TWI569225B - A Method of Estimating Traffic Information Based on Multiple Regression Model of Mobile Network Signaling - Google Patents

Description

Multivariate regression model traffic information estimation method using mobile network signaling

The invention relates to a multi-regression model traffic information estimation method using mobile network signaling. Specifically, it refers to a specific signaling data generated between mobile mobile users and the network, and designs a mobile network signaling vehicle speed estimation algorithm, comprehensively analyzes handover signaling estimated vehicle speed, general location update signaling, estimated vehicle speed, and multiple regression. The model estimates the speed of the vehicle to estimate the traffic information within one or more specified road ranges (taking the speed information as an example) to generate one or more systems and methods for specifying traffic information for the road range.

The application of mobile phone location technology to instant traffic information collection has become one of the hot topics in the field of transportation research. Due to its small investment, large coverage and abundant data, it is in the field of Intelligent Transportation System (ITS). Has great application potential. At present, a large number of theoretical and practical test researches have been developed at home and abroad, such as: UK ITIS, University of Maryland, Berkeley University, French Transportation Research Association, etc., mainly for location update (LU), routing area update (Routing) Area Update, RAU), and Handover to calculate instant traffic information. At present, the main methods can be divided into the following three methods: (1) monitoring the number of mobile phones in the control and voice transmission channels, (2) estimating the vehicle speed using the mobile phone positioning mechanism, and (3) handover signaling and general location update signaling ( The Normal Location Update (NLU) positioning mechanism estimates the speed of the vehicle.

Number of mobile phones monitoring and controlling voice transmission channels: Since the mobile phone moves to a different location area (LA) or routing area (RA) when the mobile phone is in standby mode, the mobile phone actively takes the mobile communication network. Location update or routing area update is performed, so the number of statistical LUs and RAUs can be used for traffic estimation. When the mobile phone makes a call, it will connect with the mobile communication network to establish a new call, end the call, and change the procedure. Therefore, it is possible to use the number of new calls, call, change hands, etc. to estimate the current mobile phone on the road. Quantity, and compare with historical data in the past, to observe whether there are abnormal conditions to make inferences about abnormal traffic conditions. However, although this method can reflect the general changes in traffic, it cannot accurately estimate the accurate speed information.

Estimating the speed of the vehicle by using the mobile phone positioning mechanism: The location service (LCS) standard process can be adopted in the method of the mobile phone positioning mechanism, wherein a plurality of positioning methods are mainly defined, including: A-GPS, Cell-ID, TOA (Time) Of Arrival), TDOA (Time Difference of Arriva), E-OTD (Enhanced Observed Time Difference), and OTDOA (Observed Time Difference of Arrival). After accurately positioning the location of the mobile phone through the LCS method, the mobile phone moving speed can be estimated by continuously obtaining the positioning information and the time difference twice or more. However, although this method can accurately locate the location of the mobile phone and the speed information, it needs to install a specific program on the mobile phone or it will cause a large burden on the network.

Handover signaling and general location update signaling information positioning mechanism to estimate the speed of the vehicle: In the mobile network, the base station (BS) can transmit wireless signals to provide network services, and the service area covered by the wireless signal is called a cell ( Cell). Each cell has a unique CGI (Cell Global Identity), and the base station periodically broadcasts its CGI to all mobile phones in the cell. When a mobile phone (hereinafter referred to as a Mobile Station (MS)) is connected to the mobile network, the MS will acquire and record the CGI of the cell in which it is located. When the MS dials out or connects to the call, it will be connected to the base station. At this point, if the MS in the call Handover signaling will occur when moving from the current cell to another cell. By estimating the time difference between the location of the handover signaling and the two handover signaling, the moving speed of the MS can be estimated accordingly. Similarly, general location update signaling will occur when the handset moves from the current location area to another location area. Therefore, it is possible to estimate the mobile condition and the moving speed of the mobile phone by using the general location update signaling twice. However, although this method can obtain the vehicle speed information, the number of valid samples for obtaining the vehicle speed by using the handover signaling is too small; using the general position update signaling to obtain the vehicle speed, because the range of the location area is large, the speed of the small section cannot be immediately reflected. Variety.

