CN114363996B - Heterogeneous wireless network service access control method and device based on multiple targets - Google Patents
Heterogeneous wireless network service access control method and device based on multiple targets Download PDFInfo
- Publication number
- CN114363996B CN114363996B CN202210060061.1A CN202210060061A CN114363996B CN 114363996 B CN114363996 B CN 114363996B CN 202210060061 A CN202210060061 A CN 202210060061A CN 114363996 B CN114363996 B CN 114363996B
- Authority
- CN
- China
- Prior art keywords
- particle
- particles
- formula
- network
- service
- 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
-
- 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
The invention provides a heterogeneous wireless network service access control method and a heterogeneous wireless network service access control device based on multiple targets, wherein the heterogeneous wireless network service access control method comprises the following steps: (1) acquiring parameters such as network quantity, service quantity and the like; (2) acquiring parameters such as transmission rate, channel bandwidth, signal power, noise power, the number of two-dimensional resource units and the like; (3) calculating the user information transmission rate; (4) calculating the occupied amount of the service resources; (5) calculating the network blocking rate; (6) determining a heterogeneous wireless network service access control objective function; (7) obtaining optimal particles by utilizing an improved gravity search algorithm, and updating an external archive set by the obtained optimal particles according to a domination relation; (8) performing population evolution on particles in an external filing set through a data field and a rotating base technology; (9) and when the end condition is met, outputting an optimal solution set, namely the optimal heterogeneous wireless network service access control scheme. The heterogeneous wireless network service access control method based on multiple targets can ensure that more user services are accessed to a better network, thereby providing higher-quality service experience for users.
Description
Technical Field
The invention relates to the technical field of heterogeneous wireless networks, in particular to a heterogeneous wireless network service access control method and device based on multiple targets.
Background
The wide application of the novel intelligent mobile device, the internet and the internet of things leads to the rapid increase of mobile traffic, and the trend can be continued for a long time in the future. Today, global networking devices will reach 285 billions. The unprecedented upgrade of the mobile equipment puts higher requirements on the existing network; meanwhile, user services are diversified, and different services to be accessed to a wireless network environment have different requirements on network performance, for example, in a voice session service, the service is very sensitive to network delay but requires less network bandwidth, online live video has low requirements on network delay but needs to occupy more bandwidth, users have different preferences, and some users have higher requirements on network bandwidth and lower network price, and emerging services such as virtual reality, unmanned driving, telemedicine, office, industrial automation and the like all put forward higher requirements on network performance.
The diversified development of application scenes leads to the generation of novel services, and the problem of coexistence of multiple services is increasingly prominent. For this reason, the wireless communication networks are gradually expanding in scale, and the wireless network capabilities are also continuously increasing. The mobile cellular network is developed from a mobile global communication system at the beginning to a universal mobile communication system, and then to 4G and 5G, so that a network guarantee of high-quality application is provided for users; meanwhile, on this basis, wireless networks such as a wireless local area network, a mobile Ad hoc network, an internet of vehicles, an internet of things and the like are deployed in large quantities. The establishment of a series of wireless standards, while providing high-speed wireless connectivity for users, does not allow a single network to simultaneously satisfy the diverse needs of different services for all users.
In order to solve the problems caused by the increase of mobile traffic and user services, the wireless networks need to be integrated to form a heterogeneous wireless network. In a heterogeneous wireless network environment, how to better utilize network resources on the premise of ensuring user service quality is an important problem to be solved at present in the heterogeneous wireless network, wherein service access control is a core task of fully and reasonably utilizing the network resources. Most of traditional heterogeneous wireless network service access control only optimizes a single network performance index, and when a certain performance index is improved, other performance indexes are possibly reduced by the control method. Therefore, the service access control method considering a single performance index lacks comprehensive consideration of the overall performance of the whole heterogeneous wireless network, which often fails to simultaneously meet diversified service requirements of different users and different services, and reduces network service quality.
Disclosure of Invention
Therefore, in order to comprehensively consider the diversified requirements of the user side and the different network states of the network side, the invention provides a heterogeneous wireless network service access control method and a heterogeneous wireless network service access control device based on multiple targets, and the technical scheme comprises the following steps:
step 1000: when a user accesses a network, acquiring parameters such as the number of networks and the number of user services existing in the network;
step 2000: according to the state of the heterogeneous wireless network, acquiring parameters such as the rate, channel bandwidth, signal power, noise power, two-dimensional resource unit number and the like of real-time service and non-real-time service in the heterogeneous wireless network environment;
step 3000: and (2) calculating the user information transmission rate r (x), wherein the calculation method is shown in a formula (1):
wherein, b ij Representing the channel bandwidth, S, to which service i is allocated in network j ij Representing the signal power of service i in network j, N ij Representing the noise power of the service i in the network j, and m represents the total number of the network;
step 4000: calculating the occupied quantity f (x) of the business resources, wherein the calculation method is shown in a formula (2):
wherein x is ij Denotes the case of access of service i into network j, t ij Representing the number of two-dimensional resource units occupied by the service i accessed to the network j, wherein n is the total number of the service quantity, and m is the total number of the network quantity;
step 5000: and (3) calculating the network blocking rate g (x), wherein the calculation method is shown in formula (3):
wherein x is ij The method comprises the steps that the condition that a service i is accessed into a network j is shown, n represents the total number of the service quantity, and m represents the total number of the network quantity;
step 6000: determining an objective function for controlling the service access of the heterogeneous wireless network, wherein the objective function is shown in a formula (4):
