CN115942460A - Low-orbit satellite wireless resource scheduling method and device based on resource map and countermeasure learning - Google Patents
Low-orbit satellite wireless resource scheduling method and device based on resource map and countermeasure learning Download PDFInfo
- Publication number
- CN115942460A CN115942460A CN202211356622.9A CN202211356622A CN115942460A CN 115942460 A CN115942460 A CN 115942460A CN 202211356622 A CN202211356622 A CN 202211356622A CN 115942460 A CN115942460 A CN 115942460A
- Authority
- CN
- China
- Prior art keywords
- spot beam
- resource
- network
- spot
- dimensional
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000012549 training Methods 0.000 claims abstract description 33
- 238000005520 cutting process Methods 0.000 claims abstract description 17
- 230000008569 process Effects 0.000 claims abstract description 9
- 230000003042 antagnostic effect Effects 0.000 claims abstract description 4
- 238000004891 communication Methods 0.000 claims description 45
- 230000006870 function Effects 0.000 claims description 32
- 238000009826 distribution Methods 0.000 claims description 30
- 238000013507 mapping Methods 0.000 claims description 12
- 238000010586 diagram Methods 0.000 claims description 11
- 238000003860 storage Methods 0.000 claims description 8
- 238000011478 gradient descent method Methods 0.000 claims description 7
- 230000004907 flux Effects 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 5
- 238000000926 separation method Methods 0.000 claims description 4
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 238000012935 Averaging Methods 0.000 claims description 3
- 230000003595 spectral effect Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims 1
- 238000013468 resource allocation Methods 0.000 abstract description 22
- 238000005457 optimization Methods 0.000 abstract description 18
- 238000012545 processing Methods 0.000 abstract description 16
- 230000000737 periodic effect Effects 0.000 abstract description 8
- 238000005516 engineering process Methods 0.000 abstract description 6
- 230000008859 change Effects 0.000 description 8
- 238000013473 artificial intelligence Methods 0.000 description 7
- 230000033001 locomotion Effects 0.000 description 5
- 229910052739 hydrogen Inorganic materials 0.000 description 4
- 239000001257 hydrogen Substances 0.000 description 4
- 125000004435 hydrogen atom Chemical class [H]* 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000007792 addition Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 239000006185 dispersion Substances 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 230000006855 networking Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000013526 transfer learning Methods 0.000 description 1
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
Landscapes
- Radio Relay Systems (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention provides a low-orbit satellite wireless resource scheduling method and device based on resource map and antagonistic learning, based on a beam hopping technology application scene, firstly, a constraint condition is established by utilizing periodic prior information of a satellite network based on maximum throughput and priority requirements, and the time slot number distributed to each point beam is preliminarily solved; the method comprises the steps of establishing a three-dimensional histogram of wireless network resource scheduling by taking time slots, spot beam numbers and spot beam frequencies as parameters, cutting the three-dimensional histogram into two-dimensional pictures according to time slots to construct an infinite resource map, optimizing the whole time slot allocation scheduling of the two-dimensional pictures obtained by cutting through a pre-training generation countermeasure network, utilizing the advantages of images in continuous and explicit color system processing processes, utilizing the processing of the wireless resource map to realize fine-grained optimization, and improving the accuracy of a resource allocation strategy.
Description
Technical Field
The invention relates to the technical field of communication, in particular to a low-orbit satellite wireless resource scheduling method and device based on resource maps and countermeasure learning.
Background
With the commercial use of the 5G network, the problems of massive user access, large bandwidth and low delay communication in a densely populated area are solved. However, as the activity range, the production value, the emergency and the like of people are expanded in the global range, the construction cost of a ground network is high or the ground network cannot be constructed in a sparsely populated area, and the like, the wireless resources, the coverage area and the communication rate of information interaction in the global view are seriously insufficient. In recent years, satellite internet based on large constellation, low orbit and high frequency band has become one of important solutions to solve the above problems at low cost. Currently, air-ground integrated seamless communication fusing a satellite, an air-based platform and a ground network has become an important subject of 6G research.
Different from a ground network and an existing satellite communication system, the satellite internet mainly realizes high-concurrency user mobile broadband access in a global area. Therefore, an important innovative technical idea of the satellite internet is to integrate new technical means such as a ground 5G NR air interface, introduce phased array hopping beams, combine a micro-service core network and the like, and realize more flexible, larger-capacity and higher-speed communication support. Meanwhile, the communication service experience of the satellite internet is also focused. The operator plans to simultaneously realize the mobile broadband service of providing the peak rate of a single user of more than 100Mbps, the real-time interactive service of less than 50ms and the concurrent access of more than 100 ten thousand services of the Internet of things for the global ubiquitous users based on the satellite Internet. The services have great difference, different requirements on service guarantees such as bandwidth, delay and access quantity, and the like, but the services are distributed under the same beam of the satellite internet at a high probability or the distribution dynamically changes along with the region. In addition, the same satellite coverage ground area changes at a high speed, and the channel environment of the ground users changes sharply. Therefore, based on the satellite internet, optimal wireless resource allocation and scheduling provided for different regions and different users of the same beam are the key for improving the flexibility and reliability of satellite internet communication.
The core problem of satellite internet wireless resource scheduling comprises two points, wherein the first step is that fine-grained modeling is carried out on satellite wireless communication resources such as frequency, time, space and opportunity of beams under beam hopping; and secondly, the optimal selection of the multi-constraint relation when various heterogeneous services coexist. Firstly, the introduction of a ground 5G NR new air interface enables user data in the same wave beam of satellite communication to be multiplexed by an OFDM mode, and the dynamic wave beam forming capability provided by a phased array is combined, so that the satellite communication is endowed with a multiplexing mode of different frequencies, time slots and codes for different users in different areas at different times, and due to the high-speed change of the satellite, the allocated wireless resources at different moments corresponding to each user also change at a very high speed, and meanwhile, the requirements of the wireless resources are expanded from two dimensions of the ground to three dimensions of the air, the air and the ground by considering air-based platforms such as airplanes and airships and space-based platforms such as remote sensing, meteorology and space stations. The existing wireless communication resources are modeled by adopting single-point, fixed and two-dimensional ideas, and the requirements of accurate description on spatial multi-point, fine-grained change, three-dimensional distribution and the like of the wireless communication resources in a satellite internet scene are difficult to adapt. Secondly, in a satellite internet scene, the communication service requirements of large bandwidth, low delay and high concurrency exist in the same region, and meanwhile, as the relative distance between the satellite and the user is large and the satellite covers periodically, the service request initiated by the ground user is not subjected to simple random distribution, but has more spatial and periodic regular distribution.
In the prior art, a wireless resource scheduling algorithm based on artificial intelligence focuses on link state sensing to optimally control wireless resources, emphasizes reliable connection, and does not fully utilize prior information brought by relative fixed user distribution and periodic movement of a satellite, so that the efficiency of the artificial intelligence algorithm in satellite internet wireless resource scheduling cannot be further improved. Aiming at the problems that the representation granularity of the satellite internet complex dynamic wireless resource scheduling is insufficient and the artificial intelligence algorithm does not consider the periodic prior information of the satellite network, a new low-earth-orbit satellite wireless resource scheduling method is urgently needed.
