CN110933638B - Heterogeneous network access selection strategy method applied to vehicle following queue - Google Patents
Heterogeneous network access selection strategy method applied to vehicle following queue Download PDFInfo
- Publication number
- CN110933638B CN110933638B CN201911014970.6A CN201911014970A CN110933638B CN 110933638 B CN110933638 B CN 110933638B CN 201911014970 A CN201911014970 A CN 201911014970A CN 110933638 B CN110933638 B CN 110933638B
- Authority
- CN
- China
- Prior art keywords
- network
- vehicle
- value
- matrix
- attribute
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W48/00—Access restriction; Network selection; Access point selection
- H04W48/16—Discovering, processing access restriction or access information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W48/00—Access restriction; Network selection; Access point selection
- H04W48/18—Selecting a network or a communication service
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Abstract
A heterogeneous network access selection strategy method applied to a vehicle following queue aims at the requirement that information data transmission needs to be carried out through a communication network when a vehicle in the vehicle networking is in cruising driving, and under the premise that the vehicle network and vehicle motion state information are considered at the same time, by designing a distance measurement algorithm for determining the relative position of the vehicle and a base station based on signal intensity, screening the network based on a fuzzy control algorithm and determining the network relative weight based on an approximate ideal solution algorithm, the vehicle can select the optimal communication network at the current moment through the network selection algorithm to ensure the communication quality and the correct transmission of the information, the vehicle in the vehicle networking can select the optimal data network under the condition of self-adaptive cruise control, the data information is correctly sent and received, and the error probability of the information is reduced.
Description
Technical Field
The invention belongs to the field of vehicle cruise automatic control, and discloses a heterogeneous network access selection strategy method applied to a vehicle following queue.
Background
In recent years, mobile communication technology has been rapidly developed, and communication networks have played an increasingly important role in human life. Different industries have different requirements on communication services, so a variety of communication standards are derived, and communication networks develop towards diversification, and at present, the communication networks develop into heterogeneous communication networks consisting of different types of networks. In order to meet the comfort of acquiring information and the corresponding communication quality of different crowds in different fields, it is very important to select a proper network. In the 4G era, more families have vehicles with the increase of data transmission speed and the improvement of living standard of people. Along with the sudden increase of the number of vehicles, the vehicle network is developed in order to meet the requirements of ensuring the normal running of the vehicles and avoiding traffic accidents, traffic jams and other conditions. The internet of vehicles is also a heterogeneous network, which comprises a plurality of networks with different standard types, such as 4g, wimax, wlan, and the like, and the network service is mainly used for providing good network service for vehicles and vehicle-mounted personnel, and the communication modes include vehicle-to-vehicle communication, vehicle-to-roadside base station communication, cooperative communication of the vehicle-to-roadside base station communication, and the like. After the communication network and the passing mode exist, how to select the most suitable network for the networked vehicles under the current condition is important, the networked vehicles are simultaneously provided with a plurality of different types of network interface modules, different communication networks can be switched according to commands, and the network selection algorithm has important significance in ensuring the normal operation of the network in the networked vehicles and ensuring the user experience. Macroscopically, through cooperation among various networks, it is desirable to select an optimal network to communicate with other vehicles or roadside infrastructure according to scene requirements, so that the vehicles can obtain good communication experience. In the network selection algorithm, because the meaning of the signal intensity is not great, various properties of each communication network, the motion state of the terminal and the personal preference of the user are comprehensively considered. It is valuable to find a network handover selection algorithm that comprehensively considers the specific requirements of the vehicle and the vehicle-mounted article on the network, can avoid repeated handover of the network, and ensures that the signal strength is sufficient for the user to use.
Disclosure of Invention
In order to consider the problem that poor user experience is caused by repeated switching of vehicles in different networks in the following process under the actual vehicle networking environment, the invention provides a network switching selection algorithm method which is intuitive to understand, simple in design and easy to realize.
