CN114501551B - Multi-user distributed heterogeneous network selection strategy method based on ordered potential game - Google Patents
Multi-user distributed heterogeneous network selection strategy method based on ordered potential game Download PDFInfo
- Publication number
- CN114501551B CN114501551B CN202210345625.6A CN202210345625A CN114501551B CN 114501551 B CN114501551 B CN 114501551B CN 202210345625 A CN202210345625 A CN 202210345625A CN 114501551 B CN114501551 B CN 114501551B
- Authority
- CN
- China
- Prior art keywords
- network
- video
- strategy
- user
- unmanned aerial
- 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
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
- H04W28/08—Load balancing or load distribution
- H04W28/09—Management thereof
- H04W28/0925—Management thereof using policies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
- H04L65/80—Responding to QoS
-
- 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]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Multimedia (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a multi-user distributed heterogeneous network selection strategy method based on ordered potential game, which is applied to the selection control of a plurality of video users of a unmanned aerial vehicle cluster to access a network and comprises the following steps: s1, determining a network selection model based on an ordered potential game; s2, determining a utility function in the game process; and S3, adopting a multi-video user heterogeneous network selection distributed algorithm to solve the game model. The method realizes the optimal decision aiming at the scene that a plurality of users share a plurality of heterogeneous networks for video transmission, effectively solves the problem of multi-node access network selection, and ensures that the overall video experience quality of the multi-video users is the best.
Description
Technical Field
The invention relates to the field of unmanned aerial vehicle cluster transmission control, in particular to a multi-user distributed heterogeneous network selection strategy method based on ordered potential game.
Background
For an unmanned aerial vehicle cluster formed by multiple unmanned aerial vehicles, the information carrying capacity of a single network is limited, and video transmission has certain bandwidth requirements, so that the selection of an access network cannot be arbitrary, and the decision results of multiple nodes are inevitably influenced mutually. Since the network status information is dynamic, secondly, the network selection behavior of the user may further cause dynamic changes of the network status information. For example, when the number of users in a network increases, the probability of congestion increases. This information, i.e. how many users have selected a certain same network for transmitting video, may not be known to the sending node itself. On one hand, due to the limited resources, the occupation of the network resources by each user inevitably forms a competitive relationship in the group. An individual user needs to select a "premium" wireless network that is advantageous to him/herself, e.g., a network with sufficient channel bandwidth, a low packet loss rate, and a low charging standard. On the other hand, the choices affect each other, especially in the problem of packet loss caused by congestion. For an operator of the whole unmanned aerial vehicle cluster, the cluster is a whole, and the return quality of the video needs to be measured by integrating the overall effect of all transmissions. Therefore, for a scene that a plurality of users share a plurality of networks for video transmission, a network selection algorithm needs to be found to ensure that each wireless node can compete with each other and partially cooperate with each other, so that the user experience quality brought by the video transmission of the whole system can be optimal globally, and the problem of multi-node access network selection is effectively solved. The Potential Game (PG) method is used as a branch of Game theory, and can represent the motivation of all unmanned aerial vehicles for changing the strategy as a global function, thereby providing an idea for solving the network selection problem.
Disclosure of Invention
The invention mainly aims to provide a multi-user distributed heterogeneous network selection strategy method, which aims at a scene that a plurality of users share a plurality of heterogeneous networks for video transmission, realizes an optimal decision, effectively solves the problem of multi-node access network selection and enables the overall video experience quality of the multi-video users to be the best.
Based on the above purpose, the invention provides a multi-user distributed heterogeneous network selection strategy method based on ordered potential game, which is characterized in that the method is applied to the selection control of multiple video users accessing to the network in an unmanned aerial vehicle cluster, and the method comprises the following steps:
s1, determining a network selection model based on an ordered potential game;
s2, determining a utility function in the game process;
s3, adopting a multi-video user heterogeneous network to select a distributed algorithm to solve a game model;
further, the ordering based potentialsThe network selection model of the game can be expressed asWherein, in the step (A),for unmanned aerial vehicle aggregation, i.e.An unmanned aerial vehicle video communication node needing video transmission;represents the firstA selection policy set for individual drones, wherein,Is a binary vector which represents that the unmanned aerial vehicle user is in the access network setThe network selection made in (1); wherein, unmanned aerial vehicleCan decide whether to select a networkCarry out video transmission, as shown in
Is the corresponding utility set;indicating in addition to the userSelection strategy for all other drones than that in whichRepresents a cartesian product;andin combination describe allThe behavior strategy of each unmanned aerial vehicle node is as follows
Further, unmanned aerial vehicleThe information that can be obtained is that other drones select a policy ofBy observing the congestion status of each network
When UAV j selects network k to transmit video, eachThe congestion degree of the network k faced by the unmanned aerial vehicle is reflected; network congestion may be represented by the bandwidth occupied by the network, i.e.
Wherein Z is a three-dimensional matrix,is that the kth network corresponds to a size in Z ofThe elements on the main diagonal of the two-dimensional matrix of (1) are all 0, and the rest values are all 1;video code rate vectors respectively transmitted by X unmanned aerial vehicles;is the total bandwidth of the network k,representing the congestion state of network k, then
Further, in step S2, the difference between the transmission quality and the transmission cost of the video is used as a utility function, that isWherein, in the step (A),utility vectors representing QoE corresponding to video transmissions of different drones,representing the cost vector after each node has selected the corresponding access network,is a constant coefficient with a total utility function vector of。
Further, to unmanned aerial vehicleIn particular, when the access network selected isAnd the transmission rate of the video isWhen the utility function related to the video quality can be expressed as the utility function related to the network stateFunction of correlation (6)
In the formula (6), the reaction mixture is,the video content representing the current time slot is,is a constant number of times, and is,in the form of a function of a logarithm,in the form of an exponential function of the signal,is a constant. To unmanned aerial vehicleIn other words, the frame rate of the videoAnd transmission rateAre all constant values;is aboutMonotonically increasing.
Further, the cost of a user accessing the network is related to the transmission rate of the video, i.e.
Wherein the content of the first and second substances,is as followsThe total cost factor associated with each network.
Further, unmanned aerial vehicleThe video transmission utility function can be expressedIs composed of
The network selection strategy problem based on the ordered potential game model can be expressed as
Wherein, the first and the second end of the pipe are connected with each other,a constraint condition is expressed in terms of the number of the elements,indicating the best selection strategy.
Further, in step S3, the game model is solved by using a regret matching algorithm, which has the general idea that: the probability of a certain unmanned plane changing its strategy is proportional to the regret degree of the unmanned plane not selecting other strategies at the past moment.
Further, the specific implementation steps of the algorithm include:
and S32, an iterative updating process, wherein the iterative updating process comprises two substeps of strategy updating and strategy judgment.
Further, in the policy updating step, whenThen, each node calculates the current policy separatelyAnd selecting another policyThe utility of time, and calculate the average difference between these two utilities:
wherein the content of the first and second substances,represents time and. Then, getThe value is the average regret factor;
further, in the policy decision step, in the time slotOf 1 atStrategy of individual unmanned aerial vehicleThen is atThe slot, the drone will reconsider the policy and its basis for selecting the policy will obey the following probability distribution:
wherein the content of the first and second substances,. According to the distribution rule, the strategy space can be divided intoIn be unmanned aerial vehicleThe strategy is selected probabilistically. After solving the equations (10) and (11) through multiple iterations, the calculation and selection results are not changed any more, and the algorithm is converged; and each user follows the distributed algorithm updating strategy, and the whole network selection potential game is finally converged to a balanced state.
According to the method, a multi-video user distributed access network selection model based on the ordered potential game is established aiming at the scene that a plurality of users share a plurality of heterogeneous networks for video transmission, so that the wireless nodes can compete with each other and have partial cooperation, the problem of multi-node access network selection is effectively solved, the overall video experience quality of the multi-video users is the best, the problem of network congestion is solved, reasonable selection is carried out among the networks, and the load of each network is kept balanced basically.
Drawings
Fig. 1 is a diagram of a selection result of a user corresponding to 7 segments of videos on 3 networks by using a multi-user distributed heterogeneous network selection strategy method based on ordered potential gaming in the embodiment of the present invention;
FIG. 2 is a diagram illustrating congestion levels of various networks according to an embodiment of the present invention;
FIG. 3 is a diagram illustrating the loading of various networks in an embodiment of the present invention;
FIG. 4 is a total usage graph of all video users in an embodiment of the present invention;
fig. 5 is a diagram of a network selection result of each user when transmission cost is considered in the embodiment of the present invention;
FIG. 6 is a diagram of network congestion level when transmission cost is taken into account in an embodiment of the present invention;
FIG. 7 is a diagram of network load considering transmission cost in an embodiment of the present invention;
FIG. 8 is a total utility diagram of the system in consideration of transmission cost according to an embodiment of the present invention;
fig. 9 is a diagram illustrating network load conditions at different transmission costs according to an embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of description and simplicity of description, but do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.
The following detailed description of embodiments of the invention refers to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating the present invention, are given by way of illustration and explanation only, not limitation.
Potential games can be further classified into Exact Potential Games (EPG), Weighted Potential Games (WPG), and Ordered Potential Games (OPG). In recent years, there has been an increasing number of potential gaming methods applied to study wireless channel access and network selection issues. Under the condition that joint channel interference is symmetrical and channel effectiveness is additive, a selection problem in a heterogeneous wireless network is modeled into a selection game of a resource block, the game is the sum of ordered potential games, and an ordered potential function is closely related to total congestion of the network. A distributed channel selection mechanism based on potential game is needed to solve the problem of current interruption in the dense wireless local area network. The problem of unmanned aerial vehicle allocation and channel access is solved by adopting a dynamic layered game model, wherein the channel access problem is modeled as a strict potential game.
Compared with other two potential game algorithms, the ordered potential game is more general, and only the individual income change and the change of the overall potential function have the same trend, so that more space is provided for the design of the utility function, and the setting of the utility function mainly based on the video transmission quality can be considered from more layers when the selection of the access network of a plurality of nodes is decided. In addition, in the ordered potential game, the process of searching local optimum by the participant individuals is consistent with the process of searching global optimum by the whole game, so that the realization of distributed access network selection is possible.
The application provides a multi-user distributed heterogeneous network selection strategy method based on an ordered potential game aiming at a scene that multiple users select access networks in a heterogeneous network, and optimizes a network selection strategy from the perspective of improving video transmission quality. According to the multi-user distributed heterogeneous network selection strategy method based on the ordered potential game, the unmanned aerial vehicle carrying the video shooting and wireless communication module is taken as a main video transmission user, and multiple wireless network selection problems in the communication process of aerial shooting and real-time video return of multiple unmanned aerial vehicles are researched.
The multi-user distributed heterogeneous network selection strategy method is based on the existenceUnmanned aerial vehicle cluster formed by unmanned aerial vehiclesFly in the air atNetwork set composed of heterogeneous networksWithin the common coverage area W of the first and second,unmanned aerial vehicle sharingEach wireless video node needs to select a proper network to transmit the acquired video back to a certain receiving node, and the total video transmission quality of the system is the highest.Video code rate vectors transmitted by the unmanned aerial vehicle are respectively fixed as。The network selection vector of the unmanned aerial vehicle can be expressed asCorresponding utility function of. In this scenario, the total bandwidth that each heterogeneous network can provide remains the same, i.e.The bandwidth information vector of the heterogeneous network can be expressed as。
Video data packets are transmitted in a network, and two reasons for loss generally exist, one is that a channel of a wireless network has a fading phenomenon, which easily causes error codes and packet loss; the other is that the packet forwarding delay caused by network congestion exceeds the threshold value which can be tolerated by the system, and the data packet is lost. Wherein the packet loss caused by the error code is related to the transmission characteristic of the wireless channel. The latter, i.e. packet loss due to network congestion, is mainly considered in the scenarios discussed in the present application. Since the user can freely select any one of the networks, the selected network results in different degrees of congestion of each network, and the congestion degree of each network results in different packet loss rates, which adversely affects the user selection.
In the first placeIn the network, the network is divided into a plurality of networks,the total time delay of the coded video from the sending node to the receiving node in the transmission process isThe transmission delay threshold of the video data packet is. The packet loss rate vector of each network can be expressed as. The end-to-end transmission delay can be approximated as following an exponential distribution, then at the secondThe probability of a packet being lost in a network can be expressed as. Wherein the content of the first and second substances,represents a mathematical expectation of the transmission delay of video data packets,is a smoothing parameter derived from historical observations,for the remaining bandwidth of the network,is a networkThe total bandwidth of the network (c) is,the representative selects a networkThe sum of the transmission code rates of the videos of the unmanned aerial vehicles. Thus, the network is lostThe packet rate can be expressed as:
thus, the packet loss rate of a network is related to both the bandwidth capacity of the network and the coding rate of the video transmitted by the nodes accessing the network. By usingRepresentative networkIn a congested state of
In summary, each wireless network can provide different total bandwidths to users, and the packet loss rates of the wireless networks are different, so that when a plurality of users select to access a network, the users all benefit themselves, and hope that a better video transmission effect can be obtained, and meanwhile, the selection of each user affects the network state. How to optimize the network selection of each user to maximize the total utility of the system is a key technology studied by the application.
Due to the heterogeneity of wireless networks, the available transmission bandwidths provided by each network to users are different, and with the change of the number of users accessed, QoS attributes such as the available bandwidth of a channel, packet loss rate, transmission delay and the like also change, which in turn affects the selection of video users. In the application, a multi-video user distributed access network selection model based on the ordered potential game is established aiming at a scene that a plurality of users share a plurality of heterogeneous networks for video transmission.
The method comprises the following steps:
s1, determining a network selection model based on the ordered potential game. The network selection model based on the ordered potential game can be expressed asWherein, in the step (A),for a set of participants, i.e.And the unmanned aerial vehicle video communication node is required to transmit videos.Represents the firstA set of selection policies of individual participants, wherein,Is a binary vector which represents that the unmanned aerial vehicle user is in the access network setThe network selection made in (1); wherein the participantsCan decide whether to select a networkCarry out video transmission, as shown in
For the corresponding utility set, it is related to the video QoE, which will be described in detail in the following section.Indicating in addition to the userSelection strategy of all participants except, whereinRepresenting the cartesian product.Andin combination describe allThe behavior strategy of each unmanned aerial vehicle node is as follows
In the process that the unmanned aerial vehicle selects to access the network, each participant does not know the behaviors of other participants, and therefore the unmanned aerial vehicle is an incomplete information game. First, theThe information that can be obtained by each participant is that other unmanned aerial vehicles select the strategy asBy observing the congestion status of each network
When participant j selects network k to transmit video, eachReflects the degree of congestion of the network k faced by the participant, which depends only on the transmission of other nodes occupying the network channel, which occupation is also an interference, as understood from the communication point of view. Network congestion may be represented by the bandwidth occupied by the network, i.e.
Wherein Z is a three-dimensional matrix.Is that the kth network corresponds to a size in Z ofThe elements on the main diagonal of the two-dimensional matrix of (2) are all 0, and the remaining values are all 1.Video code rate vectors respectively transmitted by X unmanned aerial vehicles;is the total bandwidth of the network k and,representing the congestion state of network k, then
And S2, determining a utility function in the game process. In order to reasonably determine the access network selection strategy according to the network parameters and accurately evaluate the performance of the network selection strategy, a utility function in the game process needs to be determined. In a distributed network selection strategy of a plurality of wireless nodes, the determination of the utility function needs to fully consider the QoE of the video. In distributed network selection, the QoE utility is calculated not through the centralized calculation of a base station, but each unmanned aerial vehicle node performs independent calculation based on a QoE mathematical model according to observed channel state information in the game process, and the QoE utility is a prediction on video transmission quality.
The present application considers using the difference between the transmission quality and the transmission cost of the video as a utility function, i.e.Wherein, in the process,utility vectors representing QoE corresponding to video transmissions of different drones,representing the cost vector after each node has selected the corresponding access network,is a constant coefficient with a total utility function vector of。
First, a video classification method is used, i.e. the video is classified into three categories of slow SM, medium GW and fast RM according to the content characteristics of the video itself, and the video is still usedThe video content representing the current time slot can be classified into one of the three types described above. To the firstFor an UAV, when the selected access network isAnd the transmission rate of the video isWhen the utility function related to the video quality can be expressed as the utility function related to the network stateFunction of correlation
In the formula (8), the reaction mixture is,the video content representing the current time slot is,is a constant number of times, and is,in the form of a function of a logarithm,in order to be an exponential function of the,is a constant. To unmanned aerial vehicleIn other words, the frame rate of the videoAnd transmission rateAll are constant values. As can be seen,is aboutMonotonically increasing.
In the present application, the cost of the user accessing the network mainly considers two aspects: one aspect is the cost of leasing the channel from the network service provider, sinceThe individual networks may belong to different network service providers, which often have different charging standards, setting cost factors during transmission of the individual networks. For non-commercial systems, the cost factor is 0. On the other hand, the energy consumption during transmission is related to the specific network environment and channel type, for example, the energy consumption between the satellite network and the ground mobile network is greatly different, and the energy consumption factor is set as. Both costs of cost and energy consumption are related to the transmission rate of the video, i.e.
Wherein the content of the first and second substances,is as followsThe total cost factor associated with each network. Then, unmanned planeThe video transmission utility function can be expressed as
The network selection strategy problem based on the ordered potential game model can be expressed as
Wherein, the first and the second end of the pipe are connected with each other,representing the constraint.
And S3, adopting a multi-video user heterogeneous network selection distributed algorithm to solve the game model. The process of solving Nash equilibrium in the game is a process of finding the optimal solution through continuous iteration. The game model is solved by utilizing a Regret Matching algorithm, and a multi-video user heterogeneous network selection distributed algorithm based on the ordered potential game is designed. The general idea of the algorithm is as follows: the probability of a certain unmanned plane changing its strategy is proportional to the regret degree of the unmanned plane not selecting other strategies at the past moment. The specific implementation steps comprise initialization and iterative update processes, wherein:
s31, initialization
At first, the methodWhen each participant is in the policy spaceOne is randomly selected. In fact, the initial policy may be any value within the scope of the policy space.
And S32, an iterative updating process. The iterative update process further comprises two sub-steps of policy update and policy decision, wherein:
s321. strategy update
When in useThen, each node calculates the current policy separatelyAnd selecting another policyThe utility of time, and calculate the average difference between these two utilities:
wherein the content of the first and second substances,represents time and. Then, getI.e. the average regret factor.
S322. strategy judgment
Suppose in a time slotOf 1 atPolicy of individual participantsThen is atThe participant will reconsider the policy and the basis of his choice of policy will obey the following probability distribution:
wherein the content of the first and second substances,is sufficiently large. According to the distribution rule, the strategy space can be divided intoThe middle is a participantThe strategy with the higher probability is selected.
After solving the equations (12) and (13) through multiple iterations, the calculation and selection results are not changed any more, and the algorithm converges. If each user follows the above distributed algorithm updating strategy, the whole network election potential game will finally converge to an equilibrium state.
In one embodiment, assuming that 7 drones are performing the filming task in the public coverage area of 3 heterogeneous networks, they independently shoot videos and send 7 different videos back to the same central user through the wireless network. Typical transmission rates for a particular video are further described in the present application, as shown in table 1 below. Generally, the slower the content of a video picture moves, the less transmission resources are occupied by the video, e.g., Akiyo. In addition, the size of the video transmission rate is related to the complexity of the scene, which also occupies more bandwidth, such as Coastguard. Here, the 7 segments of video are from several different scenes, and in practical cases, the number of videos to be transmitted may be more, the scenes are more complex, but the basic principle and the flow of the algorithm are consistent.
TABLE 1 parameters and characteristics of different videos
The 3 kinds of heterogeneous networks covering the flight area of the unmanned aerial vehicle are not particularly limited, and may be a ground mobile communication network, a satellite communication network, and the like, which are respectively referred to as network 1, network 2, and network 3, and may be used for simulation and verification of an algorithm by adjusting parameters such as bandwidth and cost factor of the network to characterize the heterogeneity of the network. Assuming that the total transmission bandwidth of the three networks is matched with the total bandwidth of video transmission, i.e. 5.2 Mbps, 4.8M bps and 4.4 Mbps, the total rate obtained by summing up the 7 video transmission rates in table 1 is about 5.485 Mbps, and the total bandwidth of any single network cannot meet the requirement of all video transmission. Therefore, it is necessary to make a reasonable choice between the networks so that the load of each network is basically balanced.
Two situations are considered, one is a user insensitive to cost, and the transmission cost of each network is not considered, and in this case, according to the utility function, the user selects the network only related to the packet loss rate of the network, that is, related to the congestion degree of the network. And the other type of the method needs to consider the difference of the heterogeneous network packet loss rate and the cost factor at the same time.
Firstly, the selection conditions of different video users for 3 heterogeneous networks are analyzed without considering the transmission cost of each network, namely in the case that the cost factor is equal to 0. Fig. 1 shows a certain selection result of 7 video users for 3 heterogeneous networks in simulation, where the ordinate is a user who selects a network and the abscissa is a network. It can be seen that two users, namely, the phone and the Football user, select the network 1, four users, namely, the Akiyo user, the Coastguad user, the Mobile user and the Table user, select the network 2, and the Forman user selects the network 3, which indicates that each user can select a corresponding network after the user nodes adopt the distributed network selection strategy provided by the application under the condition that no information interaction exists among the user nodes.
Fig. 2 shows the change of the congestion degrees of three heterogeneous networks with time, and the ordinate is the normalized network congestion degree, and it can be seen that after about 20 iterations, the congestion conditions of each network basically remain stable, which indicates that the user selection reaches a balanced state and remains stable. It can also be seen that the ideal result should be that the congestion level of each network is substantially the same, i.e. 1/3, but because the number of users in the example is small, the transmission rate difference between users is large, and the congestion of the network 1 is slightly higher due to the step effect. Fig. 3 shows the average load of the three networks, and it can be seen that the loads of the three networks are sequentially reduced, which is proportional to the total bandwidth that they can provide, and more users select the network with more total bandwidth, while the number of users selecting the network with smaller bandwidth is less.
Fig. 4 shows the total utility of all videos as a function of time, in the figure, the abscissa represents time, and the ordinate represents the total utility of the videos, and it can be seen that after about 20 time slots, the total utility tends to be stable, which indicates that each user makes a reasonable selection and keeps stable, and the multi-user heterogeneous network selection game reaches relevant balance. Meanwhile, for performance comparison, the total utility function when the user randomly selects the network is also given in the figure, as shown by the dotted line in the figure, it can be seen that the method provided by the application has obvious performance improvement.
In the simulation process of the application, the influence of the network congestion degree is considered, and the influence of the transmission cost of the network is considered, for example, in practical application, the energy consumption of a satellite network is generally larger than that of a ground mobile network. In order to make the simulation more specific, without loss of generality in the present application, cost factors of three heterogeneous networks are assumed to be 1, 2, and 3, respectively. The bandwidth and cost of the network 1 are maximum and the cost of the network 3 is minimum and the cost is maximum, and the number of the networks is 2.
Fig. 5 shows the network selection result of each user in consideration of the transmission cost, with the ordinate of the user selecting the network and the abscissa of the network. In fig. 5, the network 1 has the largest bandwidth and the lowest cost, and the network 1 is selected by the videos Akiyo, phone, Coastguard, Mobile, and Table; football selects network 2 and Foreman selects network 3.
Fig. 6 shows the congestion degree of each network in consideration of the transmission cost. In fig. 6, the ordinate is normalized network congestion degree, and since the cost of the network 1 is minimum and the bandwidth is maximum, more video users select the network 1, which results in obviously higher congestion degree of the network 1; the network 3 selects fewer users due to smaller bandwidth and higher cost, so that the congestion degree is obviously lower; the congestion level of the network 2 is in between.
Fig. 7 shows the network load of each network in consideration of the transmission cost. Similar to the case of network congestion, the user makes a selection based on the network bandwidth and cost. In fig. 7, network 1 is loaded most and network 3 is loaded least, with the rule being consistent with the network bandwidth. At the same time, the gap between the network loads is further increased compared to fig. 3, because of the effect of the cost factor. The network load is further increased because the cost of the network 1 is minimal.
Fig. 8 shows the total system utility when the transmission cost is considered, and the ordinate is the total system utility, it can be seen that the total system utility can converge to a stable value, and is significantly better than the performance of the user randomly selecting the network.
Finally, the load simulation results of three networks are shown in fig. 9, wherein the abscissa represents 1 to take the cost factor [ 321 ], the abscissa represents 2 to take the cost factor [ 222 ], and the abscissa represents 3 to take the cost factor [ 123 ]. It can be seen that for each network, as the cost factor gradually increases, the network load gradually decreases, and when the cost is the same, the ratio of the network load to the total bandwidth is the same, which indicates that the algorithm can adapt to various situations with different bandwidths and costs. According to the utility function, the cost factors, besides the factors such as energy consumption and the like which cannot be easily changed, can also be considered, and the cost factors and the like indicate that each network can automatically adjust the access condition of the user by adjusting the cost factors, so that the network load is ensured.
By adopting the distributed network selection algorithm, users can only perceive the utility no matter how the network parameters change, and the users do not need to interact with each other network selection information, and the optimal network selection can be realized by the game among multiple users to achieve balance.
In the description herein, references to the description of "an embodiment," "an example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, various embodiments or examples described in this specification and features thereof may be combined or combined by those skilled in the art without contradiction.
Although embodiments of the present invention have been shown and described, it will be understood that the embodiments are illustrative and not restrictive, and that modifications, changes, substitutions and variations may be made by those skilled in the art without departing from the scope of the present invention.
Claims (8)
1. A multi-user distributed heterogeneous network selection strategy method based on ordered potential game is characterized in that the method is applied to selection control of multiple video users accessing to a network in an unmanned aerial vehicle cluster, and comprises the following steps:
s1, determining a network selection model based on an ordered potential game;
s2, determining a utility function in the game process;
s3, adopting a multi-video user heterogeneous network to select a distributed algorithm to solve a game model;
the network selection model based on the ordered potential game can be expressed asWherein, in the step (A),for unmanned aerial vehicle sets, i.e.An unmanned aerial vehicle video communication node needing video transmission;represent unmanned aerial vehicleIn which the selection policy set is,Is a binary vector which represents that the unmanned aerial vehicle user is in the access network setThe network selection made in (1); wherein, unmanned aerial vehicleCan decide whether to select a networkCarry out video transmission, as shown by
Is the corresponding utility set;indicating in addition to the userSelection strategy for all other drones than that in whichRepresents a cartesian product;andin combination describe allThe behavior strategy of individual UAV users, therefore, has
Unmanned planeThe information that can be obtained is that other drones select a policy ofBy observing the congestion status of each network
When unmanned aerial vehicleSelecting a networkWhen transmitting video, eachAll reflect the network faced by the unmanned aerial vehicleThe congestion level of; network congestion may be represented by the bandwidth occupied by the network, i.e.
Wherein the content of the first and second substances,is a three-dimensional matrix of which the matrix is,is the firstA network is inOne size corresponding to inThe elements on the main diagonal of the two-dimensional matrix of (1) are all 0, and the rest values are all 1;is composed ofVideo code rate vectors respectively transmitted by the unmanned aerial vehicles;is a networkThe total bandwidth of the network (c) is,representative networkIn a congested state of
2. The ordered potential game-based multi-user distributed heterogeneous network selection strategy method according to claim 1, wherein: in step S2, the difference between the transmission quality and the transmission cost of the video is used as a utility function, that isWherein, in the step (A),utility vectors representing QoE corresponding to video transmissions of different drones,representing the cost vector after each node has selected the corresponding access network,is a constant coefficient with a total utility function vector of。
3. The multi-user distributed heterogeneous network selection strategy method based on ordered potential game as claimed in claim 2, wherein: to unmanned aerial vehicleIn particular, when the access network selected isAnd the transmission rate of the video isWhen the utility function related to the video quality can be expressed as the utility function related to the network stateFunction of correlation
In the formula (6), the reaction mixture is,the video content representing the current time slot is,is a constant number of times, and is,in the form of a function of a logarithm,in order to be an exponential function of the,as a constant to the unmanned planeIn other words, the frame rate of the videoAnd transmission rateAre all constant values;is aboutMonotonically increasing.
4. The ordered potential game-based multi-user distributed heterogeneous network selection strategy method according to claim 3, wherein:
the cost of a user accessing the network is related to the transmission rate of the video, i.e.
Wherein, the first and the second end of the pipe are connected with each other,is as followsTotal cost factor associated with individual network, unmanned aerial vehicleThe video transmission utility function can be expressed as
The network selection strategy problem based on the ordered potential game model can be expressed as
5. The multi-user distributed heterogeneous network selection strategy method based on ordered potential game according to any one of claims 1, 3 and 4, characterized in that in step S3, the game model is solved by using a regret matching algorithm, whose overall idea is: the probability that a certain unmanned aerial vehicle user changes the strategy is in direct proportion to the regret degree of the unmanned aerial vehicle user who does not select other strategies at the past moment.
6. The ordered potential game-based multi-user distributed heterogeneous network selection strategy method according to claim 5, wherein the specific implementation steps of the algorithm comprise:
and S32, an iterative updating process, wherein the iterative updating process comprises two substeps of strategy updating and strategy judgment.
7. The ordered potential game-based multi-user distributed heterogeneous network selection strategy method according to claim 6, wherein in the strategy updating step, when the strategy is updated, the selection strategy is executedThen, each node calculates the current policy separatelyAnd selecting another policyThe utility of time, and calculate the average difference between these two utilities:
8. The ordered potential game-based multi-user distributed heterogeneous network selection strategy method according to claim 7, wherein in the strategy judgment step, in the time slotStrategy for drone jThen is atTime slots, the strategy will be reconsidered and its basis for selecting the strategy will obey the following probability distribution:
wherein, the first and the second end of the pipe are connected with each other,according to the distribution rule, the strategy space can be divided intoIn be unmanned aerial vehicleSelecting a strategy according to the probability;
after solving the formula (10) and the formula (11) through multiple iterations, the calculation and selection results are not changed any more, and the algorithm is converged; and each user follows the distributed algorithm updating strategy, and the whole network selection potential game is finally converged to a balanced state.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210345625.6A CN114501551B (en) | 2022-04-02 | 2022-04-02 | Multi-user distributed heterogeneous network selection strategy method based on ordered potential game |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210345625.6A CN114501551B (en) | 2022-04-02 | 2022-04-02 | Multi-user distributed heterogeneous network selection strategy method based on ordered potential game |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114501551A CN114501551A (en) | 2022-05-13 |
CN114501551B true CN114501551B (en) | 2022-07-01 |
Family
ID=81487282
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210345625.6A Active CN114501551B (en) | 2022-04-02 | 2022-04-02 | Multi-user distributed heterogeneous network selection strategy method based on ordered potential game |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114501551B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116634468B (en) * | 2023-07-21 | 2023-10-31 | 中国人民解放军军事科学院系统工程研究院 | Unmanned aerial vehicle channel selection method based on accurate potential game |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105700555A (en) * | 2016-03-14 | 2016-06-22 | 北京航空航天大学 | Potential game-based multi-unmanned aerial vehicle cooperative search method |
CN107949025A (en) * | 2017-11-02 | 2018-04-20 | 南京南瑞集团公司 | A kind of network selecting method based on non-cooperative game |
CN108616302A (en) * | 2018-04-28 | 2018-10-02 | 中国人民解放军陆军工程大学 | Unmanned plane Multi folds coverage model and dispositions method under a kind of power control |
CN108965009A (en) * | 2018-07-19 | 2018-12-07 | 广东南方电信规划咨询设计院有限公司 | A kind of load known users correlating method based on gesture game |
CN109495952A (en) * | 2018-11-14 | 2019-03-19 | 北京航空航天大学 | A kind of selection method and device of honeycomb and unmanned plane integrated network |
CN112584527A (en) * | 2020-12-14 | 2021-03-30 | 中国人民解放军军事科学院国防科技创新研究院 | Bandwidth allocation method and device for multi-node video transmission and electronic equipment |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103298076B (en) * | 2013-06-21 | 2017-06-20 | 西安邮电大学 | A kind of method of the access network selection in heterogeneous network |
CN103916912B (en) * | 2014-03-25 | 2017-07-11 | 南京邮电大学 | Wireless Heterogeneous Networks are based on the node cooperation motivational techniques of non-cooperative game |
-
2022
- 2022-04-02 CN CN202210345625.6A patent/CN114501551B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105700555A (en) * | 2016-03-14 | 2016-06-22 | 北京航空航天大学 | Potential game-based multi-unmanned aerial vehicle cooperative search method |
CN107949025A (en) * | 2017-11-02 | 2018-04-20 | 南京南瑞集团公司 | A kind of network selecting method based on non-cooperative game |
CN108616302A (en) * | 2018-04-28 | 2018-10-02 | 中国人民解放军陆军工程大学 | Unmanned plane Multi folds coverage model and dispositions method under a kind of power control |
CN108965009A (en) * | 2018-07-19 | 2018-12-07 | 广东南方电信规划咨询设计院有限公司 | A kind of load known users correlating method based on gesture game |
CN109495952A (en) * | 2018-11-14 | 2019-03-19 | 北京航空航天大学 | A kind of selection method and device of honeycomb and unmanned plane integrated network |
CN112584527A (en) * | 2020-12-14 | 2021-03-30 | 中国人民解放军军事科学院国防科技创新研究院 | Bandwidth allocation method and device for multi-node video transmission and electronic equipment |
Non-Patent Citations (3)
Title |
---|
Matching Game With No-Regret Learning for IoT Energy-Efficient Associations With UAV;Safae Lhazmir;《IEEE Transactions on Green Communications and Networking》;20200713;全文 * |
无人机视频回传中的动态资源分配机制;贺超;《计算机科学》;20191130;第3-5节 * |
有序势博弈及其在智能体无线网络中的应用;于永渊;《控制与决策》;20170331;全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN114501551A (en) | 2022-05-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Pantisano et al. | Match to cache: Joint user association and backhaul allocation in cache-aware small cell networks | |
CN113038616B (en) | Frequency spectrum resource management and allocation method based on federal learning | |
Hamidouche et al. | The 5G cellular backhaul management dilemma: To cache or to serve | |
CN114205316B (en) | Network slice resource allocation method and device based on power service | |
CN114501551B (en) | Multi-user distributed heterogeneous network selection strategy method based on ordered potential game | |
US20230021435A1 (en) | Method And System For Distributing Content By Pre-Positioning And Redistributing The Content To Other Devices | |
CN113038187A (en) | Practical network bandwidth allocation method with fair video experience quality | |
CN112040512A (en) | Fair-based fog computing task unloading method and system | |
Mehrabi et al. | Joint optimization of QoE and fairness through network assisted adaptive mobile video streaming | |
Pietrabissa | A reinforcement learning approach to call admission and call dropping control in links with variable capacity | |
CN113938394B (en) | Monitoring service bandwidth allocation method and device, electronic equipment and storage medium | |
Thakolsri et al. | QoE-driven cross-layer optimization in wireless networks addressing system efficiency and utility fairness | |
CN112584527B (en) | Bandwidth allocation method and device for multi-node video transmission and electronic equipment | |
Bedogni et al. | Dynamic adaptive video streaming on heterogeneous TVWS and Wi-Fi networks | |
CN110035539B (en) | Resource optimization allocation method and device based on relevant balanced regret value matching | |
US11785088B2 (en) | Method and system for controlling the use of dormant capacity distributing data | |
Jin et al. | Joint qos control and bitrate selection for video streaming based on multi-agent reinforcement learning | |
WO2023009382A1 (en) | Method and system for distributing content by pre-positioning and redistributing the content to other devices | |
Pham et al. | QoE-based routing algorithms for H. 264/SVC video over ad-hoc networks | |
JP7388634B2 (en) | Optimization method for wireless communication system, wireless communication system and program for wireless communication system | |
CN114554496A (en) | 5G network slice resource allocation method based on machine learning | |
Song et al. | Exploiting mobility to content placement in D2D caching systems | |
CN116033584B (en) | Air interface resource scheduling method, network access equipment and communication network | |
Liu et al. | Adaptive Live Streaming for Multi-user Access with Fairness Guarantee | |
Chiariotti et al. | QoE-aware Video Rate Adaptation algorithms in multi-user IEEE 802.11 wireless networks |
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 |