Internet of vehicles group detection method based on social attributes and vehicle attributes

Abstract

The invention discloses an Internet of Vehicles group detection method based on social attributes and vehicle attributes, and aims to solve the problem that an Internet of Vehicles topological structure formed by fast moving vehicle nodes is unstable. Group detection is carried out in a more accurate and stable manner in combination with social attributes such as occupations, hobbies and interests of vehicle owners, so that the credibility of mutual communication between vehicle nodes belonging to the same group is higher, and available shared data resources are richer.

Description

Technical Field

[0001] The present invention relates to a vehicle networking method, and more particularly to a vehicle networking group detection method. Background Art

[0002] With the development of intelligent transportation systems (ITS), vehicular ad hoc networks (VANETs) have become an important part of intelligent transportation. Through the cooperation between on-vehicle communication devices (OBUs) and roadside infrastructure units (RSUs), VANETs enable vehicles to perform efficient information sharing and collaborative decision-making. However, due to the high dynamic characteristics of vehicles, the topology of VANETs is extremely unstable, resulting in increased data transmission delays and frequent link breaks, which affect communication quality and system performance.

[0003] Existing VANETs group detection methods mainly rely on location-based clustering, movement pattern-based clustering, or topology-based social relationship modeling. Among them, location-based clustering methods usually group by a preset communication range. However, they do not consider the behavioral similarity between vehicles, which easily leads to frequent group reconstructions; movement pattern-based clustering methods use information such as vehicle speed and direction to detect groups. Although they can improve the stability of the topology, they are difficult to cope with sudden traffic flow changes; social relationship-based modeling methods mainly focus on the historical behavior data of vehicle owners. However, in the actual application of VANETs, vehicles change dynamically relatively fast, and simply relying on social relationships may be difficult to adapt to complex traffic scenarios. Summary of the Invention

[0004] Object of the Invention: Aiming at the above-mentioned existing technologies, a vehicle networking group detection method based on social attributes and vehicle attributes is proposed to solve the problem of unstable topology of vehicle networking composed of fast-moving vehicle nodes.

[0005] Technical Solution: A vehicle networking group detection method based on social attributes and vehicle attributes includes:

[0006] After a vehicle node starts and drives onto the road, it selects to join a group according to the following process:

[0007] Step 1.1: If the vehicle node group list is not empty, publish all group numbers to the data management center, update the numbers according to the returned results, and set the current valid field values of all records in the list to 0; if it is empty, do nothing.

[0008] Step 1.2: Receive the group list published by the RSU along the driving route. If no group list is received after driving for several minutes, or no group is successfully joined, go to Step 1.5; otherwise, go to Step 1.3.

[0009] Step 1.3: Select a group to join from the received group list, giving priority to groups that already exist in the vehicle node group list; if none exist, calculate the difference between the vehicle node attributes and the vehicle node attributes of each group, and select the group with the smallest difference that is not greater than 30% of its own attribute value; if there is no eligible group, return to Step 1.2;

[0010] Step 1.4: After determining to join a group, send a group join notice to the local RSU and update the vehicle node group list: if the group already exists, update the group code with the current code of the group; if the group does not exist and the vehicle list has less than 5 entries, directly join; if the group does not exist and the vehicle list has 5 entries, delete the record with the farthest most recent join time and then join;

[0011] Step 1.5: If the vehicle node group list is not empty, select the record with the most recent join time and send the group number to the current RSU in the form of a notice; if it is empty, create a new group number according to the rules and add it to the list, and send the new number to the current RSU; if there is no RSU on the current road section, temporarily store the number and send it to the RSU of the next road section.

[0012] Furthermore, after the vehicle node starts driving, it receives the group list published by the RSUs along the way. If there is a group in the received list that already exists in its own vehicle node group list, set the current valid field value of the corresponding group to 1 in its own group list, update the most recent join time to the current time, and send a group join notice to the RSU to update and activate the joined group.

[0013] Furthermore, the RSU broadcasts and publishes its own group list every 30 seconds. When it receives a group join notice from a vehicle node, if the group is not in the local group list, add it; if it already exists, update the last join time field of the group to the current time.

[0014] Furthermore, the RSU regularly checks its own group list. If the most recent vehicle node join time of a group record exceeds 5 minutes from the current time, notify the group record to the data management center and delete it from the local group list.

