A Vehicle-mounted GOIP Anti-fraud Method and Device Based on a Spatiotemporal Trajectory Coincidence Degree Algorithm

Abstract

The present invention provides a vehicle-mounted Luo Man Bao - like GOIP anti - fraud method based on a spatio - temporal trajectory coincidence degree algorithm, including: S1) identifying Luo Man Bao - like GOIP fraud numbers based on the signaling behavior characteristics of devices; S2) determining whether the Luo Man Bao - like GOIP fraud numbers are vehicle - mounted Luo Man Bao - like GOIP fraud numbers according to the movement characteristics of the fraud numbers and the multi - number accompanying movement trajectory characteristics; S3) finding the number with the highest base station coincidence degree with the fraud - related Luo Man Bao GOIP device number based on the vehicle - mounted Luo Man Bao GOIP fraud numbers; S4) calculating the spatio - temporal trajectory coincidence degree between this number and the fraud number using the spatio - temporal trajectory coincidence degree algorithm to obtain the fraudster's self - used number. The present invention uses the spatio - temporal trajectory coincidence degree algorithm to identify the fraudster's self - used and active mobile phone numbers hidden behind, which can greatly improve the cracking efficiency of vehicle - mounted Luo Man Bao - like GOIP.

Description

Technical Field

[0001] The present invention relates to the technical field of communication security, and in particular to a vehicle-mounted GOIP anti-fraud method and device based on a spatio-temporal trajectory coincidence degree algorithm. Background Art

[0002] The GOIP device is a virtual dialing device for network communication, which can switch overseas numbers into local mobile phone numbers. In recent years, it has often been used by criminals to carry out telecommunications network fraud.

[0003] Generally, domestic fraud gangs set up GOIP devices for overseas fraud gangs in exchange for rewards. The overseas fraud gangs remotely control these devices overseas to call domestic victims to carry out fraud. Such virtual dialing devices can often insert dozens or even hundreds of SIM cards to make calls at the same time. The overseas fraud gangs use this method to achieve separation of humans and machines, hide themselves, and avoid being cracked down.

[0004] At the same time, in order to avoid being captured at fixed dens, domestic fraudsters often adopt the method of committing crimes by vehicle. They use GOIP to drive around the streets and alleys to commit crimes. The SIM cards used for fraud are discarded only after 30 minutes to 1 hour, which brings great difficulties to the public security organs in tracing the signal source and also poses a severe challenge to the arrest work.

[0005] Currently, the numbers identified by the conventional GOIP model based on operator signaling analysis often have a time delay of at least 5 minutes. The positioning range of the associated base station longitude and latitude information after identification is about 1 - 5 kilometers, and the expected delay when provided to the public security is more than 10 minutes. At this time, the vehicle-mounted GOIP is traveling fast (assuming a vehicle speed of 50 km / h, and it travels 8.3 km in 10 minutes). Even if the public security obtains the corresponding base station range, the number has left the original base station coverage range. Moreover, a fraud number often only uses 30 minutes to 1 hour. Even if the operator uses real-time signaling tracking technology to synchronize the latest base station range with the public security, it is extremely difficult to capture the fraud vehicle of the vehicle-mounted GOIP within the time window of 30 minutes to 1 hour. Summary of the Invention

[0006] Aiming at the deficiencies of the prior art, the present invention provides a vehicle-mounted GOIP anti-fraud method and device based on a spatio-temporal trajectory coincidence degree algorithm.

[0007] The technical solution of the present invention is as follows: A vehicle-mounted GOIP anti-fraud method based on a spatio-temporal trajectory coincidence degree algorithm includes the following steps:

[0008] S1), identifying GOIP fraud numbers based on the signaling behavior characteristics of the device;

[0009] S2), determine whether the GOIP fraud number is a vehicle-mounted GOIP fraud number according to the movement characteristics of the fraud number and the movement trajectory characteristics of multiple numbers;

[0010] S3), based on the vehicle-mounted GOIP fraud number, find out the numbers with a high degree of base station coincidence with the fraud-related GOIP device number;

[0011] S4), after screening out the numbers with a high degree of base station coincidence with the fraud-related GOIP device number on the same day, use the space-time trajectory coincidence algorithm to calculate the space-time trajectory coincidence degree between the vehicle-mounted GOIP fraud number and the number with a high degree of base station coincidence with it. If the space-time trajectory coincidence degree between this number and the fraud-related GOIP device number is similar, this number is used as the self-used number of the fraudster.

