requirements, it is only necessary to modify or re-customize the distribution rules in advance, and then store them in the distribution rule library and reinitialize them. There is no need to modify other software modules.

[0069] 3. Dynamic distribution of data link data based on data content. After obtaining the data content, the distribution rule inference can first determine whether the data content is valid, which can avoid the harm caused by the distribution of invalid data.

[0070] 4. Distribution rule reasoning returns the relative address of the data packet distribution destination in the task group, which increases the reusability of the distribution rules. For other task groups with the same data distribution requirements, there is no need to modify the distribution rules and they can be reused directly, reducing the workload. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is the overall timing diagram;

[0072] Figure 2 The schematic diagram of the widening of the parsed data packet;

[0073] Figure 3 It is a flow chart of dynamic data distribution of the embodiment;

[0074] Figure 4 This is a schematic diagram of parsing the original data packet of the embodiment;

[0075] Figure 5 A schematic diagram of a data packet widening after parsing in the embodiment;

[0076] Figure 6 Broaden the attribute preprocessing flow chart for the embodiment. DETAILED DESCRIPTION

[0077] The following is a detailed description of the embodiments of the present invention. Although the present invention will be described and illustrated in conjunction with some specific embodiments, it should be noted that the present invention is not limited to these embodiments. On the contrary, modifications or equivalent substitutions made to the present invention should all be included in the scope of the claims of the present invention.

[0078] In addition, in order to better illustrate the present invention, numerous specific details are given in the following specific embodiments. It will be understood by those skilled in the art that the present invention can also be implemented without these specific details.

[0079] See also Figure 1In order to realize dynamic and adaptive data distribution under the data link system architecture, the present invention establishes a data link data dynamic and adaptive distribution module based on data content at the data link data generating end and receiving end.

[0080] Specifically, the present invention provides a data link data dynamic distribution method based on data content, including a data platform for receiving or distributing data link data packets under a data link architecture, the data platform caches its own state information and state information of neighboring platforms, and a data distribution module of the data platform dynamically and adaptively distributes the received data. The specific implementation includes the following steps:

[0081] Step 1: Initialize the distribution rule library. Each data platform loads the distribution rules into its own data distribution module. The distribution rules include the attribute information required for reasoning.

[0082] Step 2: The data platform obtains the original data packet sent by the data generation end, and parses the data elements of the original data packet according to the data packet data structure and data element dictionary of the data link to obtain a parsed data packet;

[0083] Step 3: Assuming that the distribution rule of this data packet has been matched, check whether the parsed data packet includes all the attributes required for distribution reasoning. If all are included, proceed to step 6; if some attributes required for distribution reasoning are not included in this data packet, these attributes are called widened attributes, and proceed to the next step;

[0084] Step 4: widen the parsed data packet, that is, add a widened attribute field to the parsed data packet to obtain a widened data packet, and the widened part is called a widened field;

[0085] Step 5: Call the corresponding preprocessing function of each widening attribute to perform data preprocessing, obtain the widening attribute value, and assign it to the widening data field;

[0086] Step 5.1, for each widening attribute, obtain the data required to calculate the widening attribute from the local resource cache status information or the parsed data packet content;

[0087] Step 5.2: Perform mathematical operations on the acquired data required for calculating the widening attribute, obtain the widening attribute value and assign it to the widening field;

[0088] Step 6: Based on the attributes required for distribution reasoning in the widened data packet (if not widened, based on the parsed data packet), the relative address of the distribution destination is inferred according to the distribution rules.

[0089] Step 7: Convert the relative address of the distribution destination to the absolute address of the distribution destination.

[0090] Step 8: Restore the widened data packet (or the parsed data packet if it is not widened) to the original data packet and distribute it to the distribution destination absolute address obtained in step 7.