The technical scheme adopted by the invention for solving the technical problems is as follows:
a heterogeneous network access selection strategy method applied to a vehicle-following queue comprises the following steps:
1) In the design, the vehicle queue performs linear following, and the vehicle has a sensor module to receive and transmit speed and acceleration information of an adjacent vehicle, so that for switching, the most important is how to judge whether the vehicle is within an effective coverage range of network signals, if all the network signals are considered blindly in the algorithm process, a plurality of invalid calculations and reaction times occur, therefore, firstly, a propagation loss model method is used, the relative distance between the vehicle and a base station is estimated according to the signal intensity detected at the current moment, then, subsequent work can be performed, and the formula of the signal intensity and the distance is shown in formula (1):
wherein R (d) is the measured signal strength at the current moment, d is the distance between the vehicle and the base station at the current moment, and R (d) 0 ) For preset and adjusted signal strength, the unit is dB, d 0 Value and preset network signal intensity R (d) 0 ) In general, the boundary values are related to R (d) 0 ) Are arranged together, X σ The random noise error in the measurement process follows Gaussian distribution; n is a path loss exponent, associated with a particular environment;
when the signal strength at the current time is known, the signal strength between the vehicle and the base station can be obtained by equation (1)And then a predetermined good reception range d 1 Comparing, thereby eliminating the network signals which are too far between a part of vehicles and the base station, and screening the network signals for the first time;
2) The distance measurement algorithm in the previous step knows the relative position relationship between the user and the base station and screens the network, and the next step is the judgment on how to switch and how to optimize the switching;
the input values of the fuzzy control algorithm are received network signal strength RSSI, vehicle moving speed v and vehicle acceleration a respectively, in a fuzzy system, a membership function is the most important part of fuzzification of the input values, and different membership functions are selected, so that different fuzzy results can be obtained;
for the membership degree relation of input variables and output variables, a fuzzy logic system in a fuzzy control decision algorithm uses a triangle and a trapezoid at most in the application field, three fuzzy sets are set for the three input variables, so that 3 x 3=27 needs to be set, and 27 fuzzy rules are used in a fuzzy reasoning stage to deal with each possible input condition;
evaluating the moving speed v and the acceleration a of the terminal user and the received network signal strength RSSI through fuzzy logic to obtain an evaluation value H, wherein the obtained switching evaluation value H can reflect the applicability of each network switching, and after the network with the too low H value is removed, a more accurate candidate network group is obtained again for the step 3) to use;
3) In the environment of the internet of vehicles, factors influencing selection need to be considered in many aspects, the most suitable network is selected for the user by considering the four aspects of the user, the network, the vehicle motion attribute and the terminal, and the eight decision factors are determined to be transmission delay, transmission bandwidth, packet loss rate, jitter, price and power, vehicle speed and vehicle acceleration;
after the acceleration of the vehicle and the speed and the acceleration of the vehicle running are measured at each scanning moment, a decision matrix D with the size of m multiplied by n and containing various decision factors under the current condition of the network can be constructed m×n ：
In the above formula, d ij The value of j attribute of i network is, m in m × n represents m candidate networks in total, n represents the number of decision factors used for evaluating the weight this time, in this time, the influence of eight attributes is considered in the network selection algorithm, so n =8;
since the decision factors have different units and cannot be directly compared, the decision factors need to be normalized before subsequent steps are carried out, so that the difference of each attribute is reduced, the quality of each numerical value is conveniently measured, and a normalized decision matrix R is constructed;
in the above formula, r ij Expressing the parameter value of the attribute j in the i network after the value normalization, wherein the normalization method comprises the following steps:
revenue type:
cost type:
in the above formula max (d) ij ) And min (d) ij ) The maximum number chosen among the attributes j of all candidate networks is called max (d) ij ) Likewise, the minimum number of attribute j picks out among all candidate networks may be relatively referred to as min (d) ij )；
After the normalization of the matrix is completed, the optimal value and the worst value of each attribute are sequentially selected from the candidate network matrix R, and all the optimal values are listed as the optimal matrix R max All the worst values are listed as the worst matrix R min ；
For the decision factor, two types can be distinguished: revenue-type and cost-type. The difference between the two is that the expected value of the decision factor is different, and the expected value for the profit type is as good as the larger value is; while the expectation for cost-type is that the smaller the value the better, the normalization is also different for different kinds of decision factors, and the optimal value for profit-type is max (r) ij ) The worst value is min (r) ij ) The opposite is true for the cost model, and the optimum value is min (r) ij ) The worst value is max (r) ij ) Besides the bandwidth, the other attributes are cost type, and the bandwidth is profit type;
the optimal matrix is therefore represented as:
R max ＝[max(r 1j )min(r 2j )...min(r 8j )] (6)
the worst matrix for this is expressed as:
R min ＝[min(r 1j )max(r 2j )...max(r 8j )] (7)