where maxr (x) is to maximize the information transmission rate, b ij Representing the channel bandwidth, S, to which service i is allocated in network j ij Representing the signal power of service i in network j, N ij Representing the noise power of the service i in the network j, the value range of j is [1, m]I has a value range of [1, n](ii) a minf (x) to minimize traffic resource occupancy, x ij Denotes the case of access of service i into network j, t ij The number of two-dimensional resource units occupied by the service i accessed to the network j is represented, n represents the total number of the service quantity, and m represents the total number of the network quantity; ming (x) is to minimize network blocking rate;
step 7000: obtaining optimal particles by utilizing an improved gravity search algorithm, and updating an external archive set according to a domination relation, further comprising the following steps:
step 7010: randomly generating an initial population, setting the iteration times of the population, the maximum value of an external filing set and a universal gravitation constant G 0 Initializing the position x and the speed v of the particle;
step 7020: calculating the fitness value of each particle according to the objective function of the heterogeneous wireless network service access control in the step 6000, which is shown in a formula (4);
step 7030: calculating a weight vector of the particle: sampling M dimensional space, H is the sampling number of each target direction, and obtaining the sampling number in the optimizing space
A uniformly distributed weight vector, which specifies a sampling step size that can be expressed as δ 1/H, is output as a uniformly distributed set a of vectors, which is {0,1/H,2/H 1 +λ 2 +...+λ m 1, where λ 1 λ 2 ...λ m Is a weight vector;
step 7040: aggregating the targets by adopting a formula (5) according to the weight vector of each particle;
ming te (x|λ,z * )＝max{|λ 1 (f(x)-z 1 * )|,|λ 2 (g(x)-z 2 * )|,|λ 3 (r(x)-z 3 * )|} (5)
wherein, r (x) is the information transmission rate; (x) is the occupation amount of the service resources; g (x) is the network blocking rate; λ ═ λ 1 ,...,λ m ) As a weight, λ m The value range is [0,1 ]]；
The method is characterized in that the method is a reference point formed by optimal values of all objective functions, the advantages and disadvantages of individual fitness values of a population are judged by calculating the distance between the individual optimal value and the reference point in the evolution process, and the fitness value is better when the distance is smaller;
step 7050: assuming M of the particle ai (t)、M bi (t)、M ii (t) and M i And (t) values are equal, and the inertial mass of the particles is calculated by the calculation method shown in formula (6):
wherein M is ai (t) is the active gravitational mass of the stressed individual i at time t, M bi (t) the passive gravitational mass of the forcing entity i at time t, M ii (t) represents the inertial mass of the particle i at time t, M i (t) is the mass of particle i at a certain iteration at time t, m i (t) number of i-th particle at time tMass, m j (t) is the individual mass of the jth particle, and the value range of j is [1, N ]]N is the presence of N particles in the search space, fit i (t) represents the fitness value of the particle, worst (t) represents the fitness value of the particle with the largest quality, best (t) represents the fitness value of the particle with the smallest quality, and the definition of worst (t) and best (t) is shown in a formula (7):
wherein, the value range of i is [1, N ], N is N particles existing in the search space, and fit (t) represents the fitness value of the particles at the time t;
step 7060: calculating the acting force of the particles j to the particles i in the population in the t iteration
The calculation formula is shown in formula (8):
wherein the content of the first and second substances,
the force of particle j on particle i in the population at time t,
representing the position of the particle i in the D-dimensional space at time t,
representing the position information of the particle j in the D-dimensional space, R ij (t) represents the Euclidean distance between the particles i and j at the time t, epsilon is the minimum value, the prevention denominator is 0, M aj (t) represents the active gravity mass of the force individual j at time t, M bi (t) represents the passive gravity mass of the force application individual i at the time t, G (t) is the gravity constant at the time t, and the calculation formula is shown in formula (9):
G(t)＝G 0 ×e -αt/T (9)
wherein, G 0 The initial value of the constant coefficient is constantly set as 100, alpha is a descending coefficient and is constantly set as 20, T is the current iteration number, and T is the total iteration number;
step 7070: calculating the acceleration of the particle, the acceleration of the particle i at time t
The calculation formula is shown in formula (10):
wherein, the first and the second end of the pipe are connected with each other,
acceleration of particle i in d-dimensional space at time t, M ii (t) represents the inertial mass of the particle i at time t,
the calculation formula is shown in formula (11) for the acting force of the particle i in the d-dimensional space at the time t:
wherein, rand j The value range is [0,1 ]]，
Representing the acting force of the particles j on the particles i in the population at the moment t, wherein D is the dimension, and N is N particles in a D-dimension search space;
step 7080: memorizing the self optimal information and the population optimal information of the particles by adopting asynchronous learning factors of formulas (12) and (13);
c 1 ＝c 1_ini +(c 1_fin -c 1_ini )*t/T (12)
c 2 ＝c 2_ini +(c 2_fin -c 2_ini )*t/T (13)
wherein, c 1_ini 、c 2_ini Representing the initial learning ability, c 1_fin 、c 2_fin Representing the learning ability when the iteration is finished, wherein T represents the current iteration times, and T represents the maximum iteration times;
step 7090: mapping the sine value of the particle speed to the probability value of the change of the particle position vector, and calculating a formula (14);
where v is the velocity value of the particle and f (v) is the probability value that maps the sine of the particle velocity to a change in the particle position vector;
step 7100: changing the search path of the particle using equation (15);
wherein Levy (epsilon) is a Levy flight search path, u obeys a normal distribution curve, the value range of beta is (0,2), and u-N (0, sigma) 2 ) V to N (0,1), σ is defined as follows:
wherein the value range of beta is (0,2), gamma function is represented by gamma, and the Levy (epsilon) of the Levy flight can be determined through the formulas (16) and (15);
step 7110: updating the particles according to the non-dominated sorting of formula (17) according to the fitness value of the optimal particles obtained from step 7020 to step 7100:
wherein, for minimization problems, X 1 、X 2 There are two solutions before non-dominant ordering, solution X 1 The values on all targets are no worse than X 2 And is in X 1 Above, at least one objective function value is better than X 2 Is called X 1 Dominating X 2 ，X 1 For non-dominant solutions, X 2 To the dominant solution;
step 7120: updating the external archive set according to the obtained optimal particles according to the domination relationship, and executing the step 8000;
step 8000: the method adopts a data field and a rotating base technology to guide population evolution on the external archive obtained in the step 7120, and further comprises the following steps:
step 8100: judging the number of non-dominated solutions in the external archive set and the maximum value of the non-dominated solutions, if the number of the non-dominated solutions in the external archive set exceeds the maximum value, performing 8200 to adjust and delete the external archive set, otherwise, performing 8300 to obtain class 1 guide particles;
step 8200: mapping a non-dominant solution set in an external archive set into a gravitational data field, introducing a calculation method of potential energy in the data field, wherein the calculation formula is shown in formula (18), calculating the potential energy of each particle in the external archive set, deleting the external archive set, gradually deleting the particles with high potential energy until the number of the non-dominant solution sets is the maximum value of the external archive set, and the calculation formula of the potential energy of the data of the particles i is as follows:
wherein: psi (x) i ) Is the data potential energy of the ith particle, ρ j Representing the property value of the particle j in the data field, σ being the influence factor of the particle in the data field, x i Is the position information of the ith particle, R ij Representing the Euclidean distance between the particle i and the particle j, wherein N is N particles in the search space;
step 8300: introducing a data field potential energy formula (18) of the particles, calculating the potential energy of each particle in an external filing set, reflecting the distribution form of population particles, sequencing the potential energy of each particle, selecting guide particles according to the potential energy, wherein the particles selected by the data field potential energy principle are named as class 1 guide particles and are named as GF;
step 8400: selecting the particles with good distribution in the evolutionary population as class 2 guide particles, and naming the particles as GE, further comprising the following steps:
step 8410: carrying out non-dominant sorting on the particles in the external archive set according to a formula (17), and placing the particles in different layers according to different sorting of the particles;
step 8420: using the distribution degree index F i See formula (19), calculating the distribution degree of the solution set in the target space, wherein the smaller the value of the distribution degree, the more uniformly the solution set is distributed in the target space, and the distribution degree index F i Is defined as:
wherein K represents the number of solutions in the Pareto solution set, and
denotes the average distance between them, d i Distance of the target vector expressed as the i-th non-inferior solution to the optimal solution target vector, d i Is the formula (20):
wherein the content of the first and second substances,
the target vector representing the ith non-inferior solution,
the j and i are unequal, and the value range of K is [1, K ]]K represents the number of solutions in the Pareto solution set;
step 8430: adopting a rotation base technology, forming a rotation angle on a connecting line of a terminal point of each iteration and a drawing base point, accumulating and drawing the rotation angle to form a mapping broken line, leading the performance of the optimal solution on each target to be more balanced, leading the mapping broken line to be closer to a straight line finally, sequencing the particles in an external filing set by utilizing the relation between the mapping broken line and the performance of the optimal solution, adopting a rotation base mapping broken line distance L to carry out auxiliary judgment if the distribution indexes of the particles are the same, leading the particles to evolve if the selected particle quality is as large as possible, and leading the particles distributed as uniformly as possible by L definition according to a formula (21):
wherein, theta i 、θ j Respectively representing the angle formed by the ith rotation and the jth rotation, wherein r is the maximum rotation time, and L is the rotation base mapping broken line distance;
step 8440: finally, selecting particles with the mass as large as possible and the distribution as uniform as possible as class 2 guide particles, and naming the particles as GE;
step 8500: s with larger distance selected according to step 8400 GE The 2 nd type guide particle performs guide evolution on the population particle in the gravity search algorithm evolution mode improved in step 7000, and the specific guide mode of the 2 nd type guide particle is as follows:
wherein, the first and the second end of the pipe are connected with each other,
class 2 guide particles for time t +1,
denotes the mass of the jth type 2 guide particle at time t, R ij (t) represents the Euclidean distance between the particle i and the particle j at the time tDistance, x i (t) and x j (t) position information of the particle i and the particle j at time t, G is an attractive force constant, S GE Directing the number of particles for class 2;
step 8600: the population particles move under the common guidance of GF and GE guiding particles, and the updating mode of the speed and the position information of the particles is shown as the formula (23):
rand i is in the value range of [0,1 ]]，
Representing the velocity of the particle i at time t,
representing the velocity of particle i at time t +1,
representing the position information of particle i at time t +1,
class 1 guide particles for time t +1,
class 2 guide particles at time t + 1;
step 8700: if the fitness value is good enough or the maximum iteration number is reached, the loop is terminated, the step 9000 is executed, otherwise, the step 7020 is executed continuously;
step 9000: and outputting an optimal solution set, namely the obtained optimal heterogeneous wireless network service access control scheme.
An apparatus of a multi-target-based heterogeneous wireless network service access control method, the apparatus comprising:
a data acquisition module: is used for collecting the parameters in the steps 1000 and 2000;
an objective function determination module: determining the multi-objective optimization function model in the step 6000 according to the user information transmission rate, the service resource occupation and the network blocking rate in the steps 3000, 4000 and 5000;
a model solving module: and solving the multi-target optimization function model in the step 6000 by using the solving algorithm in the steps 7000 and 8000 to finally obtain the optimal heterogeneous wireless network service access scheme.
Compared with the prior art, the invention has the beneficial effects that:
(1) the sine mapping jump improved gravitation search algorithm based on asynchronous learning is adopted to replace an updating strategy and an individual reservation mode of the multi-objective optimization algorithm, and the convergence speed and the convergence precision of the multi-objective optimization algorithm are improved;
(2) the data field and the rotation base technology are adopted to guide evolution of the reserved optimal particles, the solution distribution and convergence of the multi-target optimization algorithm in solving the multi-target heterogeneous wireless network service access problem are enhanced, a new artificial intelligence-based method is provided for the multi-target heterogeneous wireless network service access problem, the network access decision speed is improved, waste of network resources is avoided, more user services can be guaranteed to be accessed to the network, and high-quality service experience can be provided.
Drawings
FIG. 1 is a flow chart of a heterogeneous wireless network service access control method based on multiple targets;
Detailed Description
In order that the above aspects of the present invention may be more clearly understood, the present invention will now be described in further detail with reference to the accompanying drawings. It should be noted that the specific implementation described herein is only for explaining the present application and is not used to limit the present application.
Fig. 1 is a flow chart of a heterogeneous wireless network service access control method based on multiple targets, which specifically includes the following steps:
step 1000: when a user accesses a network, acquiring parameters such as the number of networks and the number of user services existing in the network;
step 2000: acquiring parameters such as the rate, channel bandwidth, signal power, noise power, two-dimensional resource unit number and the like of real-time services and non-real-time services in a heterogeneous wireless network environment according to the state of the heterogeneous wireless network;
step 3000: calculating the user information transmission rate r (x), wherein the calculation method is shown in formula (1):
wherein, b ij Representing the channel bandwidth, S, to which service i is allocated in network j ij Representing the signal power of service i in network j, N ij Representing the noise power of the service i in the network j, and m represents the total number of the network;
step 4000: calculating the occupied quantity f (x) of the business resources, wherein the calculation method is shown in a formula (2):
wherein x is ij Denotes the case of access of service i into network j, t ij Representing the number of two-dimensional resource units occupied by the service i when the service i is accessed to the network j, wherein n is the total number of the service quantity, and m is the total number of the network quantity;
step 5000: and (3) calculating the network blocking rate g (x), wherein the calculation method is shown in formula (3):
wherein x is ij The method comprises the steps that the condition that a service i is accessed into a network j is shown, n represents the total number of the service quantity, and m represents the total number of the network quantity;
step 6000: determining an objective function for controlling the service access of the heterogeneous wireless network, wherein the objective function is shown in a formula (4):
where maxr (x) is to maximize the information transmission rate, b ij Representing the channel bandwidth, S, to which service i is allocated in network j ij Representing the signal power of service i in network j, N ij Represents the noise power of the service i in the network j, and the value range of the j is [1, m ]]I has a value range of [1, n](ii) a minf (x) to minimize traffic resource occupancy, x ij Denotes the case of access of service i into network j, t ij The number of two-dimensional resource units occupied by the service i accessed to the network j is represented, n represents the total number of the service quantity, and m represents the total number of the network quantity; ming (x) is to minimize network blocking rate;
step 7000: obtaining the optimal particles by utilizing an improved gravity search algorithm, and updating an external archive set according to a domination relation, wherein the method specifically comprises the following steps:
step 7010: randomly generating an initial population, setting the iteration times of the population, the maximum value of an external filing set and a universal gravitation constant G 0 Initializing the position x and the speed v of the particle;
step 7020: calculating the fitness value of each particle according to the objective function of the heterogeneous wireless network service access control in the step 6000, which is shown in a formula (4);
step 7030: calculating a weight vector of the particle: sampling M dimensional space, H is the sampling number of each target direction, and obtaining the sampling number in the optimizing space
A uniformly distributed weight vector, which defines a sampling step size which can be expressed as δ 1/H, and a uniformly distributed set a of vectors, which is {0,1/H, 2/H., 1}, is output, m numbers being selected from the set a such that the sum of the m numbers is 1, i.e. λ 1 +λ 2 +...+λ m 1, where λ 1 λ 2 ...λ m Is a weight vector;
step 7040: aggregating the targets by adopting a formula (5) according to the weight vector of each particle;
ming te (x|λ,z * )＝max{|λ 1 (f(x)-z 1 * )|,|λ 2 (g(x)-z 2 * )|,|λ 3 (r(x)-z 3 * )|} (5)
wherein, r (x) is the information transmission rate; (x) is the occupation amount of the service resources; g (x) is the network blocking rate; λ ═ λ 1 ,...,λ m ) As a weight, λ m The value range is [0,1 ]]；
The method is characterized by comprising the following steps that (1) the reference point is formed by optimal values of all objective functions, the advantages and disadvantages of individual fitness values of a population are judged by calculating the distance between the optimal value of each individual and the reference point in the evolution process, and the fitness value is better when the distance is smaller;
step 7050: assuming M of the particle ai (t)、M bi (t)、M ii (t) and M i And (t) values are equal, and the inertial mass of the particles is calculated by the calculation method shown in formula (6):
wherein M is ai (t) is the active gravitational mass of the stressed individual i at time t, M bi (t) the passive gravitational mass of the forcing entity i at time t, M ii (t) represents the inertial mass of the particle i at time t, M i (t) is the mass of particle i at a certain iteration at time t, m i (t) is the individual mass of the ith particle at time t, m j (t) is the individual mass of the jth particle, and the value range of j is [1, N]N is the presence of N particles in the search space, fit i (t) represents the fitness value of the particle, worst (t) represents the fitness value of the particle with the largest quality, best (t) represents the fitness value of the particle with the smallest quality, and the definition of worst (t) and best (t) is shown in a formula (7):
wherein, the value range of i is [1, N ], N is N particles existing in the search space, and fit (t) represents the fitness value of the particles at the time t;
step 7060: calculating the acting force of the particles j to the particles i in the population in the t iteration
The calculation formula is shown in formula (8):
wherein the content of the first and second substances,
the force of particle j on particle i in the population at time t,
indicating the position of the particle i in the D-dimensional space at time t,
indicating the position information of the particle j in the D-space, R ij (t) represents the Euclidean distance between the particles i and j at the time t, epsilon is the minimum value, the prevention denominator is 0, M aj (t) represents the active gravity mass of the force individual j at time t, M bi (t) represents the passive gravity mass of the force application individual i at the time t, G (t) is a gravity constant at the time t, and the calculation formula is shown in a formula (9):
G(t)＝G 0 ×e -αt/T (9)
wherein G is 0 The initial value of the constant coefficient is constantly set as 100, alpha is a descending coefficient and is constantly set as 20, T is the current iteration number, and T is the total iteration number;
step 7070: calculating the acceleration of the particle, the acceleration of the particle i at time t
The calculation formula is shown in formula (10):
wherein the content of the first and second substances,
acceleration of particle i in d-dimensional space at time t, M ii (t) represents the inertial mass of the particle i at time t,
the calculation formula is shown in formula (11) for the acting force of the particle i in the d-dimensional space at the time t:
wherein, rand j The value range is [0,1 ]]，
Representing the acting force of the particles j on the particles i in the population at the moment t, wherein D is the dimension, and N is N particles in a D-dimension search space;
step 7080: memorizing the self optimal information and the population optimal information of the particles by adopting asynchronous learning factors of formulas (12) and (13);
c 1 ＝c 1_ini +(c 1_fin -c 1_ini )*t/T (12)
c 2 ＝c 2_ini +(c 2_fin -c 2_ini )*t/T (13)
wherein, c 1_ini 、c 2_ini Representing the initial learning ability, preferably, c 1_ini Has a value of 1, c 2_ini Has a value of 0.01, c 1_fin 、c 2_fin Representing the learning ability at the end of the iteration, preferably, c 1_fin Has a value of 0.01, c 2_fin Is 1, T represents the current iteration number, and T represents the maximum iteration number;
step 7090: mapping the sine value of the particle speed to the probability value of the change of the particle position vector, and calculating the formula shown in formula (14);
where v is the velocity value of the particle and f (v) is the probability value that maps the sine of the particle velocity to a change in the particle position vector;
step 7100: changing the search path of the particle using equation (15);
wherein Levy (epsilon) is a Levy flight search path, u obeys a normal distribution curve, the value range of beta is (0,2), and u-N (0, sigma) 2 ) V to N (0,1), σ is defined as follows:
wherein the value range of beta is (0,2), gamma function is represented by gamma, and the Levy (epsilon) of the Levy flight can be determined through the formulas (15) and (16);
step 7110: updating the particles according to the non-dominated sorting of formula (17) according to the fitness value of the optimal particles obtained from step 7020 to step 7100:
wherein, for minimization problems, X 1 、X 2 There are two solutions before non-dominant ordering, solution X 1 The values on all targets are no worse than X 2 And is in X 1 Above, at least one of the objective function values is better than X 2 Is called X 1 Dominating X 2 ，X 1 For non-dominant solutions, X 2 To the dominant solution;
step 7120: updating the external archive set according to the obtained optimal particles according to the domination relationship, and executing the step 8000;
step 8000: the method adopts a data field and a rotating base technology to guide population evolution on the external archive obtained in the step 7120, and further comprises the following steps:
step 8100: judging the number of non-dominated solutions in the external archive set and the maximum value of the non-dominated solutions owned by the external archive set, if the number of the non-dominated solutions in the external archive set exceeds the maximum value of the non-dominated solutions, performing 8200 to adjust and delete the external archive set, and if not, performing 8300 to obtain class 1 guide particles;
step 8200: mapping a non-dominant solution set in an external archive set into a gravitational data field, introducing a calculation method of potential energy in the data field, wherein the calculation formula is shown in formula (18), calculating the potential energy of each particle in the external archive set, deleting the external archive set, gradually deleting the particles with high potential energy until the number of the non-dominant solution sets is the maximum value of the external archive set, and the calculation formula of the potential energy of the data of the particles i is as follows:
wherein: psi (x) i ) Is the data potential energy of the ith particle, ρ j Representing the property value of the particle j in the data field, σ being the influence factor of the particle in the data field, x i Is the position information of the ith particle, R ij Representing the Euclidean distance between the particle i and the particle j, wherein N is N particles existing in the search space;
step 8300: introducing a data field potential energy formula (18) of the particles, calculating the potential energy of each particle in an external filing set, reflecting the distribution form of population particles, sequencing the potential energy of each particle, selecting guide particles according to the potential energy, wherein the particles selected by the data field potential energy principle are named as class 1 guide particles and are named as GF;
step 8400: selecting the particles with good distribution in the evolutionary population as class 2 guide particles, and naming the particles as GE, further comprising the following steps:
step 8410: carrying out non-dominated sorting on the particles in the external archive set according to a formula (17), and placing the particles in different layers according to different sorting of the particles;
step 8420: using the distribution degree index F i See equation (19), the degree of distribution of the solution set in the target space is calculated, and the smaller the value of the degree of distribution, the more uniformly the solution set is distributed in the target space, and the distribution degree index F i Is defined as:
wherein K represents the number of solutions in the Pareto solution set, and
denotes the average distance between them, d i Distance of the target vector expressed as the ith non-inferior solution to the target vector of the optimal solution, d i Is the formula (20):
wherein the content of the first and second substances,
the target vector representing the ith non-inferior solution,
the j and i are unequal, and the value range of K is [1, K ]]K represents the number of solutions in the Pareto solution set;
step 8430: adopting a rotation base technology, forming a rotation angle on a connecting line of a terminal point of each iteration and a drawing base point, accumulating and drawing the rotation angle to form a mapping broken line, leading the performance of the optimal solution on each target to be more balanced, leading the mapping broken line to be closer to a straight line finally, sequencing the particles in an external filing set by utilizing the relation between the mapping broken line and the performance of the optimal solution, adopting a rotation base mapping broken line distance L to carry out auxiliary judgment if the distribution indexes of the particles are the same, leading the particles to evolve if the selected particle quality is as large as possible, and leading the particles distributed as uniformly as possible by L definition according to a formula (21):
wherein, theta i 、θ j Respectively representing the angle formed by the ith rotation and the jth rotation, wherein r is the maximum rotation time, and L is the rotation base mapping broken line distance;
step 8440: finally, selecting particles with the mass as large as possible and the distribution as uniform as possible as class 2 guide particles, and naming the particles as GE;
step 8500: s with larger distance selected according to step 8400 GE The 2 nd type guide particle guides the evolution of the population particles in the improved gravity search algorithm evolution mode of the step 7000, and the specific guide mode of the 2 nd type guide particle is as follows:
wherein the content of the first and second substances,
class 2 guide particles for time t +1,
denotes the mass of the jth type 2 guide particle at time t, R ij (t) denotes the Euclidean distance, x, between particle i and particle j at time t i (t) and x j (t) position information of the particle i and the particle j at time t, G is an attractive force constant, S GE Directing the number of particles for class 2;
step 8600: the population particles move under the common guidance of GF and GE guiding particles, and the updating mode of the speed and the position information of the particles is shown as the formula (23):
rand i is in the value range of [0,1 ]]，
Representing the velocity of the particle i at time t,
representing the velocity of particle i at time t +1,
representing the position information of particle i at time t +1,
class 1 guide particles for time t +1,
class 2 guide particles at time t + 1;
step 8700: if the fitness value is good enough or the maximum iteration number is reached, the loop is terminated, the step 9000 is executed, otherwise, the step 7020 is executed continuously;
step 9000: and outputting an optimal solution set, namely obtaining an optimal heterogeneous wireless network service access control scheme.
An apparatus of a multi-target-based heterogeneous wireless network service access control method, the apparatus comprising:
a data acquisition module: is used for collecting the parameters in the steps 1000 and 2000;
an objective function determination module: determining the multi-objective optimization function model in the step 6000 according to the user information transmission rate, the service resource occupation and the network blocking rate in the steps 3000, 4000 and 5000;
a model solving module: and solving the multi-target optimization function model in the step 6000 by using the solving algorithm in the steps 7000 and 8000 to finally obtain the optimal heterogeneous wireless network service access scheme.
The above description is only an example of the present invention and is not intended to limit the scope of the present invention, and it will be apparent to those skilled in the art that various modifications and variations can be made in the present invention. Any modification, equivalent replacement, or improvement made without departing from the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (2)
1. A heterogeneous wireless network service access control method based on multiple targets is characterized by comprising the following steps:
step 1000: when a user accesses a network, acquiring the network quantity and user service quantity parameters existing in the network;
step 2000: acquiring the rate, channel bandwidth, signal power, noise power and two-dimensional resource unit number parameters of real-time service and non-real-time service in the heterogeneous wireless network environment according to the state of the heterogeneous wireless network;
step 3000: calculating the user information transmission rate r (x), wherein the calculation method is shown in formula (1):
wherein, b ij Representing the channel bandwidth, S, to which service i is allocated in network j ij Representing the signal power of service i in network j, N ij Representing the noise power of the service i in the network j, and m represents the total number of the network;
step 4000: calculating the occupied quantity f (x) of the business resources, wherein the calculation method is shown in a formula (2):
wherein x is ij Denotes the case of access of service i into network j, t ij Representing the number of two-dimensional resource units occupied by the service i accessed to the network j, wherein n is the total number of the service quantity, and m is the total number of the network quantity;
step 5000: and (3) calculating the network blocking rate g (x), wherein the calculation method is shown in formula (3):
wherein x is ij The method comprises the steps that the condition that a service i is accessed into a network j is shown, n represents the total number of the service quantity, and m represents the total number of the network quantity;
step 6000: determining an objective function for controlling the service access of the heterogeneous wireless network, wherein the objective function is shown in a formula (4):
wherein maxr (x) is to maximize the information transmission rate, b ij Representing the channel bandwidth, S, to which service i is allocated in network j ij Representing the signal power of service i in network j, N ij Represents the noise power of the service i in the network j, and the value range of the j is [1, m ]]I has a value range of [1, n](ii) a minf (x) to minimize traffic resource occupancy, x ij Denotes the case of access of service i into network j, t ij The number of two-dimensional resource units occupied by the service i when the service i is accessed to the network j is represented, n represents the total number of the service quantity, and m represents the total number of the network quantity; ming (x) is to minimize network blocking rate;
step 7000: obtaining the optimal particles by utilizing an improved gravity search algorithm, and updating an external archive set according to the dominance relationship, wherein the method further comprises the following steps of:
step 7010: randomly generating an initial population, setting the iteration times of the population, the maximum value of an external filing set and a universal gravitation constant G 0 Initializing the position x and the speed v of the particle;
step 7020: calculating the fitness value of each particle according to the objective function of the heterogeneous wireless network service access control in the step 6000, which is shown in a formula (4);
step 7030: calculating a weight vector of the particle: for M dimensional spaceLine sampling, H being the number of samples in each target direction, obtained in the optimization space
A uniformly distributed weight vector, which specifies a sampling step size of δ 1/H, and outputs a uniformly distributed set a of vectors, the set a being {0,1/H,2/H 1 +λ 2 +...+λ m 1, where λ 1 λ 2 ...λ m Is a weight vector;
step 7040: aggregating the targets by adopting a formula (5) according to the weight vector of each particle;
ming te (x|λ,z * )＝max{|λ 1 (f(x)-z 1 * )|,|λ 2 (g(x)-z 2 * )|,|λ 3 (r(x)-z 3 * )|} (5)
wherein, r (x) is the information transmission rate; (x) is the occupation amount of the service resources; g (x) is the network blocking rate; λ ═ λ 1 ,...,λ m ) As a weight, λ m The value range is [0,1 ]]；
The method is characterized by comprising the following steps that (1) the reference point is formed by optimal values of all objective functions, the advantages and disadvantages of individual fitness values of a population are judged by calculating the distance between the optimal value of each individual and the reference point in the evolution process, and the fitness value is better when the distance is smaller;
step 7050: assuming M of the particle ai (t)、M bi (t)、M ii (t) and M i And (t) values are equal, and the inertial mass of the particles is calculated by the calculation method shown in formula (6):
wherein M is ai (t) is the active gravitational mass of the stressed individual i at time t, M bi (t) the passive gravitational mass of the forcing entity i at time t,M ii (t) represents the inertial mass of the particle i at time t, M i (t) is the mass of particle i at a certain iteration at time t, m i (t) is the individual mass of the ith particle at time t, m j (t) is the individual mass of the jth particle, and the value range of j is [1, N]N is the presence of N particles in the search space, fit i (t) represents the fitness value of the particles, worst (t) represents the fitness value of the particles with the largest quality, best (t) represents the fitness value of the particles with the smallest quality, and the definition of worst (t) and best (t) is shown in a formula (7):
wherein, the value range of i is [1, N ], N is N particles existing in the search space, and fit (t) represents the fitness value of the particles at the time t;
step 7060: calculating the acting force of the particles j to the particles i in the population in the t iteration
The calculation formula is shown in formula (8):
wherein, the first and the second end of the pipe are connected with each other,
the force of particle j on particle i in the population at time t,
indicating the position of the particle i in the D-dimensional space at time t,
representing the position information of the particle j in the D-dimensional space, R ij (t) represents the Euclidean distance between the particles i and j at the time t, and epsilon is the minimum value to prevent the divisionThe mother is 0, M aj (t) represents the active gravitational mass of the stressed individual j at time t, M bi (t) represents the passive gravity mass of the force application individual i at the time t, G (t) is the gravity constant at the time t, and the calculation formula is shown in formula (9):
G(t)＝G 0 ×e -αt/T (9)
wherein, G 0 The initial value of the constant coefficient is constantly set as 100, alpha is a reduction coefficient and is constantly set as 20, T is the current iteration number, and T is the total iteration number;
step 7070: calculating the acceleration of the particle, the acceleration of the particle i at time t
The calculation formula is shown in formula (10):
wherein the content of the first and second substances,
acceleration of particle i in d-dimensional space at time t, M ii (t) represents the inertial mass of the particle i at time t, F i d (t) is the acting force of the particle i in d-dimensional space at the moment t, and the calculation formula is shown in formula (11):
wherein, rand j The value range is [0,1 ]]，
Representing the acting force of the particles j on the particles i in the population at the moment t, wherein D is the dimension, and N is N particles in a D-dimension search space;
step 7080: memorizing the self optimal information and the population optimal information of the particles by adopting asynchronous learning factors of formulas (12) and (13);
c 1 ＝c 1_ini +(c 1_fin -c 1_ini )*t/T (12)
c 2 ＝c 2_ini +(c 2_fin -c 2_ini )*t/T (13)
wherein, c 1_ini 、c 2_ini Representing the initial learning ability, c 1_fin 、c 2_fin Representing the learning ability when the iteration is finished, wherein T represents the current iteration times, and T represents the maximum iteration times;
step 7090: mapping the sine value of the particle speed to the probability value of the change of the particle position vector, and calculating the formula shown in formula (14);
wherein v is the velocity value of the particle and f (v) is the probability value that maps the sine of the velocity of the particle to the change of the position vector of the particle;
step 7100: changing the search path of the particle using equation (15);
wherein Levy (epsilon) is a Levis flight search path, u obeys a normal distribution curve, the value range of beta is (0,2), and u-N (0, sigma) 2 ) V to N (0,1), σ is defined as follows:
wherein the value range of beta is (0,2), gamma function is represented by gamma, and the Levy (epsilon) of the Levy flight is determined by the formulas (15) and (16);
step 7110: updating the particles according to the non-dominated sorting of formula (17) according to the fitness value of the optimal particles obtained from step 7020 to step 7100:
wherein, for minimization problems, X 1 、X 2 There are two solutions before non-dominant ordering, solution X 1 The values on all targets are no worse than X 2 And is in X 1 Above, at least one of the objective function values is better than X 2 Is called X 1 Dominating X 2 ，X 1 For non-dominant solutions, X 2 To the dominant solution;
step 7120: updating the external archive set according to the obtained optimal particles according to the domination relationship, and executing the step 8000;
step 8000: the method adopts a data field and a rotating base technology to guide population evolution on the external archive obtained in the step 7120, and further comprises the following steps:
step 8100: judging the number of non-dominated solutions in the external archive set and the maximum value of the non-dominated solutions, if the number of the non-dominated solutions in the external archive set exceeds the maximum value, performing 8200 to adjust and delete the external archive set, otherwise, performing 8300 to obtain class 1 guide particles;
step 8200: mapping a non-dominant solution set in an external archive set into a gravitational data field, introducing a calculation method of potential energy in the data field, wherein the calculation formula is shown in formula (18), calculating the potential energy of each particle in the external archive set, deleting the external archive set, gradually deleting the particles with high potential energy until the number of the non-dominant solution sets is the maximum value of the external archive set, and the calculation formula of the potential energy of the data of the particles i is as follows:
wherein: psi (x) i ) Is the data potential energy of the ith particle, ρ j Representing the property value of the particle j in the data field, σ being the influence factor of the particle in the data field, x i Is the position information of the ith particle, R ij Representing the Euclidean distance between the particle i and the particle j, N being the existence in the search spaceN particles;
step 8300: introducing a data field potential energy formula (18) of the particles, calculating the potential energy of each particle in an external filing set, reflecting the distribution form of population particles, sequencing the potential energy of each particle, selecting guide particles according to the potential energy, wherein the particles selected by the data field potential energy principle are named as class 1 guide particles and are named as GF;
step 8400: selecting the particles with good distribution in the evolutionary population as class 2 guide particles, and naming the particles as GE, further comprising the following steps:
step 8410: carrying out non-dominant sorting on the particles in the external archive set according to a formula (17), and placing the particles in different layers according to different sorting of the particles;
step 8420: using the distribution degree index F i See formula (19), calculating the distribution degree of the solution set in the target space, wherein the smaller the value of the distribution degree, the more uniformly the solution set is distributed in the target space, and the distribution degree index F i Is defined as:
wherein K represents the number of solutions in the Pareto solution set, and
denotes the average distance between them, d i Distance of the target vector expressed as the ith non-inferior solution to the target vector of the optimal solution, d i Is the formula (20):
wherein the content of the first and second substances,
the target vector representing the ith non-inferior solution,
the j and i are unequal, and the value range of K is [1, K ]]K represents the number of solutions in the Pareto solution set;
step 8430: adopting a rotation base technology, forming a rotation angle at a connecting line of a terminal point of each iteration and a drawing base point, accumulating and drawing the rotation angle into a mapping broken line, leading the performance of the optimal solution on each target to be more balanced, leading the mapping broken line to approach to a straight line finally, utilizing the relation between the mapping broken line and the performance of the optimal solution, sequencing the particles in an external filing concentration, adopting a rotation base mapping broken line distance L to carry out auxiliary judgment if the distribution indexes of the particles are the same, selecting the particles with large mass and uniform distribution to guide the particle evolution, wherein the definition of the L is shown in a formula (21):
wherein, theta i 、θ j Respectively representing the angle formed by the ith rotation and the jth rotation, wherein r is the maximum rotation time, and L is the rotation base mapping broken line distance;
step 8440: finally, selecting the particles with large particle mass and uniform distribution as class 2 guide particles, and naming the class 2 guide particles as GE;
step 8500: s with large distance selected according to step 8400 GE The 2 nd type guide particle performs guide evolution on the population particle in the gravity search algorithm evolution mode improved in step 7000, and the specific guide mode of the 2 nd type guide particle is as follows:
wherein the content of the first and second substances,
the class 2 guide particle at time t +1,
denotes the mass of the jth type 2 guide particle at time t, R ij (t) denotes the Euclidean distance, x, between particle i and particle j at time t i (t) and x j (t) position information of the particle i and the particle j at time t, G is an attractive force constant, S GE Directing the number of particles for class 2;
step 8600: the population particles move under the common guidance of GF and GE guiding particles, and the updating mode of the speed and the position information of the particles is shown as the formula (23):
rand i is in the value range of [0,1 ]]，
Representing the velocity of the particle i at time t,
representing the velocity of particle i at time t +1,
representing the position information of particle i at time t +1,
the class 1 guide particle at time t +1,
guiding particles for class 2 at time t + 1;
step 8700: if the fitness value is good enough or the maximum iteration number is reached, terminating the loop, executing step 9000, otherwise, turning to step 7020 to continue execution;
step 9000: and outputting an optimal solution set, namely the obtained optimal heterogeneous wireless network service access control scheme.