[0091] In the above scheme, regarding how to parse the data link data packet, the formatted message transmitted in the data link uses a certain bit sequence to represent the context content, and the data link data element dictionary specifies the content, data item definition and usage scope of the data elements used in the data link, and uses the data field identifier (DFI) and the data usage domain identifier (DUI) as keywords to identify the data element items, that is, the data element is a digital unique identifier consisting of a pair of DFI and DUI. Since the construction and identification of each message completely depends on the data structure and data element dictionary of the message, and the data packet data structure and data element dictionary are opaque, the parsing process is to first identify the original data packet message header to obtain the message word type, then identify the message data structure according to the message identification field in this message word type, and finally quickly index the data element dictionary according to the message data structure, the DFI and DUI of the data element, parse the data elements in the original data packet, and obtain the parsed data packet.

[0092] For how to widen the parsed data packet, add a field of the data type size at the end of the parsed data packet according to the data type of the widening attribute. Figure 2 , and name it so that the attribute value can be obtained in the form of "data packet ['attribute name']" in distribution rule reasoning.

[0093] Regarding how to perform preprocessing to obtain the widened attribute value, corresponding preprocessing functions are preset for the attributes required in the distribution rules. When a widened field is added to the parsed data packet, the preprocessing function corresponding to this widened attribute is called to obtain the data required to calculate this attribute from the local resource cache status information or the content of the parsed data packet, perform mathematical operations on it, obtain the attribute value and assign it to the widened field.

[0094] Regarding how to convert the relative address of the distribution destination into an absolute address, in order to increase the reusability of the distribution rules, the above step 6 stipulates that the result of the distribution rule reasoning is the relative address between the platforms in the task group. After the relative address of the data packet distribution destination is obtained through distribution reasoning, the absolute address of the data packet distribution destination is mapped according to the relationship between the platforms in the task group.

[0095] For example: if the distribution rule conditions are met, it will be distributed to the maser platform. Here, the relative address of the data packet distribution destination is master. In different collaborative groups, different platforms serve as master platforms. In task group A, "Platform A" serves as the master platform, and in task group B, "Platform B" serves as the master. Therefore, in task group A, the relative address of the master platform will be mapped to the absolute address of "Platform A", and in task group B, it will be mapped to the absolute address of "Platform B".

[0096] Regarding how to restore the widened data packet to the original data packet for distribution, the widened data packet releases the widened field to obtain the parsed data packet, and the parsed data packet is de-parsed according to the data link data structure and data element dictionary to obtain the original data packet, and distributed to the distribution destination absolute address obtained in step 7.

[0097] In the above technical solution, the distribution rules are as follows:

[0098] The distribution rule exists in the form of a script function, the parameter of which is the widened data packet (if not widened, it is the parsed data packet), and the inference result is the relative address of the distribution destination of this data packet;

[0099] Use "if", "and", and "then" logical keywords to describe distribution conditions;

[0100] Use the "send_to_local" action keyword to describe data forwarding to this platform, and the parameter Dst destination defaults to this platform;

[0101] Use the "send_to" action keyword to describe whether the data packet is sent to the local platform or the neighboring platform. If it is sent to the local platform, the parameters include Type destination type and Dst destination; if it is sent to the neighboring platform, the parameters include Dst destination.

[0102] The following is an example of a specific scenario to illustrate the above process:

[0103] Scenario description: There are n+1 collaborative platforms based on data link communication that together form a star-shaped formation T1, including platforms with detection capabilities and platforms with interception capabilities. Each platform carries a resource pool, which uses a resource tree to organize and manage collaborative resources. Collaborative resources include target resources, status resources, neighbor status resources, etc. Platform A is the captain of the T1 formation, and the remaining n collaborative platforms are members of the T1 formation. In the status resources on each platform resource tree, attributes such as "formation to which it belongs", "formation role", and "platform type" are set to identify the collaborative platform.

[0104] For the formation member platform C with detection capability, after detecting the target, if the target is close to the team leader A, it is necessary to distribute the detection data to the formation leader A through the data link. The formation leader A makes a decision based on the target information, the capabilities of each formation member, etc. to dispatch a platform with interception capability to intercept the target.

[0105] The detection data distribution of the formation member platform C uses the data link dynamic data distribution method based on data content of the present invention. See the flowchart for Figure 3 The specific steps are as follows:

[0106] Step 1. Write distribution rules for this distribution requirement, as shown below. 'master' is the relative address of the distribution destination, which means distribution to the team leader.