It can be seen that there are still many shortcomings in the above-mentioned methods of use, which is not a good design, but needs to be improved.

In view of the shortcomings derived from the above-mentioned conventional methods, the inventor of the present invention has improved and innovated, and after years of painstaking research, he finally succeeded in research and development of the multi-regression model traffic information estimation method using mobile network signaling.

The object of the present invention is to provide a multi-regression model traffic information estimation method using mobile network signaling, which first uses a general location update signaling (NLU) to filter and track mobile phones that may be moving on the road, and then reuse The multi-regressive vehicle speed technology comprehensively analyzes the call arrival (Call Arrival, CA) signaling and communication completion (Call Completion, CC) signaling quantity, the number of general location update signaling, and the number of mobile phones on the road to estimate the speed of the vehicle. News. And use the established mobile network signaling speed estimation algorithm to combine the handover signaling to estimate the vehicle speed, the general position update signaling to estimate the vehicle speed, and the multiple return vehicle speed to provide the instantaneous vehicle speed information required by the passerby.

The invention provides a multi-regression model traffic information estimation method using mobile network signaling, which mainly relates to a mobile network signaling vehicle speed estimation system, which first collects mobile network signaling (including handover signaling to estimate vehicle speed). The general location update signaling estimates the vehicle speed, the number of call arrival and call end signaling, the number of general location update signaling, and the number of mobile phones on the road, and then combines the map information of the geographic information system to input the road segment data and the filtering mode. The group divides the road segments according to the location information of the base station and the location area to estimate the location where the handover between the base station and the base station and the general location update signaling occur. The vehicle speed information is generated by using the above various mobile network signalings, and input to the mobile network signaling speed estimation algorithm to obtain the vehicle speed information.

The invention provides a multi-regression model traffic information estimation method using mobile network signaling, which mainly relates to a mobile network signaling vehicle speed estimation algorithm, which uses handover signaling to estimate vehicle speed and general position update signaling to estimate vehicle speed. The five traffic information estimation factors, the number of call arrival and call end signaling, the number of general location update signaling, and the number of mobile phones on the road, are used as the mobile network signal speed estimation data to estimate the vehicle speed information. In addition, the present invention also designs a vehicle speed data collection module, and uses other traffic information estimation methods (including a global positioning system, a vehicle detector, etc.) to obtain each section according to the road section data and the section information cut by the filter module. The vehicle speed history data filters out the mobile phone and its subsequent signaling that generate general location update signaling on the observed road segment, and estimates the vehicle speed, the number of call arrival and call end signaling, and the number of general location update signaling with general location update signaling. And the number of mobile phones on the road input section speed history data machine learning module for data training and learning, the multi-regression model to estimate the relevant parameters in the vehicle speed method to provide subsequent vehicle speed estimation use. Finally, the results of the handover signaling estimated vehicle speed, the general position update signaling estimated vehicle speed, and the multiple regression model estimated vehicle speed are respectively input to the mobile network signaling vehicle speed estimation algorithm to obtain the vehicle speed information.

101‧‧‧Mobile Network Signaling Monitoring Module

102‧‧‧Base station and location area data module

103‧‧‧Geographic Information System

104‧‧‧Road section data and filter module

105‧‧‧Mobile network signaling speed estimation data module

106‧‧‧Speed Information Collection Module

107‧‧‧Segment speed history data machine learning module

108‧‧‧Mobile network signaling speed estimation method module

200‧‧‧Mobile mobile phones on the road

201‧‧‧Base station of location area 0 (LA0)

202‧‧‧Base station of location area 1 (LA1)

203~206‧‧‧Base station of location area 2 (LA2)

207‧‧‧Base station of location area 3 (LA3)