each candidate network may be normalized by the best-worst matrix:
for the revenue type:
for the cost type:
finally obtaining a constructed normalized decision matrix A
The eigenvector omega corresponding to the maximum eigenvalue of the decision matrix A is the corresponding weight of each attribute, and as the eigenvector omega is an objective algorithm, the weight of the decision attribute is not fixed according to different network conditions, and the weight of the decision attribute needs to be recalculated at each moment;
after weighting the normalization matrix, carrying out final comparison, and constructing a final weighted normalization matrix V by multiplying each network attribute with the weight matrix of the two algorithms m×n ；
v ij ＝w j ×r ij
Wherein w j That is, the normalized value of the weight value of the corresponding attribute j, r ij Then represents the normalized value of the weight attribute j in the ith network, v ij Then representing a numerical value obtained after weighting normalization of the network attribute j in the ith candidate network;
finally, all v of the same network are combined ij Adding the values;
V i finally reflecting the weight value of the candidate network i, comparing the size of each network V, obtaining the candidate network corresponding to the V with the highest weight as the preselected network at the next moment by normalizing the weight of each parameter set before and comparing, and if the difference value of the weight value corresponding to the network maintained by the vehicle at the current moment and the weight value of the preselected network is less than
The vehicle can maintain connection with the vehicle network corresponding to the current moment, and does not select the optimal solution network at the current moment, because the candidate networks in the step 2) can be used by ensuring the normal network of the vehicle through the screening of the fuzzy control, and the vehicle can be subjected to network selection after the network is finally selected to complete communication and communication work.
The technical conception of the invention is as follows: aiming at the requirement that information data transmission is required to be carried out through a communication network when a vehicle in the internet of vehicles cruises to run, on the premise of simultaneously considering the information of the vehicle network and the motion state of the vehicle, a distance measurement algorithm for determining the relative position of the vehicle and a base station based on signal intensity is designed, a network is screened based on a fuzzy control algorithm, and three commonly used network selection algorithms for determining the relative weight of the network based on an approximate ideal solution algorithm are designed, so that the vehicle can select the optimal communication network at the current moment through the network selection algorithm at each scanning moment to ensure the communication quality and the correct transmission of information, the vehicle in the internet of vehicles can select the optimal data network under the condition of self-adaptive cruise control, the data information is correctly sent and received, and the error probability of the information is reduced.
The main execution part of the invention is operated and implemented on the automatic driving control computer of the vehicle. The implementation process of the method can be divided into the following three stages:
1. setting parameters: including model parameters and controller parameters, inputting a sampling time T in a model parameter import interface S And network related parameters, after finishing, the control computer sends the setting data to the computer memory unit RAM for storage;
2. off-line debugging: clicking a 'debugging' button in the configuration interface to enable the cruise control system to enter a controller offline debugging stage, adjusting network parameters in the configuration interface, and obtaining a proper threshold value according to preset network parameters and the formula of the invention in an offline mode.
3. And (3) online operation: the method comprises the steps that a CPU of a main control computer is started to read model parameters and controller parameters, the distance error, the relative speed and the vehicle acceleration of a vehicle ahead of the vehicle at the current moment are obtained through a vehicle-mounted sensor, the vehicle acceleration of the vehicle ahead of the current moment is obtained through wireless channel transmission, the vehicle at each moment obtains relevant information of network signals through a receiver, the network is selected through an algorithm, the best network at the current moment is selected for switching, the fact that the relevant information of cruising running of the vehicle at the next moment can be smoothly transmitted is guaranteed, the relevant parameters of the vehicle are obtained through the vehicle-mounted sensor again in the next control period, the relevant information of the network signals is obtained through the receiver, and the steps are repeated in this way, and therefore the vehicle can obtain the best network experience at each moment.
The invention has the following beneficial effects:
1. in the first step of the algorithm design, the distance between the vehicle and the base station is estimated through the signal strength, so that signals which are obviously screened in the subsequent steps are eliminated, the network signals which need to be considered are further reduced through a fuzzy control algorithm, the matrix calculation during the weight calculation is simplified, the complexity of the algorithm and the time needed by the calculation are reduced, and the vehicle can distinguish and select the network signals in time;
2. the network attribute and the attribute of the vehicle during motion are simultaneously considered in the algorithm, so that the real-time state of the vehicle can be more represented;
3. by means of the fuzzy control algorithm and the weight calculation result, on the premise of guaranteeing the communication quality of the vehicle, energy waste caused by repeated switching of the vehicle among network signals in the driving process can be avoided, and the vehicle following quality is guaranteed.