Disclosure of Invention
In view of this, embodiments of the present invention provide a method and an apparatus for scheduling low-earth orbit satellite radio resources based on resource mapping and countermeasure learning, so as to eliminate or improve one or more defects in the prior art, and solve the problems that the complex dynamic radio resource scheduling of the satellite internet has insufficient granularity for representation, and the prior information caused by the relatively fixed user distribution and the periodic motion of the satellite is not fully utilized.
One aspect of the invention provides a low-orbit satellite wireless resource scheduling method based on resource map and countermeasure learning, aiming at a jump beam access scene of combining a wide beam with a dynamic spot beam of a low-orbit satellite, the method comprises the following steps:
covering a plurality of user cells by adopting a wide wave beam, covering each user cell by adopting a spot wave beam, distributing satellite resources for granularity according to a set time window, dividing the set time window into a set number of time slots, and distributing time slot resources for each spot wave beam; the users in each user cell obey two-dimensional normal distribution, and the user traffic demand intensity is determined based on the user distribution and the service priority;
based on full frequency multiplexing, averaging the channel difference of a plurality of users served by each point beam, and approximately expressing the channel capacity parameter provided by each user cell by an idealized model of Shannon flux;
setting a priority weighted value for each user cell according to the service priority of each user cell;
according to the user traffic demand strength, channel capacity parameters and priority weight values of each user cell, introducing service priority demands on the basis of maximizing throughput to establish constraints, and solving to obtain the number of time slots allocated to the spot beams corresponding to each user cell;
in a set time window, establishing a three-dimensional histogram with a time slot as an X axis, a spot beam number as a Y axis and a spot beam frequency as a Z axis, cutting the three-dimensional histogram into two-dimensional pictures according to each time slot, wherein a transverse vector of each two-dimensional picture represents the beam number, and a longitudinal quantity of each two-dimensional picture represents the spot beam frequency;
converting the wireless resource constraint condition of the beam hopping satellite communication into a map constraint condition, and carrying out constraint adjustment on the two-dimensional picture;
inputting all two-dimensional images obtained by cutting based on a single three-dimensional histogram into a generator of a pre-trained generation countermeasure network, outputting to obtain an optimized resource allocation diagram, stacking the resource allocation diagrams according to corresponding time slots to obtain a resource scheduling histogram, and performing time slot resource scheduling on each point beam according to the resource scheduling histogram. In some embodiments, based on full frequency reuse, the channel difference of multiple users served by each spot beam is averaged, and the channel capacity parameter provided by each user cell is approximately expressed by an idealized model of shannon flux, which is calculated as:
wherein,
is the actual allocated traffic volume of the wave position corresponding to the nth spot beam, i.e. the channel capacity parameter, N max The number of spot beams working for each time slot in the beam hopping system; w is the total length of the set time window; n is a radical of hydrogen i The number of time slots allocated to the ith spot beam, B is the bandwidth of the beam hopping system, log 2 (1+SINR ij ) For the ith spot beam spectral efficiency, SINR in the jth time slot ij Is the signal-to-noise ratio of the ith spot beam in the jth slot.
In some embodiments, introducing a traffic priority requirement establishment constraint based on maximizing throughput according to the user traffic requirement strength, channel capacity parameter and priority weight value of each user cell comprises:
establishing a maximum throughput objective function, wherein the expression is as follows:
establishing a priority weight target function, wherein the expression is as follows:
wherein,
intensity of user traffic demand representative of wave bit corresponding to nth spot beam, based on the intensity of the user traffic demand in the signal field>
Is the actual allocated traffic volume, N, of the nth spot beam corresponding to the wave position max The number of spot beams working for each time slot in the beam hopping system; w is the total length of the set time window; n is a radical of i Number of slots, N, allocated to ith spot beam * Represents a positive integer.
In some embodiments, the beam hopping satellite communication wireless resource constraint conditional expression is as follows:
wherein N is i The number of time slots allocated to the ith point beam is represented, and W represents the set time window;
representing the spatial separation between spot beams, r representing the radius of a spot beam, B i Denotes the ith spot beam, B j Represents the jth spot beam; o is i Representing the power, P, allocated to the ith spot beam total Representing the total power;
Representing spot beam frequency resources, p w Representing the overall frequency resources of the system.
In some embodiments, the expression of the atlas constraint is:
wherein L is num Number of pictures input, L, representing settings in the GAN network vec-x Representing the horizontal vector of the picture input, L vec-y Represents the longitudinal amount of the inputted picture, W represents the set time window, B i Which represents the (i) th spot beam,
representing the frequency parameter of the ith spot beam. />
In some embodiments, the pre-training step of generating the countermeasure network comprises:
acquiring a training sample set, wherein the training sample set comprises a plurality of samples, each sample comprises a plurality of two-dimensional pictures obtained by cutting a three-dimensional bar graph according to each time slot, the three-dimensional bar graph is established by taking the existing beam hopping wireless resource scheduling data as an X axis according to the time slot, taking the spot beam number as a Y axis and taking the spot beam frequency as a Z axis, the transverse vector of the two-dimensional picture represents the beam number, and the longitudinal quantity of the two-dimensional picture represents the spot beam frequency;
constructing and initializing a generator network and a discriminator network, and training the generator network and the discriminator network by adopting the training sample set;
the generator network performs update iteration using a minimized first loss function, the expression of which is:
wherein z is i Representing the input of the generator network, G (z) i ) Representing the output of the generator network, D (G (z) i ) Represents the output of the arbiter network;
the arbiter network performs update iteration by maximizing a second loss function and minimizing an output, where an expression of the second loss function is:
wherein x is i Representing an input to the network of discriminators,
representing the output of the generator network.
In some embodiments, the generator network and the discriminator network employ a gradient descent method for parameter updating.
In some embodiments, the generator network and the discriminator network employ Adam optimizers for parameter updates during pre-training.
In another aspect, the present invention further provides a low-earth orbit satellite radio resource scheduling apparatus based on resource mapping and countermeasure learning, which includes a processor and a memory, wherein the memory stores computer instructions, and the processor is configured to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the apparatus implements the steps of the above method.
In another aspect, the present invention also provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the above method.
The beneficial effects of the invention at least comprise:
the invention relates to a low earth orbit satellite wireless resource scheduling method and a device based on resource map and countermeasure learning, which are based on a beam hopping technology application scene, firstly, utilize periodic prior information of a satellite network to establish constraint conditions based on maximum throughput and priority requirements, and preliminarily solve the time slot number allocated to each point beam; the method comprises the steps of establishing a three-dimensional histogram of wireless network resource scheduling by taking time slots, spot beam numbers and spot beam frequencies as parameters, cutting the three-dimensional histogram into two-dimensional pictures according to time slots to construct an infinite resource map, optimizing the whole time slot allocation scheduling of the two-dimensional pictures obtained by cutting through a pre-training generation countermeasure network, utilizing the advantages of images in a continuous and dominant color system processing process, utilizing the processing of the wireless resource map to realize fine-grained optimization, and improving the accuracy of a resource allocation strategy.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and drawings.