[0107]

[0108]

[0109] Step 2: Store the prepared distribution rules into the distribution rule library, initialize the distribution rule library, and load the distribution rules into the data distribution module of the platform by team member C;

[0110] Step 3: The team member platform C obtains the original data packet detected by the detection device, and parses the data elements of the original data packet according to the data link data packet data structure and data element dictionary to obtain the parsed data packet. Figure 4 .

[0111] Step 4: Match the distribution rule written above, which requires the attributes of "data type" and "distance". It is determined that the parsed data packet only includes the "data type" attribute, but not the "distance" attribute. The "distance" attribute is called the widening attribute.

[0112] Step 5: The "distance" attribute is represented by int data 0 or 1, 0 for close and 1 for far. Open a 4-byte int space at the end of the parsed data packet to widen the "distance" attribute field. Figure 5 , get the widened data packet, and the "distance" attribute field is the widened field;

[0113] Step 6. Call the preset preprocessing function for calculating the "distance" attribute. First, obtain the "target position" attribute value from the parsed data packet, and obtain the "distance judgment threshold" and "captain A position" attribute values ​​from the resource tree of this platform C. Assume that the "distance judgment threshold" is 500 meters, then calculate the straight-line distance between the target position and the captain's position, and then compare it with 500 meters. If it is greater than 500 meters, assign 1 to the widened field "distance", otherwise assign 0, see Figure 6 ;

[0114] Step 7: Pass the widened data packet as a parameter into the distribution rule, execute the distribution rule, and obtain the distribution destination relative address as "master";

[0115] Step 8, "master" means the captain of the formation. Since the attributes such as "formation to which", "formation role", and "platform type" are set in the status resources on each platform resource tree to identify the collaborative platform, the status information of the neighboring platforms in the resource tree is queried, and the captain of the formation to which platform C belongs is obtained as platform A, that is, the absolute address of the distribution destination.

[0116] Step 9: Reversely execute step 3 according to the data link data structure and data element dictionary, release the widened field, restore the widened data packet to the original data packet, and distribute the original data packet to the distribution destination platform A obtained in step 8 through the data link.

Claims

1. A data link data dynamic distribution method based on data content, comprising a data platform for receiving or distributing data packets under a data link architecture, the data platform caches its own state information and state information of neighboring platforms, and a data distribution module of the data platform dynamically and adaptively distributes the received data. It is characterized in that Dynamic adaptive distribution includes the following steps: Step 1: Initialize the distribution rule library. Each data platform loads the distribution rules into its own data distribution module. The distribution rules include the attribute information required for reasoning. Step 2: The data platform obtains the original data packet sent by the data generation end, and parses the data elements of the original data packet according to the data packet data structure and data element dictionary of the data link to obtain a parsed data packet; Step 3: If the distribution rule of this data packet has been matched, check whether the parsed data packet includes all the attributes required for distribution reasoning. If all are included, execute step 6; if some attributes required for distribution reasoning are not included in this data packet, these attributes are called widening attributes, and execute step 4; Step 4: widen the parsed data packet, that is, add a widened attribute field to the parsed data packet to obtain a widened data packet, and the widened part is called a widened field; Step 5: Call the corresponding preprocessing function of each widening attribute to perform data preprocessing, obtain the widening attribute value, and assign it to the widening data field; Step 6: If it is not widened, then the relative address of the distribution destination is inferred based on the parsed data packet, otherwise the relative address is inferred based on the distribution rule based on the attributes required for distribution inference in the widened data packet; Step 7: Convert the relative address of the distribution destination into the absolute address of the distribution destination; Step 8: If the data packet is not widened, it is restored to the original data packet after parsing and distributed to the distribution destination absolute address obtained in step 7; otherwise, the widened data packet is restored to the original data packet and distributed to the distribution destination absolute address obtained in step 7.

2. A data link data dynamic distribution method based on data content according to claim 1, It is characterized in that Step 2 is as follows: The construction and identification of each data link message are completely dependent on the data structure and data element dictionary of the data link message, and the data packet data structure and data element dictionary are opaque, so first identify the original data packet message header to get the message word type, then identify the message data structure based on the message identification field in this message word type, and finally quickly index the data element dictionary based on the message data structure, the DFI and DUI of the data element, parse the data elements in the original data packet, and get the parsed data packet.