[0015] Furthermore, the data management center receives the group record notifications from the RSUs in real time, checks whether there are the same or similar group numbers. If a number conflict or similarity is detected, merge them into the existing group according to the attribute similarity, and then store the notified group number and the last active time information in the group record list; if the received group number already exists, update its last active time.

[0016] Furthermore, at a fixed time point every day, perform the following processing on the local group records of the data management center:

[0017] Clean up expired groups: Delete groups whose last active time exceeds 7 days from the group record list and group number comparison table, and delete their associated mapping records in the group number comparison table to optimize storage management.

[0018] Merge similar groups: Use the K-Means algorithm to randomly select a central node, calculate the difference in group numbers, and assign the group to a new group where the central node with the smallest difference is located; then recalculate the new center; repeat clustering until the center remains unchanged; then renumber the merged group and update the comparison table; finally, form a comparison table of the original and current group numbers.

[0019] Furthermore, the vehicle node is coded according to the five social attributes of the owner's occupation, hobbies, vehicle type, average vehicle speed, and average weekly driving time, and the five codes are combined with the factor weights to obtain a 16-bit binary value of the vehicle node attribute.

[0020] Furthermore, the owner's occupation is classified and encoded according to the three-level structure of major category-minor category-specific occupation, and the corresponding coding bits are 4 bits, 4 bits, and 8 bits, respectively, to form a 16-bit owner's occupation code; interests and hobbies are classified and encoded according to the three-level structure of major category-minor category-specific hobby, and the corresponding coding bits are 3 bits, 4 bits, and 5 bits, respectively, and 4 bits of 0 are added to the end of the code to form a 16-bit owner's interest and hobby code; the vehicle type code is encoded in the order of vehicle price-vehicle type-vehicle brand model, which is 3 bits, 3 bits, and 10 bits, respectively, to form a 16-bit vehicle type code; the average vehicle speed is encoded as 16 bits, the first 8 bits represent kilometers per hour in binary, and the last 8 bits are filled with zeros; the average weekly driving time is encoded as 16 bits, the first 6 bits represent the number of driving hours per week in binary, and the last 10 bits are filled with zeros.

[0021] Furthermore, the factor weights of the five social attributes are 0.19, 0.19, 0.22, 0.11, and 0.29, respectively. The binary value of the vehicle node attribute = 0.19*owner occupation code + 0.19*interest code + 0.22*vehicle type code + 0.11*vehicle average speed code + 0.29*weekly average driving time code; when calculating, each code is converted into a corresponding integer for calculation, and then the sum result is rounded up, and the rounded result is converted into a 16-bit binary number, which is the binary value of the vehicle node attribute.

[0022] Furthermore, the group number has a total of 24 bits, in the form of group vehicle node attribute value + city number + district number, and the number of coding bits corresponding to each part is 16 bits, 3 bits, and 5 bits respectively.

[0023] Beneficial effects: The present invention proposes a vehicle networking group detection method based on social attributes and vehicle attributes, which effectively improves the stability, communication efficiency, and security of groups in a high-speed vehicle movement environment. Compared with traditional group detection methods based on geographical location or movement patterns, the innovations and advantages of the present invention are reflected in the following aspects:

[0024] 1. Improve group stability and reduce frequent reorganization: Traditional vehicle clustering only considers geographical location or movement patterns, and is prone to frequent changes in the group structure due to the high-speed movement of vehicles, increasing communication overhead. By introducing vehicle attributes (such as vehicle type, average speed, driving time, etc.) and social attributes (such as occupation, hobbies, etc.), the present invention realizes more accurate group detection, making the vehicles within the group closer in behavior patterns, thus reducing the frequent changes of group members and improving topological stability.

[0025] 2. Enhance the internal trust of the group and improve communication security: In the vehicle networking environment, malicious nodes may use group communication to perform data tampering, identity forgery, or denial-of-service attacks. By analyzing the long-term driving behavior of vehicles and the social attributes of vehicle owners, the present invention quantitatively evaluates the trust between vehicles, making it more difficult for malicious nodes to penetrate high-trust groups, thereby improving the security of data exchange.