It will be appreciated by those skilled in the art that the objects and advantages that can be achieved with the present invention are not limited to what has been particularly described hereinabove, and that the above and other objects that can be achieved with the present invention will be more clearly understood from the following detailed description.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
fig. 1 is a flowchart illustrating a method for scheduling low earth orbit satellite radio resources based on resource mapping and countermeasure learning according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a three-dimensional column constructed in a low earth orbit satellite radio resource scheduling method based on resource mapping and countermeasure learning according to an embodiment of the present invention.
Fig. 3 is a diagram illustrating the slicing of fig. 2 according to time slots.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the following embodiments and accompanying drawings. The exemplary embodiments and descriptions of the present invention are provided to explain the present invention, but not to limit the present invention.
It should be noted that, in order to avoid obscuring the present invention with unnecessary details, only the structures and/or processing steps closely related to the solution according to the present invention are shown in the drawings, and other details not so related to the present invention are omitted.
It should be emphasized that the term "comprises/comprising" when used herein, is taken to specify the presence of stated features, elements, steps or components, but does not preclude the presence or addition of one or more other features, elements, steps or components.
It is also noted herein that the term "coupled," if not specifically stated, may refer herein to not only a direct connection, but also an indirect connection in which an intermediate is present.
A beam-hopping satellite communication system has great technical advantages in satellite internet applications. In order to improve the utilization rate of satellite resources, the existing wireless resource management scheme is designed under a certain constraint condition for a single service, so that the requirement of reliable communication service under dynamic and variable satellite internet environments is difficult to adapt, and the research on wireless resource management under the condition of multiple services and multiple constraints is urgently needed. Therefore, the invention mainly solves the problem of how to consider wireless resource scheduling from multi-target joint optimization scheduling of time, space, frequency and power on the premise of rapid change of user requirements and how to continuously consider restrictive indexes such as fairness, service weight and the like under the condition of taking the maximum throughput as an index, thereby further improving the utilization rate of satellite resources and the system performance. The main problems include the following two points:
first, the difficulty of the complex satellite internet radio resource model under high dynamics. The low-orbit satellite has high movement speed and constantly-changed coverage area, allocated wireless resources corresponding to different moments of each user are also changed at a high speed, and meanwhile, the requirements of the wireless resources are expanded from two dimensions of the ground to three dimensions of the air, the sky and the ground in consideration of air-based platforms such as airplanes and airships and space-based platforms such as remote sensing, meteorology and space stations. The existing wireless communication resources are modeled by adopting single-point, fixed and two-dimensional ideas, and have the defects of difficult adaptation to the requirements of accurate description on spatial multipoint, fine granularity change, three-dimensional distribution and the like of wireless communication resources in a satellite internet scene in the aspects of multi-satellite cooperative work, combined networking, different service requirements and the like in a medium-low orbit satellite constellation system, and difficult satisfaction to the requirement of multi-satellite resource combined optimization scheduling.
Secondly, under the condition of large difference of service quality guarantee requirements, the wireless resource global steady state and local fast combined efficient scheduling is difficult. Under the guarantee of service quality, different service scenes have great difference requirements, artificial intelligence is introduced into a resource scheduling algorithm to realize fuzzy and fine-grained wireless resource scheduling which cannot be solved by a determined model algorithm, but the artificial intelligence wireless resource scheduling algorithm comprises artificial intelligence wireless resource scheduling algorithms such as deep learning, reinforcement learning and transfer learning, focuses on link state sensing to carry out optimization control on wireless resources, emphasizes reliable connection, and does not fully utilize prior information brought by relative fixed user distribution and periodic satellite motion, so that the efficiency of the artificial intelligence algorithm in the wireless resource scheduling of the satellite internet cannot be further improved.
Therefore, the resource division of the time-space-frequency-machine multi-dimensional satellite resource model can more accurately meet the service requirement of the cell by constructing the time-space-frequency-machine multi-dimensional satellite resource model, and the generation countermeasure network is utilized to optimize, so that the current satellite service scene and the cell service requirement can be autonomously identified, and efficient resource scheduling is performed. Furthermore, an intelligent resource scheduling algorithm based on countermeasure learning is introduced, normal global scheduling is constructed by utilizing the relativity of user distribution and the periodicity of satellite motion, a wireless resource model is converted into an infinite resource map, and a countermeasure learning network (GAN network) is utilized for processing, so that the resource scheduling requirement of a medium-low orbit satellite constellation which is characterized by multi-satellite cooperation, combined networking and massive users can be self-adapted, and processing is carried out with finer granularity.
Specifically, the invention provides a low-orbit satellite wireless resource scheduling method based on resource map and antagonistic learning, aiming at a beam hopping access scene of combining a wide beam with a dynamic spot beam of a low-orbit satellite, as shown in fig. 1, the method comprises the following steps of S101 to S107:
step S101: covering a plurality of user cells by adopting a wide wave beam, covering each user cell by adopting a spot wave beam, distributing satellite resources for granularity according to a set time window, dividing the set time window into a set number of time slots, and distributing time slot resources for each spot wave beam; and the users in each user cell are subjected to two-dimensional normal distribution, and the user traffic demand intensity is determined based on the user distribution and the service priority.
Step S102: based on full frequency multiplexing, averaging the channel difference of multiple users served by each spot beam, and approximately expressing the channel capacity parameter provided by each user cell by an idealized model of Shannon flux.
Step S103: and setting a priority weighted value for each user cell according to the service priority of each user cell.
Step S104: and introducing service priority requirements to establish constraint on the basis of maximizing throughput according to the user traffic demand strength, the channel capacity parameters and the priority weight value of each user cell, and solving to obtain the number of time slots distributed to the spot beams corresponding to each user cell.
Step S105: in a set time window, a time slot is taken as an X axis, a spot beam number is taken as a Y axis, the spot beam frequency is taken as a Z axis to establish a three-dimensional bar graph, the three-dimensional bar graph is cut into two-dimensional pictures according to each time slot, the transverse vector of each two-dimensional picture represents the beam number, and the longitudinal quantity of each two-dimensional picture represents the spot beam frequency.
Step S106: and converting the beam hopping satellite communication wireless resource constraint condition into an atlas constraint condition, and carrying out constraint adjustment on the two-dimensional picture.
Step S107: inputting all two-dimensional images obtained by cutting based on a single three-dimensional histogram into a generator for generating a pre-trained countermeasure network, outputting to obtain an optimized resource distribution diagram, stacking each resource distribution diagram according to a corresponding time slot to obtain a resource scheduling histogram, and performing time slot resource scheduling on each point beam according to the resource scheduling histogram.