208‧‧‧NLU _IN signaling and when and where it occurred

209‧‧‧HO1 signalling and when and where it occurred

210‧‧‧HO2 signaling and when and where it occurred

211‧‧‧HO3 Signaling and when and where it occurred

212‧‧‧NLU _OUT or HO4 signaling and when and where it occurred

301‧‧‧Mobile network signalling for mobile handsets on the road

302‧‧‧Handing Signaling Estimated Vehicle Speed

303‧‧‧General location update signaling to estimate vehicle speed

304‧‧‧Number of call arrival and call completion signaling

305‧‧‧General Location Update Signaling Quantity

306‧‧‧ Number of mobile phones on the road

307‧‧‧Multivariate regression model to estimate vehicle speed

308‧‧‧Mobile network signaling speed estimation algorithm

309‧‧‧Speed Information

1 is a schematic diagram of a system architecture of a mobile network signaling vehicle speed estimation system according to the present invention; FIG. 2 is a schematic diagram of a road section; and FIG. 3 is a schematic diagram of a data model of a mobile network signaling vehicle speed estimation.

Please refer to 1~3. The multi-regression model traffic information estimation method using the mobile network signaling is firstly obtained by the base station and the location area data module 102 to obtain the location information of each base station (including the latitude and longitude). , azimuth, signal strength, and the location area number to which it belongs, combined with the information of the section near the base station (for example, 2 km) that the GIS 103 queries to the database, input to the road segment data and filter module. In 104, the mobile network signaling 301 (including handover signaling, general location update signaling, call arrival signaling, and communication end) of the mobile phone on the road between the base station and the base station is estimated according to the location information of the base station and the location area. The location of the handover signaling in the signaling) and the location of the general location update signaling are used to divide the road segments. The mobile network signaling 301 of the mobile handset on the road is filtered out as an observation and analysis target. After obtaining the mobile network signaling 301 of the mobile phone on the road, the following five traffic information estimation factors are used as the mobile network signaling speed estimation data to estimate the vehicle speed information:

a. Handover signaling estimated vehicle speed 302

b. General Location Update Signaling Estimated Vehicle Speed 303

c. Call arrival and call end signaling number 304

d. General location update signaling number 305

e. Number of mobile phones on the road 306.

In addition, the present invention obtains the speed history data of each section by using the vehicle speed data collection module 106 (Global Positioning System Detector or Vehicle Detector) according to the road section data and the section information cut by the filtering module 104, and is matched with the general position. Update signaling estimation vehicle speed 303, call arrival and call end signaling quantity 304, general location update signaling quantity 305, and mobile phone number 306 on the road input road speed history data machine learning module 107 for data training and learning, and obtain multiple The regression model estimates the relevant parameters in vehicle speed 307 to provide subsequent vehicle speed estimation use. Finally, the results of the handover signaling estimated vehicle speed 302, the general position update signaling estimated vehicle speed 303, and the multiple regression model estimated vehicle speed 307 are input to the mobile network signaling vehicle speed estimation method module 108 to obtain the vehicle speed information.

In the part cut in the road section, it is mainly considered that general location update signaling will occur when the mobile phone crosses the location area, and handover signaling will occur when the mobile phone is in the middle of the call. Taking FIG. 2 as an example, when the mobile phone 200 enters the location area 1 (LA1) 202 from the location area 0 (LA0) 201, general location update signaling will occur, and the mobile network signaling of the mobile mobile phone on the road will be Filter out and continuously collect and analyze the subsequent signaling of this phone. When the mobile phone enters the general location update signaling that occurs in location area 2 (LA2) 203, it is a type of signaling for entering the road segment for the road segment 1, which will be referred to as NLU _IN later; and occurs when the mobile phone enters LA3. The general location update signaling is a kind of signaling leaving the road segment for the road segment 1, which will be referred to as NLU _OUT . In addition, if the mobile phone enters the new base station transmission range during the call, handover signaling will occur, such as HO1, HO2, HO3 and HO4 in Figure 2, and these signalings can be used to divide the link into the link 2 (NLU). _IN position to HO1 position), road segment 3 (HO1 position to HO2 position), road segment 4 (HO2 position to HO3 position), and road segment 5 (HO3 position to HO4 position); in addition, the present invention also considers that one or more intervals can be The case where the road segments of the handover signaling position are merged, for example, the road segment 6 in FIG. 2 (the HO1 position to the HO3 position).

The data model of the mobile network signaling speed estimation proposed by the present invention (as shown in FIG. 1) is analyzed for the mobile network signaling 301 of the mobile phone that is filtered on the road after filtering the road segment data and the filtering module 104. And causing the mobile network signaling speed estimation data module 105 to move the mobile phone's mobile network signaling 301 on the road on the same road segment (including handover signaling, general location update signaling, call arrival signaling) And communication end signaling) for analysis. After analyzing the mobile network signaling speed estimation data module 105, five traffic information estimation factors can be obtained respectively: handover signaling estimated vehicle speed 302, general location update signaling estimated vehicle speed 303, call arrival and call end signaling number. 304, the general location update signaling number 305, and the number of mobile phones 306 on the road are described as follows.

Handover Signaling Estimated Vehicle Speed: The mobile network signaling vehicle speed estimation data module 105 analyzes the distance difference between the same mobile phone in the road segment (such as the road segment 3, the road segment 4, and the road segment 6 of FIG. 2) And the time difference estimates the speed of the car. For example: the handover signaling link 3 can use the estimated vehicle position l HO1 point occurs and the time t ₂ l ₂ with the position of point ₃ and HO2 occurrence time t _3, l ₂ and L ₃ by way of the path distance with t The speed difference between ₂ and t ₃ is calculated. Among them, since the intersection position between the two cells is distributed in a large interval, the vehicle speed estimated by the second-transfer position is often greatly error-free and has no utility value, for example, the road segment 2 and the road segment of FIG. 5 There will be no delivery signaling to estimate the speed of the vehicle due to the unreachable handover location.

The general location update signaling estimates the vehicle speed: the mobile network signaling vehicle speed estimation data module 105 analyzes the distance difference and the time difference estimated vehicle speed of the same mobile phone on the road segment (such as the road segment 1 of FIG. 2). For example, the handover signaling of the road segment 1 can estimate the vehicle speed using the position l ₁ and the time point t ₁ where the NLU _IN occurs and the position l ₅ and the time point t ₅ where the NLU _OUT occurs, according to the road distance between the l ₁ and the l ₅ The speed difference between t ₁ and t ₅ is calculated.

Call arrival and call end signaling number: The mobile network signaling vehicle speed estimation data module 105 counts all the call arrivals and calls of the mobile phones moving on the road segment (such as the road segment 1 to the road segment 6 in FIG. 2) during the time period. The number of signaling ends. Taking the road segment 2 of FIG. 2 as an example, it is possible to count the call arrival signaling and communication end signaling of the mobile phone moving on the road segment during the time period after the NLU _IN signaling occurs and before the HO1 signaling occurs. This factor will reflect the traffic density of the road segment.

The number of general location update signaling: the mobile network signaling vehicle speed estimation data module 105 collects general location update signaling of all mobile phones moving on the road segment (such as the road segment 1 in FIG. 2) during the time period (see FIG. 2). The number of NLU _IN signaling). This factor will reflect traffic flow on the road.

Number of mobile phones on the road: The mobile network signaling speed estimation data module 105 counts separately (please specify which module to perform statistics?) During the time period, the mobile phone enters the road section (such as the road segment 1 in Figure 2). The number of location update signaling (NLU _IN ) and the number of general location update signaling (NLU _OUT ) that have left the segment, using these two quantities to estimate the number of handsets on the road. For example, the road segment 1 can use the number of NLU _IN plus the number of mobile phones on the road in the previous time period, and then subtract the number of NLU _OUT to estimate the number of mobile phones on the road in the current cycle. This factor will reflect the traffic density of the road segment.

Since the number of call arrival and call end signaling, the number of general location update signaling, and the number of mobile phones on the road are not direct response to vehicle speed information, the present invention uses these multiple factors to combine multiple historical regression models with historical data for machine learning and estimate vehicle speed. . In the part of estimating the vehicle speed 307 in the multiple regression model, the present invention takes the linear regression method as an embodiment and can be divided into two phases: (1) the learning phase and (2) the application phase, which are detailed below.

(1) Learning phase: the vehicle speed data collection module 106 is used to obtain the vehicle speed information of the k time periods from the database, and the mobile network signaling speed estimation data module 105 obtains and counts the corresponding time period call arrival and call end information. Order quantity, general location update signaling number, number of mobile phones on the road, and define the t- th time period of the vehicle speed v _t , call arrival and call end signaling quantity c _t (for example: total in the t- th time period) When the 1000-way call arrival and the 900-way call end record occur, the t- th time period call arrival and call end signaling number c _t is 1900), and the general location update signaling quantity n _t (for example, in the t- th time period) 5000 total record general location update signaling occurs, the number of time periods t general location update signaling _n-t 5000), the phone number P _t on the road (for example: in the first time periods t -1 road p _{t -1} phone number 500, and enter the number of location area update signaling for the general location 4900, this location area away from the general location update signaling in the number of 5000, the period of time t road The phone number is 500 p _t + 4900-5000 = 400). Among them, the present invention can obtain higher accuracy by using the time attribute to classify the data and calculate the data separately. Therefore, the k time periods are classified according to the time attribute (for example, the part from Monday to Thursday can be divided into one class for work hours, one for work hours, and one for other time periods), and then the road speed history data machine The learning module 107 performs data training by formulas (1) to (3) to obtain linear regression model parameters a _c , b _c , a _n , b _n , a _p , b _p , respectively, to complete the machine learning process.

a _c : call arrival and call end signaling number vehicle speed linear regression model parameter slope

b _c : call arrival and call end signaling number vehicle speed linear regression model parameter intercept

a _n : general position update signaling quantity vehicle speed linear regression model parameter slope

b _n : general position update signaling quantity vehicle speed linear regression model parameter intercept

a _p : slope of the linear regression model parameter of the number of mobile phones on the road

b _p : number of mobile phone numbers on the road, linear regression model parameter intercept

(2) Application phase: When the mobile phone performs general location update signaling on the observed road segment, it will be filtered out of the mobile network signaling of the mobile phone on the road (including handover signaling, general location update signaling, and call arrival). Signaling and communication end signaling), and tracking the subsequent status of the mobile phone signaling of the mobile phone, and thereby counting the instantaneous call arrival and end of call signaling c _r on the road segment, the general location update signaling quantity n _r , the road when the phone number of the p _r, can use the formula (1) to (3) were calculated for each factor corresponding to the vehicle speed reaches the call signaling and call number of the end of the estimated vehicle speed of the linear regression model vehicle speed v _c, Usually the number of location update signaling The vehicle speed linear regression model estimates the vehicle speed v _n , the mobile phone number on the road, the linear regression model estimates the vehicle speed v _p , and sets the vehicle speed call arrival and the call end signaling number. The vehicle speed linear regression model estimates the vehicle speed v _c weight value w _c right right, typically the number of location update signaling linear regression model vehicle speed v _n of the weight value w _n, and the number of the phone on the road linear regression model weight of the vehicle speed v _p value w _p , and then use the formula (4) to perform a weighted average to obtain a multivariate regression model to estimate the vehicle speed v .

v = w _c × v _c + w _n × v _n + w _p × v _p where w _c + w _n + w _p =1 (4)

Finally, the present invention utilizes the mobile network signaling vehicle speed estimation method module to use the mobile network signaling vehicle speed estimation algorithm 308, and the mobile network signaling vehicle speed estimation method module 108 integrates the handover signaling to estimate the vehicle speed 302 and the general location. The information of the updated signaling estimation vehicle speed 303 and the multiple regression model estimated vehicle speed 307 generates the instantaneous vehicle speed information 309 for use by the passerby decision reference. Among them, according to the past research results, the general location update signaling estimated vehicle speed 303 will not be able to immediately reflect the traffic jam information when smoothly entering the traffic jam state, so the present invention is developed in the mobile network signaling vehicle speed estimation method module 108. A traffic jam judges the threshold value v _congestion (for example, a highway of 30 km/h), and determines the traffic jam condition when the handover signal estimation vehicle speed 302 or the multiple regression model estimates that the vehicle speed 307 is less than v _congestion . At this time, it is judged whether there is a handover signaling estimated vehicle speed 302, and if any, the vehicle speed 302 is estimated using handover signaling; otherwise, the vehicle speed 307 is estimated using a multiple regression model. When the whole road section (such as the road section 1 in Fig. 2) has smooth road conditions, the general position update signaling estimated vehicle speed 303 can provide accurate vehicle speed information, so the mobile network signaling speed estimation algorithm 308 will set a speed threshold. The value v _free (for example: 120 km/h for the expressway) is judged as a smooth condition. If the general position update signaling is used to estimate that the vehicle speed 303 is faster than v _free (that is, the entire NLU _IN to NLU _OUT is smooth) The vehicle speed 303 is estimated using general location update signaling. However, when the general position update signaling estimates that the vehicle speed 303 is not smooth, it may be that some road sections are jammed, but some of the road sections are smooth, so the vehicle speed 302 or the multiple regression model estimated vehicle speed 307 information will be used as a reference by using the handover signaling. It will first determine whether there is a handover signaling estimated vehicle speed 302, and if so, the delivery speed is estimated based on the handover signal estimation 302; otherwise, the vehicle speed 307 is estimated based on the multiple regression model. Further, although the network signal interference can cause error in the handover signaling on the estimated vehicle speed 302, but the present inventors have found that when the handover signaling 302 is higher than the estimated vehicle speed threshold limit vehicle speed _v higher (eg: motorway is 90 km / h) When it is reflected that the road section is a relatively smooth condition, the vehicle speed 302 can be estimated by referring to the handover signaling; otherwise, the vehicle speed 307 is estimated with reference to the multiple regression model. In the part of the handover signaling estimated vehicle speed 302, the present invention designs a threshold value v _free for excessive overestimation caused by network signal interference, and if it exceeds the threshold value, returns v _free ; otherwise, directly uses handover signaling estimation. Speed 302.

In summary, the disclosed embodiments provide a multiple regression model traffic information estimation method using mobile network signaling. The technology uses a pre-established mobile network signaling speed estimation system. Perform mobile network signaling acquisition, and collect sample data of communication events such as handover signaling, general location update signaling, call arrival signaling, and communication end signaling of multiple mobile users on a specified road. The mobile network signaling speed estimation algorithm estimates the handover signal estimation vehicle speed, the general position update signaling estimated vehicle speed, the multiple regression vehicle speed, and combines various vehicle speed information to generate an instant according to the mobile network signaling sample data. The speed information is used for reference by passers-by.

The traffic information estimation method based on the analysis mobile network signaling provided by the invention has the following advantages when compared with other conventional technologies:

1. The present invention can extract mobile network signaling, and can analyze and provide traffic information without affecting the use of services by the original mobile users. The data source is all the passers-by, and no need to change any equipment, and improve the traffic information to obtain instantness. And popularity.

2. The present invention utilizes multiple regression vehicle speeds to comprehensively analyze the number of call arrival and call end signaling, the number of general location update signaling, and the number of handsets on the road to analyze the instantaneous vehicle speed information. This method will provide instant information on the speed of small sections and avoid the problem of insufficient number of samples for the estimated speed of handover signaling.

3. The invention establishes a mobile network signaling vehicle speed estimation algorithm, and combines handover signaling to estimate vehicle speed, general position update signaling, estimated vehicle speed, and multiple regression vehicle speed to provide more accurate real-time vehicle speed information.

The detailed description of the present invention is intended to be illustrative of a preferred embodiment of the invention, and is not intended to limit the scope of the invention. The patent scope of this case.

In summary, this case is not only innovative in terms of technical thinking, but also has many of the above-mentioned functions that are not in the traditional methods of the past. It has fully complied with the statutory invention patents of novelty and progressiveness. If you apply in accordance with the law, you are requested to approve the application for the invention patent in order to invent the invention.

101‧‧‧Mobile Network Signaling Monitoring Module

102‧‧‧Base station and location area data module

103‧‧‧Geographic Information System

104‧‧‧Road section data and filter module

105‧‧‧Mobile network signaling speed estimation data module

106‧‧‧Speed Information Collection Module

107‧‧‧Segment speed history data machine learning module

108‧‧‧Mobile network signaling speed estimation method module

Claims (6)

A multi-regression model traffic information estimation method using mobile network signaling, the steps comprising: A. learning phase: a. obtaining location information of a plurality of base stations through a base station and a location area data module; b. The location information of the base station is combined with the map information obtained by the GIS item data and input into the road section data and the filtering module; c. the road section data and the filtering module are used to estimate the base stations based on the location information of the base stations The location of handover signaling and general location update signaling in the mobile network signaling of the mobile phone on the road to divide the road segment to filter out the movement on each road in the road segment where the general location update signaling occurs. Mobile mobile network signaling; d. Next, using the mobile network signaling speed estimation data module to obtain handover signaling estimated vehicle speed, general location update signaling, estimated vehicle speed according to mobile network signaling of mobile mobile phones on the road, The number of call arrival and call end signaling, the number of general location update signaling, and the number of mobile phones on the road, combined with the speed data collection module The collected historical speed is input to the section speed history data machine learning module to carry out data training and learning with regression model to obtain regression model parameters; B. Application stage: a. When the mobile phone has general position update signaling on the road section The road segment data and filtering module will be used to filter the mobile network signaling of the mobile phone on the road of the mobile phone where the general location update signaling occurs, and track the follow-up status of the mobile network signaling of the mobile mobile phone on the road, and then The data model of the mobile network signaling speed estimation is used to collect the handover signal estimated on the road segment, the estimated vehicle speed, the general location update signaling, the estimated vehicle speed, the call arrival and the end of the call signaling, and the general location update signal. The number and the number of mobile phones on the road, combined with the regression model parameters obtained after the training and learning of the machine learning module of the road speed history data, the multivariate regression vehicle speed estimation, the multiple regression model to estimate the vehicle speed; b. the multiple regression model The estimated vehicle speed result is input into the mobile network signaling vehicle speed estimation method module, and the vehicle speed information is estimated. The multiple regression model traffic information estimation method using the mobile network signaling according to claim 1, wherein the mobile network signaling speed estimation method module estimates the vehicle speed by using the handover signaling, the general location update signaling, and the estimated vehicle speed, Estimate the speed information by the number of call arrival and call end signaling, the number of general location update signaling, and the number of mobile phones on the road. The multiple regression model traffic information estimation method using the mobile network signaling as claimed in claim 1, wherein the vehicle speed data collection module is a global positioning system probe vehicle or a vehicle detector. The multiple regression model traffic information estimation method using the mobile network signaling according to claim 1, wherein the step of obtaining the regression model parameters for the multiple regression vehicle speed estimation comprises: (1) collecting the historical vehicle speed by using the vehicle speed data collection module, and Input to the section speed history data machine learning module to use the regression model for data training and output regression model parameter values; and (2) the number of call arrival and call end signaling, the number of general location update signaling, and the road The number of mobile phones is combined with the regression model parameter values to obtain the vehicle speed information estimated by each factor, and weighted average is generated to generate a multiple regression model to estimate the vehicle speed. The multiple regression model traffic information estimation method using the mobile network signaling according to claim 1, wherein the mobile network signaling speed estimation method module estimates the vehicle speed information, and uses the handover signaling to estimate the vehicle speed and the general position. The update signaling estimates the vehicle speed, the multiple regression model estimates the vehicle speed, and combines the final vehicle speed information. The steps include: (1) analyzing whether there is handover signaling to estimate the vehicle speed; and (2) analyzing the handover signaling to estimate the vehicle speed or multiple regression model. It is estimated whether the vehicle speed is lower than the traffic congestion threshold value v _congestion , as a traffic jam judgment; (3) analyzing the general position update signaling to estimate whether the vehicle speed is higher than the vehicle speed threshold value v _free as a smooth judgment; (4) analyzing the handover signaling Estimating whether the vehicle speed is higher than the vehicle speed upper threshold v _higher as the decision judgment for estimating the vehicle speed by using the handover signaling estimation vehicle speed or the multiple regression model; and (5) analyzing the handover signaling to estimate whether the vehicle speed exceeds the vehicle speed threshold v _free , As a judgment of reasonable speed. The multiple regression model traffic information estimation method using mobile network signaling as claimed in claim 1, wherein the mobile network signaling of the mobile phone on the road includes handover signaling, general location update signaling, call arrival signaling, and Communication end signaling.