Drawings
In the off-line debugging process of fig. 1, all preset networks are considered, and the weight value of each network is calculated through the optimal worst algorithm.
Fig. 2 shows the network selection performed by the pilot vehicle at each sampling instant.
Detailed Description
The method of the present invention is further described in detail below with reference to the attached drawings and tables.
Referring to fig. 1 and 2 and tables 1 and 2, a heterogeneous network access selection policy method applied to a vehicle longitudinal following queue includes the following steps:
1) In the design, the vehicle queue performs linear following, and the vehicle is provided with modules such as a sensor and the like to transmit and receive information such as the speed and the acceleration of an adjacent vehicle, so that for switching, the most important is how to judge whether the vehicle is within the effective coverage range of network signals, if all the network signals are considered blindly in the algorithm process, a plurality of invalid calculation and reaction times can occur, therefore, firstly, a propagation loss model method is used, the relative distance between the vehicle and a base station is estimated according to the signal intensity detected at the current moment, then, the subsequent work can be performed, and the formula of the signal intensity and the distance is shown in formula (1):
wherein R (d) is the measured signal strength at the current moment, d is the distance between the vehicle and the base station at the current moment, and R (d) 0 ) For preset and adjusted signal strength, the unit is dB, d 0 Value and preset network signal intensity R (d) 0 ) In general, the boundary values are related to R (d) 0 ) Are arranged together, X σ The random noise error in the measurement process follows Gaussian distribution; n is a path loss exponent, which is context specific;
when the signal intensity at the current time is known, the relative distance d between the vehicle and the base station is obtained by the formula (1), and then the relative distance d is compared with a preset good receiving range d 1 Comparing, thereby eliminating the network signals which are too far between a part of vehicles and the base station, and screening the network signals for the first time;
2) The distance measurement algorithm in the previous step already knows the relative position relationship between the user and the base station and screens the network, and the next step is the judgment on how to switch and how to optimize the switching; therefore, the selection priority of each network signal can be solved according to a fuzzy control algorithm;
the input values of the fuzzy control algorithm at this time are the received network signal strength RSSI, the moving speed v of the vehicle and the vehicle acceleration a respectively. In a fuzzy system, a membership function is the most important part of input value fuzzification, and different membership functions are selected to have different fuzzy results;
the membership degree relation of a fuzzy logic system in the fuzzy control judgment algorithm to input variables and output variables uses triangles and trapezoids at most in the application field. The invention sets three fuzzy sets for three input variables, so 3 × 3=27 needs to be set, and 27 fuzzy rules are used in the fuzzy inference stage to deal with each possible input condition;
evaluating the moving speed v and the acceleration a of the terminal user and the received network signal strength RSSI through fuzzy logic to obtain an evaluation value H, wherein the obtained switching evaluation value H can reflect the switching applicability of each network, and after the network with the too low H value is removed, a more accurate candidate network group is obtained again for the step 3) to use;
3) In the environment of the internet of vehicles, factors influencing selection need to be considered in many aspects, the most suitable network is selected for the user by considering the four aspects of the user, the network, the vehicle motion attribute and the terminal, and the eight decision factors are determined to be transmission delay, transmission bandwidth, packet loss rate, jitter, price and power, vehicle speed and vehicle acceleration;
at each scanning moment, by means of the acceleration of the vehicle and the speed and acceleration of the vehicle running, a containing network with the size of m x n can be constructed under the current conditionDecision matrix D of individual decision factor size m×n ：
In the above formula, d ij The value of j attribute of i network, m in m × n represents m candidate networks, n represents the number of decision factors used for evaluating weight this time, in the network selection algorithm this time, the influence of eight attributes is considered, so n =8;
since the decision factors have different units and cannot be directly compared, the decision factors need to be normalized before subsequent steps are carried out, so that the difference of each attribute is reduced, the quality of each numerical value is conveniently measured, and a normalized decision matrix R is constructed;
in the above formula, r ij Expressing the parameter value of the attribute j in the i network after the value normalization, wherein the normalization method comprises the following steps:
revenue type:
cost type:
in the above formula max (d) ij ) And min (d) ij ) The maximum number chosen among the attributes j of all candidate networks is called max (d) ij ) Likewise, the minimum number of attribute j picks out among all candidate networks may be relatively referred to as min (d) ij )；