2. An apparatus employing the multi-objective based heterogeneous wireless network service access control method according to any one of claim 1, the apparatus comprising:
a data acquisition module: is used for collecting the parameters in the steps 1000 and 2000;
an objective function determination module: determining the multi-objective optimization function model in the step 6000 according to the user information transmission rate, the service resource occupation and the network blocking rate in the steps 3000, 4000 and 5000;
a model solving module: and solving the multi-target optimization function model in the step 6000 by using the solving algorithm in the steps 7000 and 8000 to finally obtain the optimal heterogeneous wireless network service access scheme.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210060061.1A CN114363996B (en) | 2022-01-19 | 2022-01-19 | Heterogeneous wireless network service access control method and device based on multiple targets |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210060061.1A CN114363996B (en) | 2022-01-19 | 2022-01-19 | Heterogeneous wireless network service access control method and device based on multiple targets |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114363996A CN114363996A (en) | 2022-04-15 |
| CN114363996B true CN114363996B (en) | 2022-08-26 |
Family
ID=81090431
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210060061.1A Active CN114363996B (en) | 2022-01-19 | 2022-01-19 | Heterogeneous wireless network service access control method and device based on multiple targets |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114363996B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115334001B (en) * | 2022-10-18 | 2023-02-17 | 音信云(武汉)信息技术有限公司 | Data resource scheduling method and device based on priority relation |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101383845A (en) * | 2008-10-15 | 2009-03-11 | 北京邮电大学 | Integrated service access device, system and control method |
| CN102497660A (en) * | 2011-12-09 | 2012-06-13 | 电子科技大学 | Multi-service concurrency accessing system and accessing method used for heterogeneous wireless network |
| CN102752806A (en) * | 2012-06-26 | 2012-10-24 | 华为技术有限公司 | Admittance control method and device |
| CN104185248A (en) * | 2014-02-19 | 2014-12-03 | 上海物联网有限公司 | Hierarchy-based heterogeneous network joint access control method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8639267B2 (en) * | 2008-03-14 | 2014-01-28 | William J. Johnson | System and method for location based exchanges of data facilitating distributed locational applications |
| CN102946641B (en) * | 2012-11-27 | 2016-04-06 | 重庆邮电大学 | Isomery UNE bandwidth resources optimizing distribution method |
| CN103561457B (en) * | 2013-10-25 | 2017-04-05 | 华南理工大学 | A kind of multi-target networks power distribution method in heterogeneous wireless network collaboration communication |
| CN107018552A (en) * | 2016-01-27 | 2017-08-04 | 南水北调中线干线工程建设管理局 | A kind of method for selecting heterogeneous network access |
-
2022
- 2022-01-19 CN CN202210060061.1A patent/CN114363996B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101383845A (en) * | 2008-10-15 | 2009-03-11 | 北京邮电大学 | Integrated service access device, system and control method |
| CN102497660A (en) * | 2011-12-09 | 2012-06-13 | 电子科技大学 | Multi-service concurrency accessing system and accessing method used for heterogeneous wireless network |
| CN102752806A (en) * | 2012-06-26 | 2012-10-24 | 华为技术有限公司 | Admittance control method and device |
| CN104185248A (en) * | 2014-02-19 | 2014-12-03 | 上海物联网有限公司 | Hierarchy-based heterogeneous network joint access control method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114363996A (en) | 2022-04-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111585816B (en) | Task unloading decision method based on adaptive genetic algorithm | |
| CN110968426B (en) | Edge cloud collaborative k-means clustering model optimization method based on online learning | |
| CN112788605B (en) | Edge computing resource scheduling method and system based on double-delay depth certainty strategy | |
| CN111711666B (en) | Internet of vehicles cloud computing resource optimization method based on reinforcement learning | |
| CN112118312B (en) | Network burst load evacuation method facing edge server | |
| CN112784362A (en) | Hybrid optimization method and system for unmanned aerial vehicle-assisted edge calculation | |
| CN110794965B (en) | Virtual reality language task unloading method based on deep reinforcement learning | |
| CN112153145A (en) | Method and device for unloading calculation tasks facing Internet of vehicles in 5G edge environment | |
| CN114363996B (en) | Heterogeneous wireless network service access control method and device based on multiple targets | |
| CN114585006B (en) | Edge computing task unloading and resource allocation method based on deep learning | |
| CN113810233A (en) | Distributed computation unloading method based on computation network cooperation in random network | |
| CN114423063A (en) | Heterogeneous wireless network service access control method and device based on improved gravity search algorithm | |
| CN113507503A (en) | Internet of vehicles resource allocation method with load balancing function | |
| CN112996058A (en) | User QoE (quality of experience) optimization method based on multi-unmanned aerial vehicle network, unmanned aerial vehicle and system | |
| CN116321307A (en) | Bidirectional cache placement method based on deep reinforcement learning in non-cellular network | |
| CN113676357B (en) | Decision method for edge data processing in power internet of things and application thereof | |
| CN113747450B (en) | Service deployment method and device in mobile network and electronic equipment | |
| CN113139639A (en) | MOMBI-based smart city application-oriented multi-target calculation migration method and device | |
| CN116916386A (en) | Large model auxiliary edge task unloading method considering user competition and load | |
| CN116916272A (en) | Resource allocation and task unloading method and system based on automatic driving automobile network | |
| CN116112981A (en) | Unmanned aerial vehicle task unloading method based on edge calculation | |
| CN114444240B (en) | Delay and service life optimization method for cyber-physical system | |
| CN113159539B (en) | Method for combining green energy scheduling and dynamic task allocation in multi-layer edge computing system | |
| CN112165721B (en) | Multi-service task unloading and service migration method based on edge computing | |
| CN114339879A (en) | Service migration method based on reinforcement learning in mobile edge calculation |
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 |