[0012] Preferably, in step S1), the signaling behavior characteristics of the device include the characteristics of multiple numbers making outgoing calls under the same base station and multiple numbers powering on and off simultaneously under the same base station.

[0013] Preferably, in step S2), the movement characteristics of the fraud number refer to that the number passes through more than 5 base stations within one day.

[0014] Preferably, in step S2), the same movement trajectory of multiple numbers means that multiple cards on the GOIP device are used for mobile fraud on the mobile device, the movement trajectories of the multiple numbers are the same, and the base station trajectories passed by the multiple numbers in chronological order are similar and the base station coincidence degree is high.

[0015] Preferably, in step S3), since both the fraudster and the fraud-related GOIP device are moving on a mobile device, the self-used mobile phone number of the fraudster coincides with the base stations passed by the fraud-related vehicle-mounted GOIP fraud number on the same day. According to the identified fraud-related vehicle-mounted GOIP fraud number, de-duplicate and count according to the GOIP fraud number dimension, and find out the numbers with a high degree of base station coincidence with the fraud-related vehicle-mounted GOIP fraud number.

[0016] Preferably, in step S3), the specific steps to find out the numbers with a high degree of base station coincidence with the fraud-related vehicle-mounted GOIP fraud number are as follows:

[0017] Select the numbers with a high degree of coincidence according to the base station coincidence degree between the fraud-related GOIP device numbers captured in history and the self-used mobile phone numbers of the fraudsters;

[0018] Or, place a mobile phone in the standby state and the GOIP device on the mobile device, keep moving, and the GOIP device keeps making outgoing calls to simulate the real vehicle-mounted GOIP fraud scenario, and use the base station coincidence degree between the obtained fraud-related GOIP device number and the self-used mobile phone number of the fraudster as a reference value.

[0019] Preferably, in step S4), the spatio-temporal trajectory coincidence degree algorithm combines the time-weighted similarity and the space-weighted similarity to measure the similarity of spatio-temporal trajectories.

[0020] Preferably, in step S4), it specifically includes the following steps:

[0021] S401). Given two consecutive trajectories and , the overlapping time period of the two consecutive trajectories is t ∈ [α, β]. If the points of the two trajectories can be close for a long time or a long distance, then the two trajectories are considered similar;

[0022] Therefore, the time-weighted similarity is defined as:

[0023] ;

[0024] In the formula, represents , 's proximity. The proximity is integrated over time, that is, the length of time when the trajectories fit together, and then divided by the total duration to obtain the time proportion when the trajectories fit together, that is, the time-weighted similarity;

[0025] The space-weighted similarity is defined as:

[0026] ;

[0027] In the formula, refers to 's speed at time ; among them, is 's length, is the integral of the product of the proximity of the point and its speed over time as the distance when the trajectories fit together. Dividing the two is the distance proportion when the trajectories fit together, that is, the space-weighted similarity;

[0028] S402). Since in the actual process, the trajectory data is a discrete sequence of timestamps and positions and cannot be integrated, therefore, the linear interpolation method is used to infer the position of the trajectory between two sampling moments, and then the trapezoidal formula is used to approximately calculate the integral. Specifically:

[0029] ;

[0030] ;

[0031] In the formula, is the number of sampling points, The distance related to the trajectory within the i-th sampling interval;

[0032] During the calculation process, it is necessary to first interpolate the trajectory, then calculate the proximity of the points at the corresponding timestamps, and finally calculate the integral, the time-weighted similarity and the space-weighted similarity The larger it is, the higher the trajectory similarity.

[0033] The beneficial effects of the present invention are as follows:

[0034] 1. The present invention uses the spatio-temporal trajectory coincidence degree algorithm to identify GOIP and find out the self-used active mobile phone numbers of the fraudsters hiding behind, which can greatly improve the cracking efficiency of in-vehicle GOIP;

[0035] 2. The present invention measures the similarity of spatio-temporal trajectories through the time-weighted similarity and the space-weighted similarity to improve the recognition accuracy of the self-used numbers of fraudsters. Specific implementation manner

[0036] The following further describes the specific implementation manner of the present invention in conjunction with the accompanying drawings:

[0037] This embodiment provides a method for preventing fraud of in-vehicle GOIP based on the spatio-temporal trajectory coincidence degree algorithm, including the following steps:

[0038] S1). Identify GOIP fraud numbers based on the signaling behavior characteristics of the device; the signaling behavior characteristics of the device include the characteristics of multiple numbers making outgoing calls under the same base station and multiple numbers powering on and off simultaneously under the same base station.