After the normalization of the matrix is completed, all the attributes are sequentially selected from the candidate network matrix RThe best value and the worst value of the character are obtained, all the best values are listed as the best matrix R max All the worst values are listed as a worst matrix R min ；
For the decision factor, two types can be distinguished: revenue-type and cost-type. The difference between the two is that the expected value of the decision factor is different, and the expected value for the profit type is as good as the larger value is; while the expectation for cost-type is that the smaller the value the better, the normalization is also different for different kinds of decision factors, and the optimal value for profit-type is max (r) ij ) The worst value is min (r) ij ). The opposite is true for the cost model, and the optimal value is min (r) ij ) The worst value is max (r) ij ) In the design of the invention, except for the bandwidth, other attributes are cost type, and the bandwidth is profit type;
the optimal matrix is therefore represented as:
R max ＝[max(r 1j )min(r 2j )...min(r 8j )] (6)
the worst matrix for this is expressed as:
R min ＝[min(r 1j )max(r 2j )...max(r 8j )] (7)
each candidate network may be normalized by the best-worst matrix:
for the revenue type:
for the cost type:
finally obtaining a constructed normalized decision matrix A
The eigenvector omega corresponding to the maximum eigenvalue of the decision matrix A is the corresponding weight of each attribute, and as the eigenvector omega is an objective algorithm, the weight of the decision attribute is not fixed according to different network conditions, and the weight of the decision attribute needs to be recalculated at each moment;
after weighting the normalization matrix, final comparison can be carried out, and the final weighting normalization matrix V can be constructed by multiplying each network attribute with the weight matrix of the two algorithms m×n ；
v ij ＝w j ×r ij
Wherein w j That is, the normalized magnitude, r, of the weight value of the corresponding attribute j ij Then the normalized value of the weight attribute j in the ith network is shown, v ij Then representing a numerical value obtained after weighting normalization of the network attribute j in the ith candidate network;
finally, all v of the same network are combined ij The values are added.
V i The method includes the steps that weight values which reflect the advantages and disadvantages of the candidate network i finally are compared, the size of each network V is compared, due to the fact that normalization processing is conducted on the weights of the parameter sets in the prior art, the candidate network corresponding to the V with the highest weight is obtained through comparison and serves as a preselection network at the next moment, and if the difference value between the weight value corresponding to the network maintained by the vehicle at the current moment and the weight value of the preselection network is smaller than or equal to the difference value of the weight value corresponding to the network maintained by the vehicle at the current moment and the weight value of the preselection network
The vehicle can maintain connection with the vehicle network corresponding to the current moment, and does not select the optimal solution network at the current moment, because the candidate networks in the step 2) can be used by ensuring the normal network of the vehicle through the screening of the fuzzy control, and the vehicle can be subjected to network selection after the network is finally selected to complete the communication and communication work.
According to the invention, aiming at the requirement that information data transmission is required to be carried out through a communication network when a vehicle in the internet of vehicles cruises to run, on the premise of simultaneously considering the information of the vehicle network and the motion state of the vehicle, a distance measurement algorithm for determining the relative position of the vehicle and a base station based on signal intensity is designed, a network is screened based on a fuzzy control algorithm, and the relative weight of the network is determined based on an approximate ideal solution algorithm to jointly use three network selection algorithms, so that the vehicle can select the optimal communication network at the current moment through the network selection algorithm at each scanning moment to ensure the communication quality and the correct transmission of information, the vehicle in the internet of vehicles can select the optimal data network under the condition of self-adaptive cruise control, the data information is correctly sent and received, and the error probability of the information is reduced.
The embodiment is a strategy selection method for accessing a heterogeneous network of a vehicle networking, and the specific operation process is as follows:
1. in the parameter setting interface, model parameters are input as follows: time interval T S =0.1, and the preset network parameters are shown in table 1. 5 WIFI networks and 2G networks are arranged in the driving range.
TABLE 1
2. Clicking a 'debugging' button on a configuration interface to enter the debugging interface, starting a CPU (central processing unit) of a main control computer to call a 'controller calculation program' which is programmed in advance, and adjusting network parameters in the configuration interface, wherein the fuzzy algorithm rule is shown in table 2, and in an off-line mode, according to preset network parameters, according to the formula disclosed by the invention, a proper threshold value can be obtained.