[0026] 3. Optimize data sharing efficiency and improve resource utilization: Existing methods usually perform group detection based on communication range or proximity in location, resulting in large differences in the actual data requirements of vehicles within the group, thus reducing the matching degree of data sharing. By detecting vehicles with high attribute similarity into the same group, the present invention makes the points of interest and demands within the group more consistent, thereby improving the matching degree and transmission efficiency of data sharing. For example, vehicle owners with the same hobbies are more likely to form data exchanges in certain services (such as music, navigation, parking information sharing), improving bandwidth utilization. Specific implementation manners

[0027] The following further explains the present invention.

[0028] The main entity objects involved in the vehicle networking architecture used in the present invention are: vehicle nodes, RSU (Road Side Unit, roadside facility unit), and data management center. In some sections, RSU may not be deployed.

[0029] 1. Vehicle node attributes

[0030] The vehicle node has communication capabilities and can communicate wirelessly with other vehicle nodes and RSUs. The vehicle node can process the received data. Each vehicle node is encoded separately according to five social attributes: the owner's occupation, hobbies, vehicle type, average vehicle speed, and average weekly driving time. And the above 5 encoded factors are combined with factor weights and merged and calculated to obtain a 16-bit binary value of the vehicle node attribute.

[0031] (1) The owner's occupation is classified and encoded according to a three-level structure of major category - minor category - specific occupation, and the corresponding number of encoded bits is 4 bits, 4 bits, and 8 bits respectively, thus forming a 16-bit owner occupation code. In this embodiment, one encoding method of the owner's occupation is shown in Table 1, where for the case where there are only two levels of classification, the encoding of the missing level is supplemented with 0 of 4 bits.

[0032] Table 1 Occupation Encoding

[0033]

[0034]

[0035]

[0036]

[0037]

[0038]

[0039]

[0040]

[0041]

[0042]

[0043]

[0044]

[0045]

[0046]

[0047]

[0048] (2) The owner's hobbies are classified and encoded according to the three - level structure of major category - minor category - specific hobby. The corresponding number of encoding bits are 3bit, 4bit, and 5bit respectively. Add 4bit of 0 at the end of this encoding to form a 16 - bit owner's hobby encoding. In this embodiment, an encoding method of the owner's hobby encoding is shown in Table 2.

[0049] Table 2 Hobby Encoding

[0050]

[0051]

[0052]

[0053]

[0054] (3) As shown in Table 3, the vehicle type encoding is encoded in the order of vehicle price - vehicle type - vehicle brand model, which are 3bit, 3bit, and 10bit respectively, so as to form a 16 - bit vehicle type encoding.

[0055] Table 3 Vehicle Type Encoding

[0056] Vehicle price (3 bits) Vehicle type (3 bits) Vehicle brand model (10 bits)

[0057] In this embodiment, a vehicle price encoding method is as follows: 001 represents that the vehicle price is between 0 - 50,000 (excluding); 010 represents that the vehicle price is between 50,000 - 100,000 (excluding); 011 represents that the vehicle price is between 100,000 - 200,000 (excluding); 100 represents that the vehicle price is between 200,000 - 300,000 (excluding); 101 represents that the vehicle price is between 300,000 - 500,000 (excluding); 110 represents that the vehicle price is between 500,000 - 1,000,000 (excluding); 111 represents that the vehicle price exceeds 1,000,000.

[0058] In this embodiment, a vehicle type encoding method is as follows: 001 represents that the vehicle type is a gasoline sedan; 010 represents that the vehicle type is a gasoline SUV; 011 represents that the vehicle type is a pure - electric sedan; 100 represents that the vehicle type is a pure - electric SUV; 101 represents that the vehicle type is a plug - in hybrid sedan; 110 represents that the vehicle type is a plug - in hybrid SUV; 111 represents that the vehicle type is other (other types of vehicles not belonging to the above types).

[0059] In this embodiment, a coding method for vehicle brand models is as follows: By using a Hash function, the vehicle brand is mapped to a 10-bit binary number, where the form of the vehicle brand includes Chinese characters, English letters, and Arabic numerals. For example, the vehicle brand model "Audi Q3" is hashed and mapped to 1011000001; the vehicle brand model "Leapmotor C11" is hashed and mapped to 0001100010.

[0060] (4) The average vehicle speed is encoded as 16 bits. The first 8 bits represent kilometers per hour in binary, and the last 8 bits are filled with zeros. For example, if a vehicle's average speed is 60 kilometers per hour, the encoding is: 0011000000000000. The average vehicle speed is updated once for each completed driving. The update calculation method is: Average vehicle driving speed 新

[0061] = 0.8 * Average vehicle driving speed 老 + 0.2 * Average vehicle driving speed during this driving.

[0062] (5) The average driving time per week is encoded as 16 bits. The first 6 bits represent the number of hours driven per week in binary (rounded to the nearest integer). The maximum weekly driving time is 63 hours. If the weekly driving time exceeds 63 hours, it is still encoded as 111111, and the last 10 bits are filled with zeros. The average driving time per week is updated once when starting the first drive of each week. The update calculation method is: Average driving time per week 新 = 0.8 * Average driving time per week 老 + 0.2 * Driving time of the previous week.

[0063] The calculation of the binary value of the vehicle node attribute uses the analytic hierarchy process to calculate the weights of each factor. The influence weights of the five social attributes calculated by the geometric mean method are 0.19, 0.19, 0.22, 0.11, and 0.29 respectively. Then, the combined calculation is carried out through the following formula: Vehicle node attribute value (binary number) = 0.19 * Owner's occupation code + 0.19 * Interest hobby code + 0.22 * Vehicle type code + 0.11 * Average vehicle speed code + 0.29 * Average driving time per week code. When calculating, it is necessary to convert each code into the corresponding integer for calculation, then round up the sum result, and convert the rounded-up result into a 16-bit binary number, which is the binary value of the vehicle node attribute.

[0064] The encoding of vehicle node attributes is illustrated with an example: A certain vehicle has a brand model of Audi Q3 and a vehicle price of 220,000 yuan. Since this vehicle model is an SUV, the vehicle type encoding is 1000101011000001 (35521D); the owner's occupation is a college teacher, and the occupation encoding is 0010100000000001 (10241D); the owner's main hobby is calligraphy art, and the encoding is 0100001001000000 (16960D); the average driving speed of the vehicle is 40 kilometers per hour, and the average driving speed encoding is 0010100000000000 (10240D); the average driving time of the vehicle per week is 5 hours, and the average driving time encoding per week is 0001010000000000 (5120D); thus, the vehicle node attribute value is: 0011110011101010 (15594D).

[0065] 2. Group Number

[0066] The group number consists of 24 bits in total. The group is autonomously created by vehicle nodes, and the data management center can merge or delete existing groups. The group number adopts the vehicle node attribute value of the group vehicles (16 bits) + city number (3 bits) + urban area number (5 bits).

[0067] Among them, the vehicle node attribute value of the group vehicles (16 bits) is the average value of the vehicle node attribute values of all vehicles joining this group. This value is updated by the data management center once a day.

[0068] The city number is 3 bits, and the urban area number is 5 bits. For counties (cities), the urban area number can be the township number. Usually, this invention is only applied to one city, and the 3-bit city number can be set to all 0s. Taking Chongchuan District, Nantong City, Jiangsu Province as an example in this embodiment, the example prefix of the group number is shown in Table 4.

[0069] Table 4 Urban Area Encoding of Nantong City

[0070]

[0071] In this invention, the data communication between vehicle nodes, between vehicle nodes and RSU, and between RSU and the data management center can be realized by wireless or wired means, and the communication is reliable, that is, this invention focuses on the application layer and does not involve low-level communication methods. Each vehicle node completes the encoding of vehicle attribute nodes according to the aforementioned vehicle node attribute encoding rules. Each vehicle node locally stores a list of joined groups. In this embodiment, the list can store at most 5 joined groups at most. The structure of the vehicle node group list is as follows:[[]]

[0072] Group number (24 bits) Most recent join time (month 4 bits / day 5 bits) Currently valid (7 bits)

[0073] Among them, the currently valid field is a logical value indicating whether the current vehicle node is in the group. 1 represents being in the group, and 0 represents not being in the group.

[0074] In the present invention, after the vehicle node starts and drives onto the road, it selects to join a group according to the following process:

[0075] Step 1.1: If the vehicle node group list is not empty, publish all group numbers in the group list to the data management center. According to the result returned by the data management center, update the group numbers in the group list, and set the currently valid field values of all records in the vehicle node group list to 0. If the vehicle node group list is empty, no operation is performed in this step.