In steps S101 to S104, a jump beam access scheme system communication model combining the wide beam of the low-orbit satellite and the dynamic spot beam is established, and a technical scheme that the wide beam and the spot beam are matched with each other and the wide beam assists in guiding the switching of the jump beam is adopted. And establishing a system model based on the prior information of the access scheme, the working frequency of the spot beam and the user distribution characteristics. In the beam hopping technology, a beam hopping period is a second-level range, spatial distribution of service requests in a coverage area can be regarded as static invariance in the hopping period, and based on the consideration, the method and the device mainly aim at scheduling on-satellite resources in one beam hopping period. Specifically, one adjustment variable period may be determined as a set time window, and a plurality of time slots are further divided in the set time window to serve as the maximum allocation unit for resource scheduling. The service block covered by the wide wave beam comprises a plurality of user cells, each user cell is covered by the spot wave beam to carry out resource allocation, and in a time window, the resource allocation is carried out by allocating corresponding time slot numbers for the spot wave beams.
And constructing a system communication model of the beam hopping access scheme based on prior information, wherein the main points are service requirements in each user cell and resources which can be provided by each point beam. In each service block, the positions of users are randomly distributed and obey the characteristic of two-dimensional normal distribution, the user service demand intensity of each user cell is related to the number of users, the service priority is synchronously introduced in the application to determine the user service demand intensity, the related service types of the specified user cells can be predicted, the corresponding service priority can be set for each user according to the known service types, and the user service demand intensity of each user cell is calculated.
In step S101, for the jth user served by the ith spot beam, the number of users in each user cell is q i J =1,2,3 \ 8230q i User u ij Can be expressed as
Wherein +>
In contrast, the probability density function is:
user u ij Position of
Obey mean value of mu 1 And mu 2 Variance is σ 1 And σ 2 ρ =0, and the variance describes the degree of dispersion of users in the cell. At the introduction of traffic priority w i In this case, the intensity of the user traffic demand can be expressed as:
further, in step S102, each point beam adopts a full frequency reuse mode, so that in each time slot, the channel capacity provided by the satellite to each cell can be approximated by an ideal shannon capacity. In some embodiments, based on full frequency reuse, the channel difference of multiple users served by each spot beam is averaged, and the channel capacity parameter provided by each user cell is approximately expressed by an idealized model of shannon flux, which is calculated as:
wherein,
is the actual allocated traffic volume of the wave position corresponding to the nth spot beam, i.e. the channel capacity parameter, N max The number of spot beams working for each time slot in the beam hopping system; w is the total length of the set time window; n is a radical of i The number of time slots allocated to the ith spot beam, B is the bandwidth of the beam hopping system, log 2 (1+SINR ij ) For the ith point beam spectrum efficiency in the jth time slot, SINR ij The signal-to-noise ratio of the ith spot beam in the jth slot.
In step S104, according to the user traffic demand strength, the channel capacity parameter, and the priority weight value of each user cell, a traffic priority demand establishment constraint is introduced on the basis of maximizing throughput, which includes:
establishing a maximum throughput objective function, wherein the expression is as follows:
establishing a priority weight target function, wherein the expression is as follows:
wherein,
intensity of user traffic demand representative of wave bit corresponding to nth spot beam, based on the intensity of the user traffic demand in the signal field>
Is the actual allocated traffic volume, N, of the nth spot beam corresponding to the wave position max The number of spot beams working for each time slot in the beam hopping system; w is the total length of the set time window; n is a radical of hydrogen i Number of time slots, N, allocated to ith spot beam * Representing a positive integer.
Until step S104, the number of time slots allocated to each point beam in the set time window is obtained through preliminary solution, but the coarse-grained discrete characteristics are presented, and the solution performed only by the prior information cannot meet the requirement of fine-grained continuous scheduling in the complex communication environment. In steps S105-S107, the scheduling strategy of the preliminary solution is converted into an infinite resource map for optimization, and the advantages of the images in the continuous and dominant color system processing process are utilized to introduce the generation of the countermeasure network for continuous fine-grained processing optimization.
First, in step S105, as shown in fig. 2 and 3, a three-dimensional histogram is established with the time slot as the X axis, the spot beam number as the Y axis, and the spot beam frequency as the Z axis based on the preliminarily obtained resource allocation policy. And is divided into a plurality of two-dimensional pictures according to the time slots. If a set time window comprises K time slots, the three-dimensional histogram is divided into K two-dimensional pictures, each two-dimensional picture corresponds to one time slot, the horizontal vector of each two-dimensional picture represents a beam number, and the vertical quantity of each two-dimensional picture represents the spot beam frequency.
In step S106, a preliminary constraint is performed based on the beam hopping satellite communication radio resource constraint condition, the beam hopping satellite communication radio resource constraint condition is first converted into a map constraint condition, and constraint adjustment is performed on each two-dimensional picture.
In some embodiments, the beam hopping satellite communication wireless resource constraint condition expression is as follows:
wherein N is i The number of time slots allocated to the ith spot beam is represented, and W represents the set time window;
representing the spatial separation between spot beams, r representing the radius of a spot beam, B i Denotes the ith spot beam, B j Represents the jth spot beam; p is i Representing the power, P, allocated to the ith spot beam total Represents the total power;
Representing spot beam frequency resources, p w Presentation systemA body frequency resource. />
Further, the map constraint condition expression obtained by conversion is as follows:
wherein L is num Indicating the number of pictures input, L, set in the GAN network vec-x Representing the horizontal vector of the picture input, L bec-y Representing the longitudinal amount of the picture input, W representing the set time window, B i Which represents the (i) th spot beam,
representing the frequency parameter of the ith spot beam.
In step S107, in order to achieve fine-grained continuous optimization and improve the overall generalization capability of the optimization process, the service requirements of multiple user cells are met. Optimizing all two-dimensional picture integral input values obtained by dividing the same three-dimensional histogram in a generator for generating a countermeasure network in a pre-training mode, outputting an optimized resource allocation diagram, re-stacking to obtain a resource scheduling histogram, and performing time slot resource scheduling on each point beam according to the resource scheduling histogram to realize fine-grained continuous optimization.
Specifically, in some embodiments, the pre-training step of generating the countermeasure network includes steps S201 to S202:
step S201: the method comprises the steps of obtaining a training sample set, wherein the training sample set comprises a plurality of samples, each sample comprises a plurality of two-dimensional pictures obtained by cutting a three-dimensional histogram according to each time slot, the three-dimensional histogram is established by taking the existing beam hopping wireless resource scheduling data as an X axis according to the time slot, the spot beam number as a Y axis and the spot beam frequency as a Z axis, the transverse vector of the two-dimensional picture represents the beam number, and the longitudinal quantity of the two-dimensional picture represents the spot beam frequency.
Step S202: and constructing and initializing a generator network and a discriminator network, and training the generator network and the discriminator network by adopting a training sample set.