3. A data link data dynamic distribution method based on data content according to claim 1, It is characterized in that Step 4 is as follows: According to the data type of the widened attribute, a field of the size of this data type is added to the end of the parsed data packet and named so that the attribute value can be obtained in the form of "data packet ['attribute name']" in the distribution rule reasoning.

4. A data link data dynamic distribution method based on data content according to claim 1, It is characterized in that In step 5: Corresponding preprocessing functions are preset for the attributes required in the distribution rules. When a widened field is added to the parsed data packet, the preprocessing function corresponding to the widened attribute is called to obtain the data required to calculate this attribute from the local resource cache status information or the content of the parsed data packet, perform mathematical operations on it, obtain the attribute value and assign it to the widened field.

5. A data link data dynamic distribution method based on data content according to claim 1, It is characterized in that The distribution rules described in step 6 are as follows: The distribution rule exists in the form of a script function, the parameter of which is the widened data packet or the parsed data packet, and the inference result is the relative address of the distribution destination of this data packet; Use "if", "and", "then" logical keywords to describe distribution conditions; Use the "send_to_local" action keyword to describe data forwarding to this platform, and the parameter Dst defaults to this platform; Use the "send_to" action keyword to describe the data packet sent to the local platform or the neighboring platform. If it is sent to the local platform, the parameters include Type destination type and Dst destination; if it is sent to the neighboring platform, the parameters include Dst destination.

6. A data link data dynamic distribution method based on data content according to claim 1, It is characterized in that Step 7 is as follows: In order to increase the reusability of the distribution rules, the result of the distribution rule reasoning specified in the above step 6 is the relative address between the platforms in the task group. After the relative address of the data packet distribution destination is obtained through distribution reasoning, the absolute address of the data packet distribution destination is mapped according to the relationship between the platforms in the task group.

7. A data link data dynamic distribution method based on data content according to claim 1, It is characterized in that Step 8 is as follows: For the widened data packet, release the widened field to obtain the parsed data packet, and de-parse the parsed data packet according to the data link data structure and data element dictionary to obtain the original data packet, and distribute it to the distribution destination absolute address obtained in step 7.

8. A data link data dynamic distribution device based on data content, comprising a data platform for receiving or distributing data packets under a data link architecture, the data platform caches its own status information and the status information of neighboring platforms, and a data distribution module of the data platform dynamically and adaptively distributes the received data. It is characterized in that Dynamic adaptive distribution is achieved through the following modules: Distribution rule base initialization module: Distribution rule base initialization, each data platform loads distribution rules into its own data distribution module, the distribution rules include attribute information required for reasoning; The original data packet parsing module and the data platform obtain the original data packet sent by the data generation end, parse the data elements of the original data packet according to the data packet data structure and data element dictionary of the data link, and obtain the parsed data packet; The judgment module checks whether the parsed data packet contains all the attributes required for distribution reasoning if the distribution rule of the data packet has been matched. If all the attributes are included, the reasoning module is executed. If some attributes required for distribution reasoning are not included in the data packet, these attributes are called widening attributes, and the widening module is executed. The widening module widens the parsed data packet, that is, adds a widening attribute field to the parsed data packet to obtain a widened data packet. The widened part is called a widening field. The preprocessing function corresponding to each widening attribute is called to preprocess the data to obtain the widening attribute value and assign it to the widening data field. The reasoning module, if not widened, then based on the parsed data packet, otherwise based on the attributes required for distribution reasoning in the widened data packet, infers the relative address of the distribution destination according to the distribution rules; An address conversion module converts the relative address of the distribution destination into the absolute address of the distribution destination; The data packet restoration module, if not widened, restores the parsed data packet to the original data packet and distributes it to the distribution destination absolute address obtained in the address translation module; otherwise, the widened data packet is restored to the original data packet and distributed to the distribution destination absolute address obtained in the address translation module.

9. A storage medium, It is characterized in that When the processor executes the program in the storage medium, a data link data dynamic distribution method based on data content as described in any one of claims 1 to 7 is implemented.

Images