[0076] Step 1.2: Receive the group list published by the RSU along the driving route. If no group list published by any RSU is received after driving for 3 minutes, or if it fails to successfully join any group, go to Step 1.5; otherwise, go to Step 1.3. Among them, the structure of the group list published by the RSU is as follows:

[0077] Number of groups (8 bits) Group 1 number (24 bits) ... Group n number (24 bits)

[0078] Step 1.3: The vehicle node selects a group from the received group list to join and enters Step 1.4. The selection basis is as follows: If there is a group in the group list record (group code) that already exists in its own vehicle node group list, select this group to join; if there are multiple groups, select the group with the most recent join time closest to the current time to join; if there is no group that already exists in its own vehicle node group list, calculate the difference between the vehicle node attributes and the vehicle node attributes of each group in the group list published by the RSU, and select the group with the smallest difference and the difference not greater than 30% of its own vehicle node attribute value to join. If there is no such group in the group list, return to Step 1.2.

[0079] Step 1.4: After the vehicle node determines to join a certain group, send a group join notice to the RSU where it is located and update its own vehicle node group list: If the group already exists in its own vehicle node group list, update the group code with the current group code and update the most recent join time field to the current time; if the group is not in its own vehicle node group list and the number of records in its own vehicle node group list is less than 5, directly add the group record to its own vehicle node group list; if the group is not in its own vehicle node group list and the number of records in its own vehicle node group list is 5, delete a record with the most recent join time farthest from the current time from the list and add the group record, thus completing the joining of the group.

[0080] Step 1.5: If the list of vehicle node group lists of its own vehicle is not empty, select a record with the closest join time to the current time from the list, set the current valid field value of this record to 1, and publish the group number in the record to the current RSU in the form of a group join notification; if the list of vehicle node group lists of its own vehicle is empty, according to the aforementioned group number rule, create a new group number based on the current location of the vehicle node, and the vehicle node attribute value in the group number is the vehicle node attribute value of this vehicle node. Add this record to the list of vehicle node group lists of its own vehicle, and the corresponding current valid field value is 1. Publish the newly created group number to the current RSU in the form of a group join notification. If there is no RSU on the current road section, the group number can be temporarily stored and published to the RSU of the next road section to complete the group joining.

[0081] In the present invention, after the vehicle node starts driving on the road, it updates and activates the joined group according to the following workflow:

[0082] Receive the group list published by the RSU along the driving route. If there is a group in the group list record (group code) that already exists in the list of vehicle node groups of its own vehicle, in the list of vehicle node groups of its own vehicle, set the current valid field value to 1, update the closest join time to the current time, and at the same time send a group join notification to the RSU.

[0083] In the present invention, the RSU broadcasts and publishes its own group list every 30 seconds and works according to the following workflow:

[0084] Step 3.1: Receive the group join notification from the vehicle node. If the group is not in its own group list, add it to the local group list; if the group is already in the local group list, update the last join time field of the group to the current time. Among them, the structure of the RSU local group list is as follows:

[0085] Group number Most recent vehicle node join time

[0086] Step 3.2: Check its own group list. If the time since the last vehicle node joined a certain group record is more than 5 minutes from the current time, notify this group record to the data management center, delete this record from the local group list, and then return to Step 3.1.

[0087] In the present invention, the data management center performs the following operations:

[0088] Receive the group record announcements from the RSU in real time, and the data management center checks whether there are the same or similar group numbers. If a number conflict or similarity is detected, the system merges them into the existing group according to the attribute similarity. Then, information such as the announced group number and the last active time is stored in the group record list. If the received group number already exists, its last active time is updated. The structure of the group record list in the data management center is as follows:

[0089] Group number Last active time

[0090] Illustrated with an example, the attribute value of a vehicle node is: 0100110101100010 (19810D), and the current group list record is empty. The following group list is received from a certain RSU:

[0091] 3D 3537921D 3059201D 4981761D

[0092] It can be calculated that the average value of the vehicle node attributes in group 3537821D is 13820D,

[0093] The difference between the vehicle node attribute value and the average value of the vehicle node attributes in this group is: (12694 - 13820) / 12694 = -0.089, which is the smallest among the differences of the average values of the vehicle node attributes in the three groups, and the difference is less than 30%. Therefore, this vehicle node can join group 3537921D. After joining, the local group list of this vehicle node is:

[0094] 3537921D Most recent join time 1’ 1D

[0095] At the same time, this vehicle node sends a group join message to the RSU. After receiving the message, the RSU updates the local group list record as:

[0096] 3537921D Time 1’ 3059201D Time 2 4981761D Time 3

[0097] When a vehicle node starts or changes groups, it sends a group join request to the data management center. The data management center looks up the group number cross-reference table. If there is a mapping relationship for the original group number, a new group number is returned; otherwise, the original number is retained. The structure of the group number cross-reference table is as follows:

[0098] Original group number Current group number Last active time

[0099] In the present invention, the following processing is performed on the local group records of the data management center at 0:00 every day:

[0100] Step 4.1: Delete the groups whose last active time exceeds 7 days from the group record list and the group number cross-reference table to optimize storage and management. At the same time, delete all associated mapping records in the group number cross-reference table.

[0101] Step 4.2: Merge similar groups, that is, groups with group numbers differing by only a small number of bits. Use the K-Means algorithm to perform group merging on the group record table, and record the merging result into the group number comparison table. The specific steps are as follows:

[0102] (1) The data management center randomly selects a group as the central node, calculates the difference between the group numbers of other groups and the central node, finds the central point with the smallest difference, and adds this group to the new group represented by this central point;

[0103] (2) Recalculate the center of these groups as the central point for the next clustering;

[0104] (3) Repeat the above steps for clustering until the center of the group no longer changes;

[0105] (4) Reassign numbers to the merged groups and update the group number comparison table;

[0106] (5) Form a group number comparison table with the original group numbers and the current group numbers.

[0107] To solve the problem of unstable topology structure of the vehicle ad-hoc network (VANET) formed by fast-moving vehicle nodes, the method of the present invention constructs clusters of vehicle nodes with similar attributes and consistent driving directions. Vehicle nodes within a cluster can maintain a more stable communication connection. However, clustering of vehicle nodes still cannot effectively solve the problems of communication behavior attacks by malicious vehicle nodes and efficient resource sharing. Therefore, based on clustering, aiming at the high dynamic characteristics of vehicle nodes in VANETs, the present invention proposes a stable, efficient and secure group detection method to improve the communication quality and data sharing ability of the VANET. While retaining the advantage of stable communication of clustered vehicle nodes, it can effectively improve the mutual trust degree among nodes within a group, which helps to improve the sharing efficiency of interested resources.

[0108] The method of the present invention not only comprehensively considers the physical characteristics of vehicles, such as vehicle type, average speed, driving time, etc., but also combines social attributes such as the occupation and hobbies of vehicle owners to perform group detection in a more accurate and stable manner, so that the mutual communication trust degree among vehicle nodes belonging to the same group is higher, and the shared data resources that can be provided are more abundant.

[0109] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A vehicle networking group detection method based on social attributes and vehicle attributes, characterized in that Including: After the vehicle node starts and drives onto the road, it selects to join a group according to the following process: Step 1.1: If the vehicle node group list is not empty, publish all group numbers to the data management center, update the numbers according to the returned results, and set the current valid field values of all records in the list to 0; if it is empty, do nothing. Step 1.2: Receive the group list published by the RSU along the driving route. If no group list is received after driving for several minutes, or no group is successfully joined, go to Step 1.

5. Otherwise, go to Step 1.

3. Step 1.3: Select a group to join from the received group list, and preferentially select a group that already exists in the vehicle node group list; if not, calculate the difference between the vehicle node attributes and the vehicle node attributes of each group, and select the group with the smallest difference and not greater than 30% of its own attribute value. If there is no eligible group, return to Step 1.

2. Step 1.4: After determining to join a group, send a group join notice to the local RSU and update the vehicle node group list: if the group already exists, update the group code with the current code of the group; if the group does not exist and the vehicle list is less than 5 items, directly join. If the group does not exist and the vehicle list is full of 5 items, delete the record with the farthest recent join time and then join. Step 1.5: If the vehicle node group list is not empty, select the record with the most recent join time, and send the group number to the current RSU in the form of a notice; if it is empty, create a new group number according to the rules and add it to the list, and send the new number to the current RSU. If there is no RSU in the current section, temporarily store the number and send it to the RSU in the next section.