The generator network performs update iteration by using a minimized first loss function, the expression of which is:
wherein z is i Representing the input of the generator network, G (z) i ) Representing the output of the generator network, D (G (z) i ) 0 represents the output of the arbiter network.
The arbiter network performs update iteration by using a maximized second loss function and a minimized output, wherein the expression of the second loss function is as follows:
wherein x is i Representing an input to the network of discriminators,
representing the output of the generator network.
In some embodiments, the generator network and the discriminator network employ a gradient descent method for parameter updating.
In some embodiments, the generator network and the discriminator network employ Adam optimizers for parameter updates during the pre-training process.
In another aspect, the present invention further provides a low-earth orbit satellite radio resource scheduling apparatus based on resource mapping and countermeasure learning, which includes a processor and a memory, wherein the memory stores computer instructions, and the processor is configured to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the apparatus implements the steps of the above method.
In another aspect, the present invention also provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the above method.
The invention is illustrated below with reference to specific examples:
in the embodiment, the architecture analysis and problem modeling are carried out on the beam hopping satellite wireless resources, and the change characteristics of the multidimensional resources along with time slots are analyzed. And then integrating multiple application scenes and a multi-target function, continuously considering resource allocation strategies of indexes such as fairness, service weight and the like under the condition that the maximum throughput is used as the index, and solving the target function and analyzing a simulation result. And finally, optimizing by generating a countermeasure network, thereby obtaining a fine-grained accurate resource allocation strategy.
In this embodiment, as shown in fig. 1, the specific process is as follows:
step 1: wireless resource model construction of time-frequency-space-machine satellite internet beam hopping system
And establishing a communication model of a hopping beam access scheme system of the low earth orbit satellite wide beam combined with the dynamic spot beam. The satellite end antenna transmits two beams of a wide beam and a spot beam, and broadband communication of a low-orbit satellite coverage area is completed by adopting a scheme of mutual cooperation of the wide beam and the spot beam, namely beam hopping switching under the auxiliary guidance of the wide beam. A resource scheduling scenario for a satellite communication system is shown. The satellite communication system consists of M multi-beam communication satellites and dynamically covers ground users, and coverage areas of different satellite beams are overlapped to a certain extent. The area range which can be covered by the system is divided into N ground areas, and the communication tasks respectively belong to different ground areas according to the positions of the initiating users.
With the beam hopping technique, the beam hopping period is typically in the order of seconds. Thus, the spatial distribution of traffic requests within the coverage area can be considered as static invariant within a beam hopping period, and based on the above considerations, the present invention mainly describes the on-satellite resource scheduling problem within one beam hopping period. Setting a time window of satellite resource allocation to Tw i Scheduling satellite communication resources with granularity of time window length, and Tw 1 Time window, tw, representing the current beam resource allocation 3 A time window representing the next beam resource scheduling, the start and the end of the time window being ST i And ET i . The time window is further divided into N k Each resource scheduling window is the maximum allocation unit of the task, and each resource scheduling window lasts for K time slots.
Taking a specific beam coverage cell as an example, one wide beam is composed of 9 user cells, and meanwhile, a spot beam performs coverage resource allocation on a single user cell. The space Resource of the satellite system is a beam Resource, the space Resource is divided according to the beam, each beam includes time and frequency resources, that is, each beam includes RB (Resource Block, 12 subcarriers continuous in frequency) in the system range, and the space Resource is allocated in a time window unit according to the demand of the ground user according to an allocation strategy. Thus, the [ 2 ]]Multidimensional joint resource allocation is performed on the satellite communication system. Because the resources are in the time window Tw i In which time slots are divided into time units, so that each time window Tw is set i The total length is W, i.e. each time window comprises W time slots; n is a radical of hydrogen i Is the ith spot beam B i The number of time slots allocated, i.e. N i Is a time slot resource in the resource model; p is a radical of i Representing a frequency parameter. Then the overall resource model E i The mathematical model of (a) is expressed as:
E i ＝f{(p i ，B i ，N i )，i≤i max }；
each wide beam C according to the satellite coverage model i Dividing the satellite spot beam into n areas to correspond to the coverage positions (wave positions) of the satellite spot beams, and defining each spot beam coverage area as a user cell marked as B n . Users u in each user cell ij Where i represents the number of wave bits, j represents the user number, and the number of users in each user cell is assumed to be q i J =1,2,3 \ 8230q i . In each user cell, the user positions are randomly distributed, the users obey two-dimensional normal distribution in the satellite coverage cell area, and a ground cell B is set n In, user u ij Location from its position
To express, in which,
and &>
Mutually opposed, the probability density function is:
wherein, user u ij Position of
Obey mean value of mu 1 And mu 2 Variance is σ 1 And σ 2 ρ =0, and the variance describes the degree of dispersion of users in the cell. For the intensity of the service volume demand of the users in each user cell, the embodiment is only related to the number of users, and the number of users obeys the user distribution, i.e. the two-dimensional normal distribution, in addition, the intensity of the service volume demand of the users is also related to the service priority w i In relation, therefore, the user traffic demand intensity can be expressed as:
and 2, step: differentiated service guarantee constraint-wireless resource scheduling based on multi-objective optimization function
In a time window Tw i In (2), the number of spot beams existing in each satellite is S, and each spot beam is marked as B i . Because full frequency multiplexing is adopted, and the channel difference among different users in the beam is averaged to be used as the channel capacity parameter of the corresponding wave position of the spot beam
In each time slot, for the convenience of modeling, an idealized model approximation of shannon flux is usedRepresenting the channel capacity resource provided by the satellite beam to each terrestrial cell:
wherein,
is the actual allocated traffic volume of the wave position corresponding to the nth spot beam, i.e. the channel capacity parameter, N max The number of spot beams working for each time slot in the beam hopping system; w is the total length of the set time window; n is a radical of hydrogen i The number of time slots allocated to the ith spot beam, B is the bandwidth of the beam hopping system, log 2 (1+SINR ij ) For the ith point beam spectrum efficiency in the jth time slot, SINR ij Is the signal-to-noise ratio of the ith spot beam in the jth slot.
The resource scheduling algorithm of this embodiment takes the maximum throughput of the system as an optimization target, and establishes the following maximum throughput objective function:
when the service volume of a user in a ground cell is small, but under the emergency conditions, such as emergency communication, burst service and the like, if the maximum throughput of a system is taken as a principle, less resources are allocated and the service requirement of the ground cell cannot be met, so a weight objective function is introduced, on the basis of taking the maximum throughput of the system, the service of the user is compared to determine the priority of each service, and different weight values w are set i (the weight value is set according to specific analysis of specific service scenes) to meet the special requirements of the ground users. Therefore, the method needs to be based on system scenes and business requirementsBy combining the system and the constraint conditions mentioned above, the corresponding priority weight objective functions are listed as follows:
wherein,
represents the intensity of the user traffic demand on the wave bit corresponding to the nth spot beam, and/or>
Is the amount of traffic actually allocated to the wave bit for the nth spot beam, N max The number of spot beams working for each time slot in the beam hopping system; n is the total length of the set time window; n is a radical of i Number of slots, N, allocated to ith spot beam * Representing a positive integer.
And 3, step 3: method for establishing wireless resource map of hopping beam satellite communication
Setting up in 6G internet satellite covered cell B n In (2), the number of hop beams existing in a single satellite is S. Each wave beam adopts a full frequency multiplexing mode as a frequency parameter P of the wave beam i In each time slot, for ease of modeling, the channel capacity provided by the satellite to each cell is approximated by the ideal shannon capacity:
where B is the bandwidth of the beam hopping system, log 2 (1+SINR ij ) For spectral efficiency, SINR ij The signal-to-noise ratio of the ith cell wave bit in the jth time slot. The division of each beam is the division of space resource, so each beam is usedThe covered ground cell labels to distinguish the beams, beam C i Are spatial resources in the resource model.
When the resource allocation problem focuses on a time window Tw i The wave position B corresponding to the middle point wave beam n The satellite acquires the service request traffic of each cell
And channel allocation capacity->
Allocating a beam slot N on the Y-axis i Size, for which resource allocation problems can be graphically characterized in a time window: as shown in FIG. 2, a three-dimensional histogram w is built mod Each wave position B n Time slot resource N of wave position distinguished by different cylinders i Then it is represented by the width of the column, frequency resource P i The height of the column is shown, and the space resource is shown among the beams.
And 4, step 4: mapping of hopping beam satellite communication radio resource constraints and map processing constraints
The three-dimensional hopping beam satellite communication wireless resource model w can be obtained by carrying out mapping processing on the satellite on-satellite resource model mod The beam hopping technology optimizes the satellite power and bandwidth resources from the time domain by using the time slicing technology, allocates the whole satellite on-satellite resources to each beam by taking a time slot as a unit, and can generate a countermeasure network by taking a picture as input, so that the countermeasure network can be generated according to the unit time slot N i For three-dimensional resource model w mod And cutting the two-dimensional picture as input, on the basis of which, the mapping relation exists between the wireless resource constraint of the beam hopping satellite communication and the map processing constraint, and the two-dimensional picture is constrained and adjusted on the basis of the map constraint condition obtained after mapping.
The constraint condition expression of the beam hopping satellite communication wireless resource is as follows:
wherein N is i The number of time slots allocated to the ith spot beam is represented, and W represents the set time window;
representing the spatial separation between spot beams, r represents the radius of a spot beam, B i Denotes the ith spot beam, B j Represents the jth spot beam; p is i Denotes the power, P, allocated to the ith spot beam total Representing the total power;
Representing spot beam frequency resources, p w Representing the overall frequency resources of the system.
The expression of the map constraint condition is as follows:
wherein L is num Indicating the number of pictures input, L, set in the GAN network vec-x Representing the horizontal vector of the picture input, L vec- Representing the longitudinal amount of the picture input, W representing the set time window, B i Which represents the (i) th spot beam,
representing the frequency parameter of the ith spot beam. In this embodiment, the time window W is set to 16 time slots, and the resource scheduling graph is divided according to unit time slots, so C4 indicates that the number of input picture sets is 16; the spot beams are divided according to the number of the coverage of the ground cell, and N =16, so that C5 represents that the horizontal vector of the input picture is 16 units; and because of adopting full frequency reuse, the frequency resource P is divided i Dividing into 16 frequency bands, so converting into picture constraint as: c6 denotes that the vertical vector of the input picture is 16 units.
And 5: wireless resource map optimal processing based on counterstudy
By matching the three-dimensional model map w mod According to time slot N i To perform two-dimensional picture slicing, which means to divide each cell B in each unit time slot n The resource scheduling algorithm takes the obtained two-dimensional picture as the input of the GAN network, the resource allocation map of the GAN network after optimizing each time slot can be obtained after the GAN network is optimized, and the two-dimensional picture is piled up through Matlab to obtain an optimized time window Tw i Within the resource scheduling histogram.
The pre-training step of the GAN network comprises the following steps:
acquiring a training sample set, wherein the training sample set comprises a plurality of samples, each sample comprises a plurality of two-dimensional pictures obtained by cutting a three-dimensional histogram according to each time slot, the three-dimensional histogram is established by taking the existing beam hopping wireless resource scheduling data as an X axis according to the time slot, the spot beam number as a Y axis and the spot beam frequency as a Z axis, the transverse vector of the two-dimensional picture represents the beam number, and the longitudinal quantity of the two-dimensional picture represents the spot beam frequency;
initializing a satellite beam hopping resource allocation scenario, initializing a discriminator network (D) parameter (theta) d for D), initializing the generator network (G) parameters
Wherein the training of the discriminator network comprises:
1) And extracting a training set from the existing data, specifically, calculating a resource distribution diagram in a time window according to the selected target function and the constraint conditions thereof, and cutting a two-dimensional picture to obtain the training set. Two-dimensional picture in each sample { x (1) ，x (2) ，x (3) 8230The is used as the inherent parameter index of the discriminator network.
2) Random noise of the same dimension is taken as input to the generator network, and the picture is in the form of z (1) ，z (2) ，z (3) …}
3) Deriving data from the output of a generator network
Input to the D-network.
4) Updating a parameter θ of a discriminator network d To maximize
The larger the better, and->
The smaller the better, the better the training is carried out by using a loss function and a gradient descent method, and the expression is as follows:
the gradient descent method updates the network parameter expression of the discriminator as follows:
training of the generator network includes:
1) Extracting a training set from the existing data, specifically, calculating a resource distribution diagram in a time window according to a selected target function and constraint conditions thereof, and carrying out two-dimensional picture segmentation to obtain the training set which is used as the input of a generator network, wherein the picture is in a form of { z (1) ，z (2) ，z (3) …}
2) Updating parameters of G-network
Minimizing the loss function, training by using a loss function and a gradient descent method, and expressing the following expression:
the gradient descent method updates the generator network parameter expression as follows:
jointly solving to obtain N i I.e. each wave position B n Number of allocated time slots, so N i That is, the parameters for generating the countermeasure network are adjusted to make the generated probability distribution and the real data distribution as close as possible, and the loss function of the GAN network is designed as follows:
wherein,
and &>
Representing the probability of the real data and the generated data.
By adjusting the network parameter theta of the GAN network d And
so that the value of the loss function L (G, D) becomes smaller and smaller, and the robustness of the whole algorithm is better, and then the predicted value which generates the antagonistic network is calculated by the loss function formula>
And the actual value N i The difference between the two parameters is used for updating all parameters theta for generating the countermeasure network through back propagation d And &>
Thereby reducing the difference between the actual value and the predicted value such that the predicted value generated by the GAN network->
Real time slot allocation N i And then optimizing the whole resource allocation strategy.