[0039] S2). Determine whether the GOIP fraud number is an in-vehicle GOIP fraud number according to the movement characteristics of the fraud number and the movement trajectory characteristics of multiple numbers; in this embodiment, the movement characteristics of the fraud number refer to that the number passes through more than 5 base stations within one day.

[0040] Among them, the same movement trajectories of multiple numbers mean that multiple cards on the GOIP device move for fraud on the mobile device, the movement trajectories of multiple numbers are the same, and the base station trajectories passed by multiple numbers in chronological order are similar and the base station coincidence degree is high.

[0041] S3), Identify numbers with a high degree of base station overlap with the GOIP device numbers involved in fraud based on in-vehicle GOIP fraud numbers; since both the fraudster and the GOIP device involved in fraud move on a mobile device, the self-use mobile phone number of the fraudster coincides with the base stations passed by the in-vehicle GOIP fraud number involved in fraud on the same day. According to the identified in-vehicle GOIP fraud numbers involved in fraud, perform deduplication statistics according to the GOIP fraud number dimension to find numbers with a high degree of base station overlap with the in-vehicle GOIP fraud numbers involved in fraud.

[0042] In this embodiment, the specific steps for finding numbers with a high degree of base station overlap with the in-vehicle GOIP fraud numbers involved in fraud are as follows:

[0043] Select numbers with a high degree of overlap based on the base station overlap between the GOIP device numbers involved in fraud captured historically and the self-use mobile phone numbers of fraudsters;

[0044] Or, place a mobile phone in the standby state and the GOIP device on a mobile device, move continuously, and let the GOIP device make continuous outgoing calls to simulate a real in-vehicle GOIP fraud scenario. Use the base station overlap between the GOIP device numbers involved in fraud and the self-use mobile phone numbers of fraudsters obtained thereby as a reference value.

[0045] S4), After screening out numbers with a high degree of overlap with the base stations passed by the GOIP device numbers involved in fraud on the same day, use the spatio-temporal trajectory overlap algorithm to compare the spatio-temporal trajectory overlap and calculate numbers with trajectories similar to the GOIP device numbers involved in fraud.

[0046] The spatio-temporal trajectory overlap algorithm combines time-weighted similarity and space-weighted similarity to measure the similarity of spatio-temporal trajectories.

[0047] Preferably, in step S4), the following steps are specifically included:

[0048] S401), Given two consecutive trajectories and , the overlapping time period of the two consecutive trajectories is t ∈ [α, β]. If the points of the two trajectories can be close for a long time or a long distance, then the two trajectories are considered similar;

[0049] Therefore, define the time-weighted similarity as:

[0050] ;

[0051] In the formula, represents , 's proximity, and the proximity is the integral of time, that is, the length of time the trajectories fit, and then divided by the total duration That is, the time ratio of trajectory fitting, i.e., the time-weighted similarity;

[0052] Define the space-weighted similarity as:

[0053] ;

[0054] In the formula, refers to the speed at time ; among which, is the length of ; the product of the proximity of the point and its speed integrated over time is used as the distance of trajectory fitting, and the division of the two is the distance ratio of trajectory fitting, i.e., the space-weighted similarity;

[0055] S402) In the actual process, since the trajectory data is a discrete time stamp and position sequence and cannot be integrated, therefore, the method of linear interpolation is used to infer the position of the trajectory between two sampling times, and then the trapezoidal formula is used to approximately calculate the integral, specifically:

[0056] ;

[0057] ;

[0058] In the formula, is the number of sampling points, is the distance related to the trajectory within the i-th sampling interval;

[0059] In the calculation process, it is necessary to first interpolate the trajectory, then calculate the proximity of the points at the corresponding time stamps, and finally calculate the integral. The larger the time-weighted similarity and the space-weighted similarity , the higher the trajectory similarity.

[0060] The above embodiments and descriptions in the specification only illustrate the principles and the best embodiments of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed.

Claims

1. A vehicle-mounted GOIP anti-fraud method based on the spatio-temporal trajectory coincidence degree algorithm, characterized in that, it includes the following steps: S1). Identify GOIP fraud numbers based on the signaling behavior characteristics of the device; S2). Determine whether the GOIP fraud numbers identified in step S1) are vehicle-mounted GOIP fraud numbers according to the movement characteristics of the fraud numbers and the multi-number movement trajectory characteristics; S3). Based on the vehicle-mounted GOIP fraud numbers, find the numbers with a high base station coincidence degree with the fraud-related GOIP device numbers; S4). After screening out the numbers with a high coincidence degree of the cumulative base stations passed by the fraud-related GOIP device numbers on the same day, use the spatio-temporal trajectory coincidence degree algorithm to calculate the spatio-temporal trajectory coincidence degree between the vehicle-mounted GOIP fraud numbers and the numbers with a high base station coincidence degree. The spatio-temporal trajectory coincidence degree algorithm combines time-weighted similarity and space-weighted similarity to measure the similarity of spatio-temporal trajectories; specifically includes the following steps: S401), Given two consecutive trajectories and , the time period during which the two consecutive trajectories coincide is t ∈ [α, β]. If the points of the two trajectories can approach each other for a long time or over a long distance, then the two trajectories are considered to be similar; Therefore, define the time-weighted similarity as follows: ； In the formula, represents , proximity, and the proximity is integrated over time, that is, the length of time of trajectory fitting, and then divided by the total duration to obtain the time ratio of trajectory fitting, that is, the time-weighted similarity; Define spatial weighted similarity as follows: ； In the formula, refers to the speed at time ; where is the length of the integral of the product of the proximity of a point and its speed with respect to time as the distance of trajectory fitting. The division of the two is the distance ratio of trajectory fitting, that is, the similarity of spatial weighting; S402). Since in the actual process, the trajectory data is a discrete time stamp and position sequence and cannot be integrated, therefore, the method of linear interpolation is used to infer the position of the trajectory between two sampling moments, and then the trapezoidal formula is used to approximately calculate the integral, specifically as follows: ； ； Wherein, is the number of sampling points, is the distance related to the trajectory within the i-th sampling interval; During the calculation process, it is necessary to first interpolate the trajectory, then calculate the proximity of points at the corresponding timestamps, and finally calculate the integral, the time-weighted similarity and the space-weighted similarity The larger it is, the higher the trajectory similarity 2. A vehicle-mounted GOIP anti-fraud method based on the spatio-temporal trajectory coincidence degree algorithm according to claim 1, characterized in that: In step S1), the signaling behavior characteristics of the device include the characteristics of multiple numbers making outgoing calls under the same base station and multiple numbers powering on and off simultaneously under the same base station.

3. A vehicle-mounted GOIP anti-fraud method based on the spatio-temporal trajectory coincidence degree algorithm according to claim 1, characterized in that: In step S2), the movement characteristic of the fraud number means that the number of base stations passed by the number within one day is greater than 5.

4. A vehicle-mounted GOIP anti-fraud method based on the spatio-temporal trajectory coincidence degree algorithm according to claim 1, characterized in that: In step S2), the same multi-number movement trajectory means that multiple cards on the GOIP device are used for mobile fraud on the mobile device, and the movement trajectories of the multiple numbers are the same, and the base station trajectories passed by the multiple numbers in chronological order are similar and the base station coincidence degree is high.

5. A vehicle-mounted GOIP anti-fraud method based on the spatio-temporal trajectory coincidence degree algorithm according to claim 1, characterized in that: In step S3), according to the identified fraud-related vehicle-mounted GOIP fraud numbers, perform deduplication statistics according to the GOIP fraud number dimension, and find the numbers with a high base station coincidence degree with the fraud-related vehicle-mounted GOIP fraud numbers.

6. A vehicle-mounted GOIP anti-fraud method based on the spatio-temporal trajectory coincidence degree algorithm according to claim 5, characterized in that: The specific steps of finding the numbers with a high base station coincidence degree with the fraud-related vehicle-mounted GOIP fraud numbers in step S3) are: Select the numbers with a high coincidence degree according to the base station coincidence degree of the fraud-related GOIP device numbers captured historically and the mobile phone numbers used by the fraudsters themselves; Alternatively, place a mobile phone in the standby state and a GOIP device on a mobile device, continuously move, and let the GOIP device make continuous outgoing calls to simulate a real in-vehicle GOIP fraud scenario. Use the base station coincidence degree between the GOIP device number involved in fraud and the mobile phone number used by the fraudster obtained thereby as a reference value.

Images