2. The vehicle networking group detection method based on social attributes and vehicle attributes according to claim 1, characterized in that After the vehicle node drives on the road, receive the group list published by the RSU along the way. If there is a group in the received list that already exists in its own vehicle node group list, set the current valid field value of the corresponding group in its own group list to 1, update the most recent join time to the current time, and send a group join notice to the RSU to update and activate the joined group.

3. The vehicle networking group detection method based on social attributes and vehicle attributes according to claim 2, wherein The RSU broadcasts and publishes its own group list every 30 seconds. When receiving the group join notice of the vehicle node, if the group is not in the local group list, add it. If it already exists, update the last join time field of the group to the current time.

4. The vehicle networking group detection method based on social attributes and vehicle attributes according to claim 3, wherein The RSU regularly checks its own group list. If the most recent vehicle node join time of a group record exceeds 5 minutes from the current time, notify the group record to the data management center and delete it from the local group list.

5. The vehicle networking group detection method based on social attributes and vehicle attributes according to claim 4, characterized in that, The data management center receives the group record notifications from the RSU in real time, checks whether there are the same or similar group numbers. If a number conflict or similarity is detected, merge them into the existing group according to the attribute similarity, and then store the notified group number and the last active time information in the group record list. If the received group number already exists, update its last active time.

6. The method for detecting a vehicle networking group based on social attributes and vehicle attributes according to claim 5, wherein Perform the following processing on the local group records of the data management center at a fixed time point every day: Clean up expired groups: Delete the groups whose last active time exceeds 7 days from the group record list and the group number comparison table, and delete their associated mapping records in the group number comparison table to optimize storage management. Merge similar groups: Use the K-Means algorithm to randomly select a central node, calculate the difference in group numbers, and assign the group to a new group where the central node with the smallest difference is located; Recalculate the new center again; repeat clustering until the center remains unchanged; Then the merged groups are renumbered and the comparison table is updated; finally, a comparison table of the original and current group numbers is formed.

7. The vehicle networking group detection method based on social attributes and vehicle attributes according to any one of claims 1-6, characterized in that The vehicle node is coded according to the five social attributes of the owner's occupation, hobbies, vehicle type, average vehicle speed, and average weekly driving time. The five codes are combined with the factor weights to obtain a 16-bit binary value of the vehicle node attribute.

8. The vehicle networking group detection method based on social attributes and vehicle attributes according to claim 7, characterized in that, The owner's occupation is classified and coded according to the three-level structure of major category-minor category-specific occupation, and the corresponding coding bits are 4 bits, 4 bits, and 8 bits, respectively, thus forming a 16-bit owner's occupation code; hobbies are classified and coded according to the three-level structure of major category-minor category-specific hobby, and the corresponding coding bits are 3 bits, 4 bits, and 5 bits, respectively. 4 bits of 0 are added to the end of the code to form a 16-bit owner's interest and hobby code; the vehicle type code is encoded in the order of vehicle price-vehicle type-vehicle brand model, which is 3 bits, 3 bits, and 10 bits, respectively, thus forming a 16-bit vehicle type code; the average vehicle speed is encoded as 16 bits, the first 8 bits represent kilometers per hour in binary, and the last 8 bits are filled with zeros; the average weekly driving time is encoded as 16 bits, the first 6 bits represent the number of hours driven per week in binary, and the last 10 bits are filled with zeros.

9. The vehicle networking group detection method based on social attributes and vehicle attributes according to claim 7 or 8, characterized in that, The factor weights of the five social attributes are 0.19, 0.19, 0.22, 0.11, and 0.29, respectively. The binary value of the vehicle node attribute = 0.19*owner occupation code + 0.19*hobby code + 0.22*vehicle type code + 0.11*vehicle average speed code + 0.29*weekly average driving time code; when calculating, each code is converted into a corresponding integer for calculation, and then the sum is rounded up, and the rounded result is converted into a 16-bit binary number, which is the binary value of the vehicle node attribute.

10. The vehicle networking group detection method based on social attributes and vehicle attributes according to claim 7, wherein, The group number has a total of 24 bits, in the form of group vehicle node attribute value + city number + district number, and the corresponding coding bit numbers of each part are 16 bits, 3 bits, and 5 bits respectively.

Images