TABLE 2
3. And (3) online operation: the method comprises the steps of starting a CPU (central processing unit) of a main control computer to read model parameters and controller parameters, obtaining the distance error, the relative speed and the acceleration of a vehicle between a vehicle and a vehicle at the current moment through a vehicle-mounted sensor, obtaining the acceleration of the vehicle at the current moment through wireless channel transmission, obtaining relevant information of network signals by a receiver of the vehicle at each moment, selecting the network through an algorithm, selecting the best network at the current moment to switch, ensuring that the relevant information of cruising running of the vehicle at the next moment can be smoothly transmitted, obtaining the relevant parameters of the vehicle through the vehicle-mounted sensor again in the next control period, obtaining the relevant information of the network signals through the receiver, and repeating the steps in this way, so that the vehicle can obtain the best network experience at each moment.
In the embodiment, the selected vehicle length L =2m, the initial speed of the pilot vehicle is 0,0-8 s, and the internal acceleration is 1m/s 2 The internal acceleration of 14-18 s is-1 m/s 2 The internal acceleration of 32-36 s is-0.5 m/s2, and the internal acceleration of 60-64 s is 1m/s 2 The internal acceleration of 80-84 s is-0.5 m/s 2 And then keeping constant-speed straight driving, wherein the minimum safe distance d0=8m. The actual control effect is shown in fig. 1 and 2. Fig. 1 shows weight values of various networks calculated by the optimal worst algorithm when all preset networks are taken into consideration during offline debugging. Fig. 2 shows the network selection of the pilot vehicle at each sampling instant during actual operation.
The foregoing illustrates the effect of the excellent heterogeneous network access selection policy method applied to a vehicle-following queue, which is shown in an embodiment of the present invention. It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that any modifications made within the spirit of the invention and the scope of the appended claims fall within the scope of the invention.
Claims (1)
1. A heterogeneous network access selection strategy method applied to a vehicle-following queue is characterized by comprising the following steps:
1) The vehicle queue performs linear following, the vehicle is provided with a sensor module to receive and transmit speed and acceleration information of an adjacent vehicle, firstly, a propagation loss model method is used, the relative distance between the vehicle and a base station is estimated according to the signal intensity detected at the current moment, then the subsequent work can be performed, and the formula of the signal intensity and the distance is shown in formula (1):
wherein R (d) is the signal strength measured at the current time, d is the distance between the vehicle and the base station at the current time, and R (d) 0 ) For preset and adjusted network signal strength, the unit is dB, d 0 Value and preset network signal intensity R (d) 0 ) Related to a boundary value with R (d) 0 ) Are arranged together, X σ Obey Gaussian distribution for random noise errors in the measurement process; n is a path loss exponent, associated with a particular environment;
when the signal intensity at the current time is known, the distance d between the vehicle and the base station is obtained by the formula (1), and then the distance d is compared with a preset good receiving range d 1 Comparing, thereby eliminating the part of network signals which are too far between the vehicle and the base station, and screening the network signals for the first time;
2) According to a fuzzy control algorithm, the selection priority of each network signal is solved;
the input values of the fuzzy control algorithm are respectively received network signal strength RSSI, vehicle moving speed v and vehicle acceleration a, in the fuzzy system, a membership function is the most important part for fuzzification of the input values, and different membership functions are selected, so that different fuzzy results can be obtained;
27 fuzzy rules are set for the three input variables, and the 27 fuzzy rules are used for handling each possible input condition in the fuzzy reasoning stage;
evaluating the moving speed v and the acceleration a of the terminal user and the received network signal strength RSSI through fuzzy logic to obtain an evaluation value H, wherein the obtained switching evaluation value H reflects the suitability of each network switching, and after the network with the too low H value is removed, a more accurate candidate network group is obtained again for the step 3) to use;
3) The eight decision factors are transmission time delay, transmission bandwidth, packet loss rate, jitter, price and power, vehicle speed and vehicle acceleration;
after the eight decision factors of the vehicle are obtained at each scanning moment, a decision matrix D with the size of m multiplied by n and containing the sizes of the decision factors under the current condition of the network is constructed m×n ：