Step 6: conversion of radio resource map optimal processing and beam hopping satellite communication radio resource scheduling
Obtaining each time slot N through optimization of GAN network i The optimal map within. These optimal maps-N per time slot i The resource allocation map carries out the stacking of two-dimensional pictures, and then a time window Tw optimized by the GAN network is obtained i The input map is two-dimensionally cut according to unit time slot, so that the obtained optimal map is divided according to time slot N i And performing two-dimensional tiling to obtain a resource optimal allocation map. And the two-dimensional top view with the optimal resources and the top view before optimization are subjected to time slot analysis to determine the wave beam time slot N for each cell for generating the countermeasure network i Degree of optimization of (2), time slot N i The optimized wave position B can be obtained by combining with the wave position resource traffic formula n Actual allocated traffic
And then, the generation of the countermeasure network is converted from map optimization to resource scheduling optimization, and the problems of continuous change and various service types of low-orbit 6G satellite internet service scenes can be solved.
The embodiment can obviously improve the utilization rate of the satellite system resources in a time window. And the satisfaction degree of users in each cell is balanced, so that the method is suitable for the changeable service scenes of the satellite.
In summary, the method and the device for scheduling the low earth orbit satellite wireless resources based on the resource map and the countermeasure learning of the invention are based on the application scene of the beam hopping technology, firstly, the periodic prior information of the satellite network is utilized to establish constraint conditions based on the maximum throughput and the priority requirement, and the time slot number allocated to each point beam is preliminarily solved; the method comprises the steps of establishing a three-dimensional histogram of wireless network resource scheduling by taking time slots, spot beam numbers and spot beam frequencies as parameters, cutting the three-dimensional histogram into two-dimensional pictures according to time slots to construct an infinite resource map, optimizing the whole time slot allocation scheduling of the two-dimensional pictures obtained by cutting through a pre-training generation countermeasure network, utilizing the advantages of images in a continuous and dominant color system processing process, utilizing the processing of the wireless resource map to realize fine-grained optimization, and improving the accuracy of a resource allocation strategy.
In accordance with the above method, the present invention also provides an apparatus/system comprising a computer device including a processor and a memory, the memory having stored therein computer instructions, the processor being configured to execute the computer instructions stored in the memory, the apparatus/system implementing the steps of the method as described above when the computer instructions are executed by the processor.
Embodiments of the present invention further provide a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to implement the foregoing steps of the edge computing server deployment method. The computer readable storage medium may be a tangible storage medium such as Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, floppy disks, hard disks, removable storage disks, CD-ROMs, or any other form of storage medium known in the art.
Those of ordinary skill in the art will appreciate that the various illustrative components, systems, and methods described in connection with the embodiments disclosed herein may be implemented as hardware, software, or combinations of both. Whether this is done in hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention. When implemented in hardware, it may be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, plug-in, function card, or the like. When implemented in software, the elements of the invention are the programs or code segments used to perform the required tasks. The program or code segments can be stored in a machine-readable medium or transmitted by a data signal carried in a carrier wave over a transmission medium or a communication link.
It is to be understood that the invention is not limited to the specific arrangements and instrumentality described above and shown in the drawings. A detailed description of known methods is omitted herein for the sake of brevity. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present invention are not limited to the specific steps described and illustrated, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps, after comprehending the spirit of the present invention.
Features that are described and/or illustrated with respect to one embodiment may be used in the same way or in a similar way in one or more other embodiments and/or in combination with or instead of the features of the other embodiments in the present invention.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made to the embodiment of the present invention by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A low orbit satellite wireless resource scheduling method based on resource map and countermeasure learning is characterized in that aiming at a jump beam access scene of combining a wide beam with a dynamic spot beam of a low orbit satellite, the method comprises the following steps:
covering a plurality of user cells by adopting a wide wave beam, covering each user cell by adopting a spot wave beam, distributing satellite resources for granularity according to a set time window, dividing the set time window into a set number of time slots, and distributing time slot resources for each spot wave beam; the users in each user cell obey two-dimensional normal distribution, and the user traffic demand intensity is determined based on the user distribution and the service priority;
based on full frequency multiplexing, averaging the channel difference of a plurality of users served by each point beam, and approximately expressing the channel capacity parameter provided by each user cell by an idealized model of Shannon flux;
setting a priority weighted value for each user cell according to the service priority of each user cell;
according to the user traffic demand strength, channel capacity parameters and priority weight values of each user cell, introducing service priority demands on the basis of maximizing throughput to establish constraints, and solving to obtain the number of time slots allocated to the spot beams corresponding to each user cell;
in a set time window, establishing a three-dimensional bar graph by taking a time slot as an X axis, a spot beam number as a Y axis and a spot beam frequency as a Z axis, cutting the three-dimensional bar graph into two-dimensional pictures according to each time slot, wherein the transverse vector of each two-dimensional picture represents the beam number, and the longitudinal quantity of each two-dimensional picture represents the spot beam frequency;
converting the wireless resource constraint condition of the beam hopping satellite communication into a map constraint condition, and carrying out constraint adjustment on the two-dimensional picture;
inputting all two-dimensional images obtained by cutting based on a single three-dimensional histogram into a generator for generating a pre-trained countermeasure network, outputting to obtain an optimized resource distribution diagram, stacking each resource distribution diagram according to a corresponding time slot to obtain a resource scheduling histogram, and performing time slot resource scheduling on each point beam according to the resource scheduling histogram.
2. The method of claim 1, wherein the channel differences of the users served by each spot beam are averaged based on full frequency reuse, and the channel capacity parameter provided by each user cell is approximately expressed by an idealized model of shannon throughput, and the calculation formula is as follows:
wherein,
is the actual allocated traffic volume of the wave position corresponding to the nth spot beam, i.e. the channel capacity parameter, N max The number of spot beams working for each time slot in the beam hopping system; w is the total length of the set time window; n is a radical of i The number of time slots allocated to the ith spot beam, B is the bandwidth of the beam hopping system, log 2 (1+SINR ij ) For the ith spot beam spectral efficiency, SINR in the jth time slot ij The signal-to-noise ratio of the ith spot beam in the jth slot.
3. The method of claim 2, wherein introducing traffic priority requirements establishing constraints based on maximizing throughput according to user traffic demand strength, channel capacity parameters and priority weight values of each user cell comprises:
establishing a maximum throughput objective function, wherein the expression is as follows:
establishing a priority weight target function, wherein the expression is as follows:
wherein,
intensity of user traffic demand representative of wave bit corresponding to nth spot beam, based on the intensity of the user traffic demand in the signal field>
Is the actual allocated traffic volume, N, of the nth spot beam corresponding to the wave position max The number of spot beams working for each time slot in the beam hopping system; w is the total length of the set time window; n is a radical of i Number of slots, N, allocated to ith spot beam * Representing a positive integer.
4. The resource map and countermeasure learning based low orbit satellite radio resource scheduling method of claim 1, wherein the beam hopping satellite communication radio resource constraint condition expression is:
wherein N is i The number of time slots allocated to the ith spot beam is represented, and W represents the set time window;
representing the spatial separation between spot beams, r represents the radius of a spot beam, B i Denotes the ith spot beam, B j Represents the jth spot beam; p i Representing the power, P, allocated to the ith spot beam total Representing the total power;
Representing spot beam frequency resources, p w Representing the overall frequency resources of the system.
5. The resource map and countermeasure learning-based low-earth orbit satellite radio resource scheduling method according to claim 4, wherein the expression of the map constraint condition is:
wherein L is num Number of pictures input, L, representing settings in the GAN network vec-x Representing the horizontal vector of the picture input, L vec-y Representing the longitudinal amount of the picture input, W representing the set time window, B i Which represents the (i) th spot beam,
representing the frequency parameter of the ith spot beam.
6. The method for low-earth orbit satellite radio resource scheduling based on resource mapping and countermeasure learning of claim 5, wherein the pre-training step for generating the countermeasure network comprises:
acquiring a training sample set, wherein the training sample set comprises a plurality of samples, each sample comprises a plurality of two-dimensional pictures obtained by cutting a three-dimensional histogram according to each time slot, the three-dimensional histogram is established by taking the existing beam hopping wireless resource scheduling data as an X axis according to the time slot, the spot beam number as a Y axis and the spot beam frequency as a Z axis, the transverse vector of the two-dimensional picture represents the beam number, and the longitudinal quantity of the two-dimensional picture represents the spot beam frequency;
constructing and initializing a generator network and a discriminator network, and training the generator network and the discriminator network by adopting the training sample set;
the generator network performs update iteration using a minimized first loss function, the expression of which is:
wherein z is i Representing the output of the generator networkIn, G (z) i ) Representing the output of the generator network, D (G (z) i ) Represents the output of the discriminator network;
the arbiter network performs update iteration by maximizing a second loss function and minimizing an output, where an expression of the second loss function is:
wherein x is i Representing an input to the network of discriminators,
representing the output of the generator network.
7. The method according to claim 6, wherein the generator network and the discriminator network perform parameter updating by using a gradient descent method.
8. The method for low-earth orbit satellite radio resource scheduling based on resource mapping and antagonistic learning of claim 7, wherein the generator network and the discriminator network are updated with parameters by using an Adam optimizer in a pre-training process.
9. An apparatus for scheduling low-earth orbit satellite radio resources based on resource mapping and countermeasure learning, comprising a processor and a memory, wherein the memory stores computer instructions, the processor is configured to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the apparatus implements the steps of the method according to any one of claims 1 to 8.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211356622.9A CN115942460A (en) | 2022-11-01 | 2022-11-01 | Low-orbit satellite wireless resource scheduling method and device based on resource map and countermeasure learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211356622.9A CN115942460A (en) | 2022-11-01 | 2022-11-01 | Low-orbit satellite wireless resource scheduling method and device based on resource map and countermeasure learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115942460A true CN115942460A (en) | 2023-04-07 |
Family
ID=86653359
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211356622.9A Pending CN115942460A (en) | 2022-11-01 | 2022-11-01 | Low-orbit satellite wireless resource scheduling method and device based on resource map and countermeasure learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115942460A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116567648A (en) * | 2023-07-05 | 2023-08-08 | 四川创智联恒科技有限公司 | Method and system for generating wave bitmap spectrum under data wave beam |
| CN117811645A (en) * | 2024-03-01 | 2024-04-02 | 南京控维通信科技有限公司 | Satellite frequency resource allocation and utilization rate calculation method |
-
2022
- 2022-11-01 CN CN202211356622.9A patent/CN115942460A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116567648A (en) * | 2023-07-05 | 2023-08-08 | 四川创智联恒科技有限公司 | Method and system for generating wave bitmap spectrum under data wave beam |
| CN116567648B (en) * | 2023-07-05 | 2023-09-19 | 四川创智联恒科技有限公司 | Method and system for generating wave bitmap spectrum under data wave beam |
| CN117811645A (en) * | 2024-03-01 | 2024-04-02 | 南京控维通信科技有限公司 | Satellite frequency resource allocation and utilization rate calculation method |
| CN117811645B (en) * | 2024-03-01 | 2024-05-31 | 南京控维通信科技有限公司 | Satellite frequency resource allocation and utilization rate calculation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115942460A (en) | Low-orbit satellite wireless resource scheduling method and device based on resource map and countermeasure learning | |
| CN114048689B (en) | Multi-unmanned aerial vehicle aerial charging and task scheduling method based on deep reinforcement learning | |
| CN110518956B (en) | Time slot allocation algorithm-based hopping pattern optimization method and device, and storage medium | |
| CN114362810A (en) | Low-orbit satellite beam hopping optimization method based on migration depth reinforcement learning | |
| CN113162682A (en) | PD-NOMA-based multi-beam LEO satellite system resource allocation method | |
| CN115474281B (en) | Resource allocation method in star-ground fusion network | |
| CN111555795A (en) | Network resource allocation method and device based on user grouping scheduling | |
| CN112583453A (en) | Downlink NOMA power distribution method of multi-beam LEO satellite communication system | |
| Kisseleff et al. | A new optimization tool for mega-constellation design and its application to trunking systems | |
| CN113259946A (en) | Ground-to-air full coverage power control and protocol design method based on centralized array antenna | |
| Li et al. | Spectrum allocation with asymmetric monopoly model for multibeam-based cognitive satellite networks | |
| CN116505998A (en) | Multi-beam satellite communication resource distribution system and method based on deep reinforcement learning | |
| CN117614520B (en) | Method for optimizing large-scale MIMO (multiple input multiple output) resources by removing cells based on unmanned aerial vehicle-satellite cooperation | |
| Lee et al. | Optimal frequency reuse and power control in multi-UAV wireless networks: Hierarchical multi-agent reinforcement learning perspective | |
| Wei et al. | Dynamic beam scheduling of multibeam low earth orbit satellites based on an enhanced artificial bee colony algorithm | |
| CN116546624B (en) | Method and device for predicting wave-hopping satellite service and distributing multidimensional link dynamic resources | |
| CN117295090A (en) | Resource allocation method for Unmanned Aerial Vehicle (UAV) through-sense integrated system | |
| CN115694612A (en) | Low-orbit satellite communication resource scheduling system and method based on phased array antenna | |
| EP4331299A1 (en) | Determing allocation of unmanned aerial vehicle base stations in a wireless network | |
| CN108632767B (en) | The transmission method and device of heterogeneous network collaboration multicast | |
| CN116248163B (en) | Frequency compatibility analysis method for large-scale low-orbit constellation | |
| CN108449119B (en) | The transmission method and device of heterogeneous network time-division collaboration multicast | |
| CN114710195B (en) | Low-orbit satellite energy-efficient resource allocation method based on beam hopping technology | |
| CN118524069B (en) | Large-scale satellite network service level general calculation cooperative resource allocation device and method | |
| Deng et al. | The investigation of resource allocation on heterogeneous space-based networks based on sdn framework |
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 |