In the above formula, d ij The value of j attribute of i network, m in m × n represents m candidate networks, n represents the number of decision factors used for evaluating weight this time, in the network selection algorithm this time, the influence of eight attributes is considered, so n =8;
constructing a normalized decision matrix R;
in the above formula, r ij Expressing the parameter value of the attribute j in the i network after the value normalization, wherein the normalization method comprises the following steps:
revenue type:
cost type:
in the above formula max (d) ij ) And min (d) ij ) The maximum number chosen among the attributes j of all candidate networks is called max (d) ij ) Likewise, the minimum number of attribute j picks out among all candidate networks may be relatively referred to as min (d) ij )；
After the normalization of the matrix is completed, the optimal value and the worst value of each attribute are sequentially selected from the normalization decision matrix R, and all the optimal values are listed as the optimal matrix R max All the worst values are listed as the worst matrix R min ；
For the decision factor, two types are distinguished: the profit type and the cost type have different normalization modes for different decision factors, and the optimal value of the profit type is max (r) ij ) The worst value is min (r) ij ) The opposite is true for the cost model, and the optimum value is min (r) ij ) The worst value is max (r) ij ) Besides the bandwidth, the other attributes are cost type, and the bandwidth is profit type;
the optimal matrix is therefore represented as:
R max ＝[max(r 1j ) min(r 2j ) ... min(r 8j )] (6)
the worst matrix against this is expressed as:
R min ＝[min(r 1j ) max(r 2j ) ... max(r 8j )] (7)
each candidate network may be normalized by the best-worst matrix:
for the profit type:
for the cost type:
finally obtaining a normalized matrix A
The eigenvector omega corresponding to the maximum eigenvalue of the normalization matrix A is the corresponding normalized weight of each attribute, and as the eigenvector omega is an objective algorithm, the weight of the decision attribute is not fixed according to different network conditions, and the weight of the decision attribute needs to be recalculated at each moment;
after weighting the normalization matrix A, carrying out final comparison to construct a final weighted normalization matrix V m×n ；
v ij ＝w j ×r ij
Wherein w j That is, the normalized magnitude, r, of the weight value of the corresponding attribute j ij The normalized value of the attribute j in the ith network is shown, v ij Then, a numerical value obtained after weighting normalization of the network attribute j in the ith candidate network is represented;
finally, all v of the same network are combined ij Adding the values;
V i the weight values of the candidate networks i are finally reflected, and the weight values V of the networks are compared, because of the weight valuesThe weights of all parameter sets are normalized, a candidate network corresponding to the V with the highest weight is obtained through comparison and serves as a preselected network at the next moment, and if the difference value between the weight value corresponding to the network maintained by the vehicle at the current moment and the weight value of the preselected network is smaller than that
The vehicle can maintain connection with the vehicle network corresponding to the current moment, and does not select the optimal solution network at the current moment, because the candidate networks in the step 2) can be used by ensuring the normal network of the vehicle through the screening of the fuzzy control, and the vehicle can be subjected to network selection after the network is finally selected to complete the communication and communication work.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911014970.6A CN110933638B (en) | 2019-10-24 | 2019-10-24 | Heterogeneous network access selection strategy method applied to vehicle following queue |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911014970.6A CN110933638B (en) | 2019-10-24 | 2019-10-24 | Heterogeneous network access selection strategy method applied to vehicle following queue |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110933638A CN110933638A (en) | 2020-03-27 |
| CN110933638B true CN110933638B (en) | 2023-01-31 |
Family
ID=69849334
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911014970.6A Active CN110933638B (en) | 2019-10-24 | 2019-10-24 | Heterogeneous network access selection strategy method applied to vehicle following queue |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110933638B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111861905B (en) * | 2020-06-17 | 2022-07-15 | 浙江工业大学 | SAR image speckle noise suppression method based on Gamma-Lee filtering |
| CN111885543B (en) * | 2020-06-28 | 2021-08-27 | 北京交通大学 | Internet of vehicles switching method and device based on ultra-dense networking |
| CN116088401B (en) * | 2023-04-10 | 2023-07-07 | 之江实验室 | Heterogeneous network-based automatic driving vehicle remote control device and method |
| CN117354755B (en) * | 2023-12-05 | 2024-02-02 | 鹰驾科技（深圳）有限公司 | Vehicle-mounted multimedia management cloud platform of wireless transmission technology |
| CN117395691B (en) * | 2023-12-11 | 2024-03-01 | 泉州市三川通讯技术股份有限责任公司 | Vehicle-mounted terminal communication optimization method under weak signal environment |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103139860A (en) * | 2013-02-04 | 2013-06-05 | 重庆邮电大学 | Uniting switch method based on speed self-adaption in isomerism cognition radio network |
| CN109151953A (en) * | 2018-09-25 | 2019-01-04 | 华北电力大学（保定） | A kind of network insertion selection calculation method based on user and network bilateral income |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9866610B2 (en) * | 2011-12-16 | 2018-01-09 | Genband Us Llc | Methods, systems, and computer readable media for selecting a codec pair based on network conditions |
-
2019
- 2019-10-24 CN CN201911014970.6A patent/CN110933638B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103139860A (en) * | 2013-02-04 | 2013-06-05 | 重庆邮电大学 | Uniting switch method based on speed self-adaption in isomerism cognition radio network |
| CN109151953A (en) * | 2018-09-25 | 2019-01-04 | 华北电力大学（保定） | A kind of network insertion selection calculation method based on user and network bilateral income |
Non-Patent Citations (4)
| Title |
|---|
| Distributed_Predictive_Control_for_String_Stability_of_Heterogonous_Vehicle_Platoons_with_Multiple_Communication_Topologies;LU Liangye, HE Defeng, YU Shiming, DONG Shijian;《Proceedings of the 38th Chinese Control Conference》;20190730;第2930-2935页 * |
| Energy-Efficient_Opportunistic_Packet_Scheduling_in_Mobile_Relay_Systems;Zhaoyang Zhang;《IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY》;20160730;第65卷(第7期);第5327-5336页 * |
| 基于直觉模糊决策的异构车联网网络选择方案研究;陈璇宇;《中国优秀硕士学位论文辑》;20200131;论文全文 * |
| 异构网络传输的车辆自适应巡航控制;宋秀兰;《工业控制计算机》;20181030;第31卷(第10期);第61-63页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110933638A (en) | 2020-03-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110933638B (en) | Heterogeneous network access selection strategy method applied to vehicle following queue | |
| CN110809306A (en) | Terminal access selection method based on deep reinforcement learning | |
| Zhang et al. | A neural network prediction-based adaptive mode selection scheme in full-duplex cognitive networks | |
| Shen et al. | Design and implementation for deep learning based adjustable beamforming training for millimeter wave communication systems | |
| Kunarak et al. | Algorithmic vertical handoff decision and merit network selection across heterogeneous wireless networks | |
| CN117581248A (en) | Determining target policies for managing an environment | |
| CN113423110B (en) | Multi-user multi-channel dynamic spectrum access method based on deep reinforcement learning | |
| CN110996365B (en) | Heterogeneous network vertical switching algorithm and system based on multi-objective optimization model | |
| Islam et al. | Improvement of the handover performance and channel allocation scheme using fuzzy logic, artificial neural network and neuro-fuzzy system to reduce call drop in cellular network | |
| CN115190489A (en) | Cognitive wireless network dynamic spectrum access method based on deep reinforcement learning | |
| Zhou et al. | Multi-agent few-shot meta reinforcement learning for trajectory design and channel selection in UAV-assisted networks | |
| Hazarika et al. | Multi-agent DRL-based computation offloading in multiple RIS-aided IoV networks | |
| CN111200855B (en) | Adaptive multi-criterion fuzzy logic 5G heterogeneous network switching decision method | |
| CN115811788B (en) | D2D network distributed resource allocation method combining deep reinforcement learning and unsupervised learning | |
| CN111130698A (en) | Wireless communication receiving window prediction method and device and wireless communication equipment | |
| CN113490248B (en) | Multi-mode terminal switching method and device | |
| CN114390625B (en) | Control method and device for base station switching, storage medium and electronic equipment | |
| CN113890653B (en) | Multi-agent reinforcement learning power distribution method for multi-user benefits | |
| CN114337881B (en) | Wireless spectrum intelligent sensing method based on multi-unmanned aerial vehicle distribution and LMS | |
| CN115297518A (en) | Network switching method and system based on mobile user position | |
| Guo et al. | AI-Aided channel quality assessment for Bluetooth adaptive frequency hopping | |
| Yu et al. | Fast automatic link establishment: A new metric and the value of spectrum prediction | |
| CN106937326A (en) | Method of coordinated transmission and first base station between base station | |
| CN113630830B (en) | Network switching method, device and equipment based on multi-attribute fuzzy neural network | |
| Xin et al. | Deep Learning Based MAC via Joint Channel Access and Rate Adaptation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |