WO2020134145A1

WO2020134145A1 - Time-frequency null resource allocation method, computer apparatus, and computer-readable storage medium

Info

Publication number: WO2020134145A1
Application number: PCT/CN2019/103384
Authority: WO
Inventors: 詹勇
Original assignee: 中兴通讯股份有限公司
Priority date: 2018-12-26
Filing date: 2019-08-29
Publication date: 2020-07-02
Also published as: CN111372312A

Abstract

The present application relates to a time-frequency null resource allocation method, a computer apparatus, and a computer-readable storage medium, The time-frequency null resource allocation method comprises: acquiring a user to be scheduled (S101); determining, according to a pre-set algorithm, a pairing policy for the user to be scheduled (S102); acquiring an evaluation mechanism for the pairing policy (S103); and obtaining, according to the evaluation mechanism, an optimal time-frequency resource allocation method corresponding to the pairing policy, and scoring the pairing policy (S104).

Description

Time-frequency space resource allocation method, computer device and computer readable storage medium

cross reference

This application refers to the Chinese patent application No. 201811602906.5 filed on December 26, 2018, entitled "Time-frequency-space resource allocation method, computer device, and computer-readable storage medium", which is fully incorporated by reference into this application.

Technical field

The present application relates to the technical field of processing of the upstream/downstream MAC (Media Access Control) layer, in particular to a method for allocating time-frequency space resources, a computer device, and a computer-readable storage medium.

Background technique

The wireless communication system has evolved into the 4G era. By introducing multi-antenna technology and MU-MIMO (Multi-User Multiple Input Multiple Output) system, the wireless SE (Spectrum Efficiency Spectrum Efficiency)/cell throughput has been significantly improved. In the 5G era, multiple antennas and MU-MIMO are still the most effective methods to improve the SE of wireless systems. Among them, the space division user pairing and time-frequency resource allocation method are the key technologies of the MU-MIMO system: the space division user pairing determines which users are composed of each empty group in each TTI (Transmission Time Interval); and The time-frequency resource allocation method determines how much frequency resources each empty packet and other frequency division users occupy in each TTI.

Generally, in a MU-MIMO system, the quality of each user's wireless channel and its service requirements are very different. At the same time, the horizontal/vertical spacing between users is also very different. Therefore, the wireless SE/cell throughput caused by different space division user pairing methods and time-frequency resource allocation methods may vary greatly. Therefore, while considering fairness, how to obtain the optimal space-division user pairing and time-frequency resource allocation strategy to maximize wireless SE/cell throughput is a significant and urgent problem to be solved.

Summary of the invention

In order to solve the above technical problem or at least partially solve the above technical problem, the present application provides a time-frequency space resource allocation method, a computer device, and a computer-readable storage medium.

In view of this, in the first aspect, embodiments of the present application provide a method for allocating space-time resources, including: acquiring users to be scheduled; determining a matching strategy for users to be scheduled according to a preset algorithm; obtaining an evaluation mechanism for the matching strategy; The mechanism obtains the optimal time-frequency resource allocation method corresponding to the pairing strategy and scores the pairing strategy.

The time-frequency-space resource allocation method provided in this application first obtains the user to be scheduled, and then uses a preset algorithm to determine the matching strategy of the user to be scheduled; further, after determining the matching strategy of the user to be scheduled, the evaluation mechanism of the matching strategy is obtained, and then based on The evaluation mechanism obtains the optimal time-frequency resource allocation method corresponding to the matching strategy and scores the matching strategy. Specifically, the optimal time-frequency resource allocation method is a time-frequency resource allocation method that can maximize the wireless SE and cell throughput. It is worth noting that this application uses a preset algorithm to determine the pairing strategy of the users to be scheduled, and evaluates the optimal time-frequency resource allocation method corresponding to the pairing strategy through the evaluation mechanism. The two cooperate with each other to determine the optimal air separation user Pairing and time-frequency resource allocation strategies maximize wireless SE (Spectrum Efficiency)/cell throughput while considering fairness.

According to the above-mentioned time-frequency space resource allocation method of the present application, it may also have the following additional technical features.

In the above technical solution, the step of obtaining users to be scheduled specifically includes: determining the target number of users to be scheduled according to the number of users that can be scheduled at the current transmission time interval and the number of active users at the current transmission time interval; according to preset screening rules Filter out the target number of users to be scheduled from the active users in the current transmission time interval.

In any of the above technical solutions, the step of determining the target number of users to be scheduled according to the number of users that can be scheduled in the current transmission time interval and the number of active users in the current transmission time interval, specifically includes: when the current transmission time interval can be scheduled When the number of users is greater than or equal to the number of active users in the current transmission time interval, the target number is equal to the number of active users in the current transmission time interval; when the number of users that can be scheduled in the current transmission time interval is less than the number of active users in the current transmission time interval, the target The number is equal to the maximum number of users that can be scheduled in the current transmission time interval.

In any of the above technical solutions, the preset screening rule is one of the following rules: polling rule, proportional fairness rule, and enhanced proportional fairness maximum load-to-interference ratio rule.

In any of the above technical solutions, the preset algorithm is one of the following algorithms: particle swarm optimization, particle swarm optimization with Gaussian mutation, genetic algorithm, adaptive genetic algorithm.

In any of the above technical solutions, the step of obtaining the evaluation mechanism of the pairing strategy specifically includes: obtaining the optimal time-frequency resource allocation method corresponding to the current transmission time interval pairing strategy and the corresponding maximum available throughput; Score as a matching strategy.

In any of the above technical solutions, the step of obtaining the optimal time-frequency resource allocation method corresponding to the current transmission time interval pairing strategy and the corresponding maximum throughput that can be obtained specifically includes: after the current transmission time interval is given the pairing strategy, Use the evaluation mechanism to obtain the time-frequency resource allocation method that maximizes the current transmission time interval throughput; obtain the maximum throughput according to the time-frequency resource allocation strategy that maximizes the current transmission time interval throughput.

In a second aspect, the present application provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it is implemented as any one of the first aspect of the present application. Steps of the time-frequency space resource allocation method.

The computer device proposed in the second aspect of the present application includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it is implemented as any one of the first aspect of the present application. Steps of frequency-space resource allocation method. Therefore, it has all the beneficial effects of the time-frequency space resource allocation method of any of the above technical solutions.

In a third aspect, the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for allocating space-time resources according to any one of the first aspect of the present application are implemented.

A computer-readable storage medium proposed in the third aspect of the present application has a computer program stored thereon, and when the computer program is executed by a processor, a time-frequency-space resource allocation method as described in any one of the above technical solutions is implemented. Therefore, it has all the beneficial effects of the time-frequency space resource allocation method of any of the above technical solutions.

BRIEF DESCRIPTION

The drawings herein are incorporated into and constitute a part of this specification, show embodiments consistent with this application, and are used together with the specification to explain the principles of this application.

In order to more clearly explain the embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings required in the embodiments or the description of the prior art. Obviously, for those of ordinary skill in the art In other words, other drawings can be obtained based on these drawings without paying any creative labor.

FIG. 1 is a schematic flowchart of a method for allocating space-time space-frequency resources according to an embodiment of the present application;

FIG. 2 shows a schematic flowchart of a method for allocating space-time space-frequency resources provided by another embodiment of the present application;

FIG. 3 shows a schematic flowchart of a method for allocating space-time resources in another embodiment of the present application;

FIG. 4 shows a schematic flowchart of a method for allocating space-time resources in a specific embodiment of the present application;

FIG. 5 is a schematic flow chart of using the particle swarm optimization algorithm and the space-separation user pairing strategy evaluation mechanism to determine the optimal space-separation user pairing and time-frequency resource allocation strategy in the time-frequency-space resource allocation method of the embodiment shown in FIG. 4; FIG.

6 is a schematic flow chart of determining the optimal space-division user pairing and time-frequency resource allocation strategy by using the particle swarm optimization algorithm with Gaussian mutation and the space-division user pairing strategy evaluation mechanism in the time-frequency-space resource allocation method of the embodiment shown in FIG. 4;

FIG. 7 is a schematic flow chart of determining the optimal space-division user pairing and time-frequency resource allocation strategy using the genetic algorithm and the space-division user pairing strategy evaluation mechanism in the time-frequency-space resource allocation method of the embodiment shown in FIG. 4;

FIG. 8 is a schematic flow chart of using an adaptive genetic algorithm in conjunction with the space-division user pairing strategy evaluation mechanism to determine the optimal space-division user pairing and time-frequency resource allocation strategy in the time-frequency space resource allocation method of the embodiment shown in FIG. 4;

9 shows a structural block diagram of a computer device provided by an embodiment of the present application.

detailed description

To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments It is a part of the embodiments of this application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without making creative efforts fall within the protection scope of the present application.

In the first aspect, the present application provides a method for allocating space-time resources.

FIG. 1 shows a schematic flowchart of a method for allocating space-time resources according to an embodiment of the present application. As shown in FIG. 1, the time-frequency-space resource allocation method includes: S101, acquiring users to be scheduled; S102, determining a pairing strategy of users to be scheduled according to a preset algorithm; S103, acquiring an evaluation mechanism of the pairing strategy; S104, according to an evaluation mechanism Obtain the optimal time-frequency resource allocation method corresponding to the pairing strategy, and score the pairing strategy.

The time-frequency-space resource allocation method provided in this application first obtains the user to be scheduled, and then uses a preset algorithm to determine the matching strategy of the user to be scheduled; further, after determining the matching strategy of the user to be scheduled, the evaluation mechanism of the matching strategy is obtained, and then based on The evaluation mechanism obtains the optimal time-frequency resource allocation method corresponding to the matching strategy and scores the matching strategy. Specifically, the optimal time-frequency resource allocation method is the time-frequency resource allocation method that maximizes the wireless SE and cell throughput. It is worth noting that this application determines the pairing strategy of air separation users through a preset algorithm, and evaluates the optimal time-frequency resource allocation method corresponding to the pairing strategy through an evaluation mechanism. The two cooperate with each other to determine the optimal air separation user The pairing and time-frequency resource allocation strategy maximizes wireless SE/cell throughput while considering fairness.

FIG. 2 shows a schematic flowchart of a method for allocating space-time resources according to another embodiment of the present application. As shown in FIG. 2, the time-frequency space resource allocation method includes: S201, determining the target number of users to be scheduled according to the number of users that can be scheduled in the current transmission time interval and the number of active users in the current transmission time interval; S202, according to the Set filtering rules to filter out the target number of users to be scheduled from the active users in the current transmission time interval; S203, determine the pairing strategy of the user to be scheduled according to a preset algorithm; S204, obtain the evaluation mechanism of the pairing strategy; S205, obtain according to the evaluation mechanism The optimal time-frequency resource allocation method corresponding to the matching strategy and scoring the matching strategy.

In this embodiment, the number of active users in a cell's current transmission time interval and the number of users that can be scheduled in the current transmission time interval often do not match. Specifically, the number of activated users in the current transmission time interval is greater than the number of users that can be scheduled. Therefore, the activated users in the current transmission time interval need to be screened, and the filtered activated users are regarded as users to be scheduled, so that the filtered The number of activated users matches the number of users that can be scheduled in the current transmission time interval, thereby ensuring that the base station can stably transmit information for users to be scheduled. Specifically, the screening of the activated users in the current transmission time interval is performed according to preset screening rules, and the preset screening rules may be set in advance according to actual needs.

In an embodiment of the present application, the step of determining the target number of users to be scheduled according to the number of users that can be scheduled in the current transmission time interval and the number of active users in the current transmission time interval specifically includes: when the current transmission time interval can be When the number of scheduled users is greater than or equal to the number of active users in the current transmission time interval, the target number is equal to the number of active users in the current transmission time interval; when the number of users that can be scheduled in the current transmission time interval is less than the number of active users in the current transmission time interval, The target number is equal to the maximum number of users that can be scheduled in the current transmission time interval.

In this embodiment, in a given cell, the number of active users in the current transmission time interval and the number of users that can be scheduled in the current transmission time interval generally do not match, and there are the following two relationships: the first is the current When the number of users that can be scheduled in the transmission time interval is greater than or equal to the number of active users in the current transmission time interval, the representative base station can provide services for all active users at this time, so there is no need to filter, and the total number of active users in the current transmission time interval is directly As a user to be scheduled; the second is that the number of users that can be scheduled in the current transmission time interval is less than the number of active users in the current transmission time interval. At this time, it means that the base station cannot provide services for all activated users. The activated users in the current transmission time interval are screened, and the filtered activated users are regarded as users to be scheduled, so that the number of activated users after screening matches the number of users who can be scheduled in the current transmission time interval. Specifically, after the handsome selection, the target number of users to be scheduled is equal to the maximum number of users that can be scheduled in the current transmission time interval, and on the premise of ensuring the normal operation of the base station, try to provide services for as many users as possible.

In an embodiment of the present application, the preset screening rule is one of the following rules: polling rule, proportional fairness rule, and enhanced proportional fairness maximum load-to-interference ratio rule.

In this embodiment, when the number of active users in the current transmission time interval of a cell is greater than or equal to the number of users that can be scheduled in the current transmission time interval, the number of active users in the current transmission time interval needs to be screened. Use any of the polling rules, proportional fairness rules, and enhanced proportional fairness maximum load-to-interference ratio rules to obtain a target number of users to be scheduled. Specifically, what kind of preset filtering rules to use can be selected according to actual needs, which is not limited herein.

In an embodiment of the present application, the preset algorithm is one of the following algorithms: particle swarm optimization, particle swarm optimization with Gaussian mutation, genetic algorithm, and adaptive genetic algorithm.

In this embodiment, in the process of determining the pairing strategy of the air separation user according to the preset algorithm, one of the particle swarm algorithm, the particle swarm algorithm with Gaussian mutation, the genetic algorithm, and the adaptive genetic algorithm can be selected according to the actual needs to determine the space For the user-pairing strategy, the above intelligent algorithms can all obtain the optimal air-separation user matching strategy. Specifically, what kind of preset algorithm is adopted can be selected according to actual needs, which is not limited herein.

FIG. 3 shows a schematic flowchart of a method for allocating space-time resources in another embodiment of the present application. As shown in FIG. 3, the time-frequency-space resource allocation method includes: S301, determining the target number of users to be scheduled according to the number of users that can be scheduled in the current transmission time interval and the number of active users in the current transmission time interval; S302, according to the Set filtering rules to filter out the target number of users to be scheduled from the active users in the current transmission time interval; S303, determine the pairing strategy of the user to be scheduled according to a preset algorithm; S304, obtain the optimal time-frequency corresponding to the current transmission time interval pairing strategy The resource allocation method and the corresponding maximum throughput that can be obtained; S305, the maximum throughput is used as the score of the pairing strategy; S306, the optimal time-frequency resource allocation method corresponding to the pairing strategy is obtained according to the evaluation mechanism, and the pairing strategy is scored.

In this embodiment, in the process of obtaining the evaluation mechanism of the pairing strategy, the maximum throughput obtainable by the current transmission time interval pairing strategy is first obtained, and then the maximum throughput obtainable by the current transmission time interval pairing strategy is taken as the Evaluation mechanism. In other words, the evaluation of the pairing strategy is based on its maximum throughput in the current transmission time interval. Whether the pairing strategy is optimal is proportional to its maximum throughput in the current transmission time interval. Therefore, in order to maximize the wireless SE/cell throughput, it is necessary to select the pairing strategy with the largest maximum throughput in the current transmission time interval.

In an embodiment of the present application, the step of obtaining the maximum throughput obtainable by the current transmission time interval pairing strategy specifically includes: after the current transmission time interval is given a pairing strategy, obtaining the time when the current transmission time interval throughput can be maximized Frequency resource allocation strategy; the maximum throughput is obtained according to the time-frequency resource allocation strategy that maximizes the throughput of the current transmission time interval.

In this embodiment, in the process of obtaining the maximum throughput obtainable by the current transmission time interval pairing strategy, when the current transmission time interval pairing strategy is given, the time-frequency resource that can maximize the throughput of the current transmission time interval is obtained. The allocation strategy can be called the time-frequency resource allocation strategy as the target time-frequency resource allocation strategy, and then the throughput of the target time-frequency resource allocation strategy in the current transmission time interval is taken as the maximum throughput that can be obtained by the current transmission time interval pairing strategy.

FIG. 4 shows a schematic flowchart of a method for allocating space-time resources in a specific embodiment of the present application. As shown in FIG. 4, the time-frequency space resource allocation method includes: S401, screening out users to be scheduled; S402, determining an air separation user pairing strategy evaluation mechanism; S403, using an intelligent algorithm to cooperate with the air separation user pairing strategy evaluation mechanism to determine the optimal Space division user pairing and time-frequency resource allocation strategy.

In this embodiment, the intelligent algorithm may use particle swarm optimization, particle swarm optimization with Gaussian mutation, genetic algorithm, or adaptive genetic algorithm.

In the following, the intelligent algorithm is described as particle swarm optimization, particle swarm optimization with Gaussian mutation, genetic algorithm, and adaptive genetic algorithm.

One: Use the RR (Round Robin Robin Algorithm) method to screen out the current TTI downlink users to be scheduled and use the particle swarm algorithm to determine the pairing strategy of the air separation users.

The basic idea of this embodiment is to use the RR method to screen out the current TTI downlink users to be scheduled; determine the air separation user pairing strategy evaluation mechanism; use the particle swarm optimization algorithm and the air separation user pairing strategy evaluation mechanism to determine the downlink optimal air separation user pairing and Time-frequency resource allocation strategy.

In S401, U users are selected from the current TTI downlink active users as the current TTI downlink to-be-scheduled users in a polling manner; wherein, the value of U is determined by the base station's downlink scheduling capability per TTI and the current number of TTI downlink active users : If the base station’s downlink scheduling capability per TTI is not less than the current number of TTI downlink active users, no user screening is performed, that is, U is equal to the current TTI downlink active users; if the base station’s downlink scheduling capability per TTI is less than the current TTI downlink active users, proceed For user selection, U is equal to the maximum number of downlink scheduled users per TTI supported by the base station. In the embodiment of the present application, it is assumed that the current number of TTI downlink active users is 22, and the cell downlink scheduling capability is 16, that is, U=16.

In S402, an air separation user pairing strategy evaluation mechanism is determined.

Let u∈{1,...,U} denote the current TTI downlink user index to be scheduled.

Let g∈{1,...,G _MU +G _SU } denote the current TTI downlink empty packet/frequency division index, G _MU denote the current TTI downlink empty packet number, and G _SU denote the current TTI downlink frequency division user number; where, When g∈{1,...,G _MU },g is the current TTI downlink empty group index, when g∈{G _MU +1,...,G _MU +G _SU },g is the current TTI downlink frequency division User index.

Let I _u,g indicate whether the user u belongs to the downlink empty packet/frequency division g at the current TTI; where I _u,g =1 means that the user u belongs to the downlink empty packet/frequency division g at the current TTI, I _u,g =0 Indicates that user u does not belong to the downlink empty packet/frequency division g at the current TTI.

Let RB _g denote the number of RBs (Resource Block) allocated by the current TTI for downlink empty packets/frequency division g, let RB_Total denote the total amount of available RBs in the current TTI downlink, let

Indicates the number of RBs that user u needs to obtain at least in the current TTI; in the embodiment of this application, it is assumed that the total downlink available RBs in the current TTI is 100, and each user needs to obtain at least 1 RB in the current TTI, that is, RB_Total=100,

Let SE _u,g denote the number of bits that the current TTI user u can send per RB in the downlink empty packet/frequency division g; where, let g'denote the downlink empty packet assigned to user u under the current TTI given empty allocation pair strategy/ Frequency division index, if g≠g', SE _u,g = 0; if g ==g', SE _u,g is equal to the number of bits that user u can send per RB in the downlink empty packet/frequency division g (this value is determined by The quality of the channel where the user u is located and the downlink empty packet/frequency division g include which users jointly decide).

Let BSR _u denote the number of bits to be sent by the current TTI user u; in the embodiment of this application, it is assumed that all users to be scheduled in the current TTI downlink have an infinite BSR, that is, BSR _u = ∞,

To sum up, in the current TTI, given any pairing strategy I _u,g of the space-separation users, we can obtain the current TTI of the space allocation pair strategy corresponding to the optimal time-frequency resource allocation strategy RB _g and available by solving the following convex optimization problem Maximum cell throughput

Among them, the above convex optimization problem is used to solve RB _g , real-time frequency resource allocation strategy; the objective function of the above convex optimization problem is to maximize the downlink throughput of the current TTI cell; the first constraint of the above convex optimization problem refers to any empty packet of the current TTI The number of RB (Resource Block) obtained by frequency division is not greater than the total amount of RBs available in the current TTI, as far as possible to meet the minimum RB requirements of all users of empty grouping/frequency division; the second constraint of the above convex optimization problem is It means that the total number of RBs obtained from all empty packets/frequency divisions of the current TTI is not greater than the total available RBs of the current TTI.

In S403: the particle swarm optimization algorithm is used in conjunction with the air separation user pairing strategy evaluation mechanism to determine the optimal space separation user pairing and time-frequency resource allocation strategy.

FIG. 5 is a schematic flow chart of using the particle swarm optimization algorithm and the space-separation user pairing strategy evaluation mechanism to determine the optimal space-separation user pairing and time-frequency resource allocation strategy in the time-frequency-space resource allocation method of the embodiment shown in FIG. 4; As shown, the time-frequency space-time resource allocation method includes: S501, setting the number of particles, the maximum number of iterations, the number of initial iterations, the position of the particles, and the speed; S502, obtaining the corresponding empty allocation pair strategy for each particle according to the position of the particles, thereby obtaining each particle Corresponding to the value of I _u,g , finally get the value of each particle corresponding to SE _u,g , G _MU , G _SU ; S503, the value of I _u,g , SE _u,g , G _MU , G _SU corresponding to the particle Substitute into the air separation user pairing strategy evaluation mechanism, solve the convex optimization problem to obtain the time-frequency resource allocation strategy and maximum cell throughput corresponding to each particle, and set the particle fitness to this maximum cell throughput; iteration number +1; S504, Determine whether the number of iterations reaches the maximum number of iterations, when the result is no, execute S505, otherwise execute S506; S505, update the position and velocity of each particle, and return to S502; S506, output the empty allocation pair strategy and time frequency corresponding to the optimal particle Resource allocation strategy.

In S501, set the number of particles, the maximum number of iterations, initialize the number of iterations, particle position, and velocity. It is assumed that the cell supports at most one downlink empty packet, and the number of users to be scheduled at the same time is U=16, so there are 2 ¹⁶ =65536 possible empty allocation pairs; therefore, we set the particle position selectable range from 0 to 65535, particle speed The available range is -65535 to 65535; here, we set the number of particles to 100, the maximum number of iterations is 100, and initialize the number of iterations to 0 at the same time. In addition, let p ∈ {1,...P} denote the particle index, and P denote the number of particles (in this embodiment, P=100). Let Pbestpop _p denote the optimal position obtained by particle p in the iterative process, Pbestvalue _p denote the corresponding fitness of particle p in obtaining the optimal position in the iterative process, and let Gbestpop denote the global optimal position obtained by all particles in the iterative process , Gbestvalue represents the corresponding fitness of all particles when they obtain the global optimal position in the iterative process. At the same time, Pbestpop _p = 0, Pbestvalue _p = -∞, Gbestpop = 0, Gbestvalue = -∞ are initialized.

In S502, the empty allocation pair strategy corresponding to each particle is obtained according to the particle position, thereby obtaining the value corresponding to I _u,g for each particle, and finally obtaining the value corresponding to SE _u,g , G _MU , and G _SU for each particle.

For example, the location of a particle is 4758.44. First, we rounded the particle position to get 4758. Then, the rounded position information is subjected to binary conversion (if the number of digits of the value obtained after the binary conversion is less than U, it is padded with 0 until the number of digits of the value obtained after the binary conversion is equal to U), and 0001001010010110 is obtained. We let 0 in the binary value represent the frequency division, and 1 represents the empty group 1, so we can get the space division strategy represented by the above particles: users 1, 2, 3, 5, 6, 8, 10, 11, 13, 16 frequency division ; User 4, 7, 9, 12, 14, 15 enter the empty packet 1. From this, we get the value of I _u,g (due to space limitations, the value of I _u,g is not given here), G _MU =1, G _SU =10. Finally, the frequency division user SE is obtained according to the channel quality of the frequency division user, and the space division user SE is obtained according to the channel quality of the space division user and how many other users are included in the space group and the correlation between the users in the space group (due to space limitations, SE _u,g values are not given here).

In S502, we get the values of I _u,g , SE _u,g , G _MU , and G _SU corresponding to each particle. In S503, we divide each particle I _u,g , SE _u,g , G The values of _MU and G _SU are substituted into the air separation user pairing strategy evaluation mechanism:

Solve the above convex optimization problem to obtain the optimal time-frequency resource allocation strategy corresponding to each particle and the downlink throughput of each particle corresponding to the current TTI cell

Use this as the particle fitness; after all particle fitness calculations are completed, update the individual optimal position Pbestpop _{p of} each particle particle and the corresponding individual optimal fitness Pbestvalue _p and the global optimal position of all particles Gbestpop and corresponding global optimal Fitness Gbestvalue. After completing the above work, the number of iterations will be +1.

In S503, the space division user pairing strategy evaluation mechanism is used to obtain the throughput of the cell corresponding to each particle as its fitness; the number of iterations is +1.

In S505, the process of updating the position and velocity of each particle is as follows:

For p=1: P% traverses all particles

% Update each particle velocity, where Velocity _p represents particle p velocity, rand() function generates a random decimal from 0 to 1, pop _p represents particle p position, max_Velocity represents particle maximum velocity, min_Velocity represents particle minimum velocity

Velocity _p = max(min_Velocity, min(max_Velocity, 0.7298·(Velocity _p +

2.05·rand()·(Pbestpop _p -pop _p )+2.05·rand()·(Gbestpop-pop _p ))))

% Update each particle position, where max_Position represents the maximum particle position and min_Position represents the minimum particle position

pop _p = max(min_Position,min(max_Position,pop _p +Velocity _p ))

End for

After the particle position and velocity are updated, the process returns to S502.

In S506, the empty allocation pair strategy and the time-frequency resource allocation strategy corresponding to the optimal particles are output.

The particles with the highest fitness in the previous iterations are output to obtain their corresponding empty allocation pair strategy and time-frequency resource allocation strategy.

2. Use PF (Proportional Fairness proportional fairness algorithm) to screen out the current TTI downlink users to be scheduled and use the particle swarm algorithm with Gaussian mutation to determine the pairing strategy of air separation users.

The basic idea of this embodiment is to select the current TTI downlink users to be scheduled by PF; determine the air separation user pairing strategy evaluation mechanism; use the particle swarm optimization algorithm with Gaussian mutation and the air separation user pairing strategy evaluation mechanism to determine the optimal downlink space Sub-user pairing and time-frequency resource allocation strategy.

In S401, the current TTI downlink to-be-scheduled users are selected by PF.

According to the size of the PF factor, select the U users with the largest PF factor from the current TTI downlink active users as the current TTI downlink users to be scheduled; where the value of U is determined by the base station's TTI downlink scheduling capability and the current number of TTI downlink active users: if The base station's downlink scheduling capability per TTI is not less than the current number of TTI downlink active users, user screening is not performed, that is, U is equal to the current TTI downlink active users; if the base station's downlink scheduling capability per TTI is less than the current TTI downlink active users, user screening , U is equal to the maximum number of downlink scheduled users per TTI supported by the base station; in the embodiment of the present application, it is assumed that the current number of TTI downlink active users is 40, and the cell downlink scheduling capability is 32, that is, U=32.

Let u∈{1,...,U} denote the current TTI downlink user index to be scheduled.

Let RB _g denote the number of RBs (Resource Blocks) allocated by the current TTI for downlink empty packets/frequency division g, let RB_Total denote the total amount of available RBs in the current TTI downlink, let

Let SE _u,g denote the number of bits that the current TTI user u can send per RB in the downlink empty packet/frequency division g; where, let g′ denote the downlink empty packet assigned to user u under the current TTI given empty allocation pair strategy/ Frequency division index, if g≠g', SE _u,g = 0; if g ==g', SE _u,g is equal to the number of bits that user u can send per RB in the downlink empty packet/frequency division g (this value is determined by The quality of the channel where the user u is located and the downlink empty packet/frequency division g include which users jointly decide).

Among them, the above convex optimization problem is used to solve RB _g , real-time frequency resource allocation strategy; the objective function of the above convex optimization problem is to maximize the downlink throughput of the current TTI cell; the first constraint of the above convex optimization problem refers to any empty packet of the current TTI /The number of RBs obtained by frequency division should not exceed the total amount of RBs available in the current TTI, and as far as possible meet the minimum RB requirements of all users/frequency division; the second constraint of the above convex optimization problem is that all empty groups of the current TTI / The total number of RBs obtained by frequency division is not greater than the total available RBs of the current TTI.

In S403, a particle swarm optimization algorithm with Gaussian mutation is used in conjunction with the space-separation user pairing strategy evaluation mechanism to determine the optimal space-separation user pairing and time-frequency resource allocation strategy.

FIG. 6 is a schematic flow chart of determining the optimal space-division user pairing and time-frequency resource allocation strategy by using the particle swarm optimization algorithm with Gaussian mutation and the space-division user pairing strategy evaluation mechanism in the time-frequency-space resource allocation method of the embodiment shown in FIG. 4. As shown in FIG. 6, the time-frequency space-time resource allocation method includes: S601, setting the number of particles, the maximum number of iterations, the number of initial iterations, the position of the particles, and the speed; S602, according to the position of the particles, the corresponding empty allocation strategy for each particle is obtained, thereby Obtain the value of each particle corresponding to I _u,g , and finally obtain the value of each particle corresponding to SE _u,g , G _MU , G _SU ; S603, the corresponding I _u,g , SE _u,g , G _MU , The value of G _SU is substituted into the space-division user pairing strategy evaluation mechanism to solve the convex optimization problem to obtain the time-frequency resource allocation strategy and maximum cell throughput corresponding to each particle, and set the particle fitness to this maximum cell throughput; number of iterations + 1; S604, Gaussian mutation; S605, to determine whether the number of iterations reaches the maximum number of iterations, when the result is no, execute S606, otherwise execute S607; S606, update the position and velocity of each particle, and return to S602; S607, output the optimal particles Corresponding empty allocation pair strategy and time-frequency resource allocation strategy; for example, in S601, set the number of particles, the maximum number of iterations, initialize the number of iterations, particle position, speed.

In the embodiment of the present application, it is assumed that the cell supports at most one downlink empty packet and the number of users to be scheduled at the same time is U=32, so there are 2 ³² = 4,294,967,296 possible empty allocation pairs; therefore, we set the particle position selectable range to 0～4,294,967,295, the particle speed can be selected from -4,294,967,295～4,294,967,295; here, we set the number of particles to 100, the maximum number of iterations is 100, and initialize the number of iterations to 0. In addition, let p ∈ {1,...P} denote the particle index, and P denote the number of particles (in this embodiment, P=100). Let Pbestpop _p denote the optimal position obtained by particle p in the iterative process, Pbestvalue _p denote the corresponding fitness of particle p in obtaining the optimal position in the iterative process, and let Gbestpop denote the global optimal position obtained by all particles in the iterative process , Gbestvalue represents the corresponding fitness of all particles when they obtain the global optimal position in the iterative process. At the same time, Pbestpop _p = 0, Pbestvalue _p = -∞, Gbestpop = 0, Gbestvalue = -∞ are initialized.

In S602, the empty allocation pair strategy corresponding to each particle is obtained according to the particle position, thereby obtaining the value corresponding to each particle I _u,g , and finally obtaining the value corresponding to each particle SE _u,g , G _MU , G _SU .

For example, the position of a particle is 294967295.44. First, we rounded the particle position to get 294967295. Then, the rounded position information is subjected to binary conversion (if the number of digits of the value obtained after the binary conversion is less than U, then 0 is filled in front of it, until the number of digits of the value obtained after the binary conversion is equal to U), 00010001100101001101011111111111 is obtained. We let 0 in the binary value represent the frequency division and 1 represents the empty group 1, so we can get the space division strategy represented by the above particles: users 1, 2, 3, 5, 6, 7, 10, 11, 13, 15, 16. , 19, 21 frequency division; users 4, 8, 9, 12, 14, 17, 18, 20, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32 enter the empty packet 1. From this, we get the value of I _u,g (due to space limitations, the value of I _u,g is not given here), G _MU =1, G _SU =13. Finally, the frequency division user SE is obtained according to the channel quality of the frequency division user, and the space division user SE is obtained according to the channel quality of the space division user and how many other users are included in the space group and the correlation between the users in the space group (due to space limitations, SE _u,g values are not given here).

In S603, the space division user pairing strategy evaluation mechanism is used to obtain the cell throughput corresponding to each particle as its fitness; the number of iterations is +1.

In S602, we get the values of I _u,g , SE _u,g , G _MU , and G _SU corresponding to each particle. In S603, we divide each particle I _u,g , SE _u,g , G The values of _MU and G _SU are substituted into the air separation user pairing strategy evaluation mechanism:

In S604, Gaussian mutation.

The Gaussian mutation process is shown in the pseudocode below:

% Velocity mutation operation, where Velocity_mutation represents the set of mutated particle velocities (in this embodiment, we generate 10 mutated particles per iteration), max_Velocity represents the maximum velocity of the particle, min_Velocity represents the minimum velocity of the particle, and Gbest_Velocity represents the current global optimal particle Speed, normrnd(0,1,10) function generates 10 standard normally distributed random numbers

Velocity_mutation=max(min_Velocity,min(max_Velocity,Gbest_Velocity·e ^{normrnd(0,1,10)} ))% position mutation operation, where pop_mutation represents the set of mutated particle positions (in this embodiment, we generate 10 mutated particles per iteration ), max_Position represents the maximum particle position, min_Position represents the minimum particle position

pop_mutation=max(min_Position,min(max_Position,Gbestpop+

Velocity_mutationnormrnd(0,1,10)))

% Calculate the fitness of the mutated particles. The function Fitvalue_mutation=Fitness(pop_mutation) is used to calculate the fitness of each particle in the particle set pop_mutation and return the fitness of each particle to Fitvalue_mutation

Fitvalue_mutation=Fitness(pop_mutation)

% Update the global optimal particle, the function [Gbest_Fitness,II]=max(Fitvalue_mutation) finds the largest fitness value in the set Fitvalue_mutation and returns it to Gbest_Fitness, and returns the index of the largest numerical position to II; after that, the Gbest_Fitness and the current global most Comparison of optimal particle fitness Gbestvalue, if Gbest_Fitness is greater, update the current global optimal particle position, velocity and fitness

[Gbest_Fitness,II]=max(Fitvalue_mutation)

If Gbest_Fitness>Gbestvalue

Gbestpop=pop_mutaion(II)

Gbest_Velocity=Velocity_mutaion(II)

Gbestvalue=Gbest_Fitness

End

In S605, it is determined whether the maximum number of iterations is reached, and it is determined whether the current number of iterations is equal to the maximum number of iterations. If not, S606 is executed; otherwise, S607 is executed.

In S606, the position and velocity of each particle are updated, and the process returns to S602.

See the pseudocode below for updating the position and velocity of each particle:

For p=1: P% traverses all particles

Velocity _p = max(min_Velocity, min(max_Velocity, 0.7298·(Velocity _p +

2.05·rand()·(Pbestpop _p -pop _p )+2.05·rand()·(Gbestpop-pop _p ))))

pop _p = max(min_Position,min(max_Position,pop _p +Velocity _p ))

End for

After the particle position and velocity are updated, the process returns to S602.

In S607, the empty allocation pair strategy and the time-frequency resource allocation strategy corresponding to the optimal particles are output.

Three: Use Max C/I (Maximum Carrier to Interference Maximum Carrier-to-Interference Ratio) method to screen out the current TTI uplink to be scheduled users and use genetic algorithm to determine the pairing strategy of air separation users.

The basic idea of this embodiment is to use the Max C/I method to screen out the current TTI uplink users to be scheduled; determine the air separation user pairing strategy evaluation mechanism; use genetic algorithm to cooperate with the air separation user pairing strategy evaluation mechanism to determine the uplink optimal air separation user Pairing and time-frequency resource allocation strategy.

In S401, U users with the best channel quality are selected from the current TTI uplink active users according to the channel quality of each user as the current TTI uplink to-be-scheduled user; where the value of U is determined by the base station's TTI uplink scheduling capability and current The number of TTI uplink active users is determined: if the base station's TTI uplink scheduling capability is not less than the current TTI uplink active users, no user screening is performed, that is, U is equal to the current TTI uplink active users; if the base station's TTI uplink scheduling capability is less than the current TTI uplink The number of activated users is selected, and U is equal to the maximum number of uplink scheduled users per TTI supported by the base station. In this embodiment of the present application, it is assumed that the current number of TTI uplink active users is 16, and the cell uplink scheduling capability is 48, that is, U=16 .

Let u∈{1,...,U} denote the current TTI uplink user index to be scheduled.

Let g∈{1,...,G _MU +G _SU } denote the current TTI uplink empty packet/frequency division index, G _MU denote the current TTI uplink empty packet number, and G _SU denote the current TTI uplink frequency division user number; where, When g∈{1,...,G _MU },g is the current TTI uplink empty packet index, when g∈{G _MU +1,...,G _MU +G _SU },g is the current TTI uplink frequency division User index.

Let I _u,g indicate whether user u belongs to upstream empty packet/frequency division g at current TTI; where I _u,g =1 means user u belongs to upstream empty packet/frequency division g at current TTI, I _u,g =0 It indicates that user u does not belong to the uplink empty packet/frequency division g at the current TTI.

Let RB _g denote the number of RBs (Resource Blocks) allocated by the current TTI for uplink empty packets/frequency division g, let RB_Total denote the total amount of RBs available in the current TTI uplink, let

Indicates the number of RBs that user u needs to obtain at least in the current TTI; in the embodiment of the present application, it is assumed that the total amount of available RBs in the current TTI uplink is 100, and each user needs to obtain at least 1 RB in the current TTI, that is, RB_Total=100,

Let SE _u,g denote the number of bits that the current TTI user u can send in each uplink RB in the uplink empty packet/frequency division g; where, let g′ denote the uplink empty packet assigned to user u under the current TTI given empty allocation pair strategy/ Frequency division index, if g≠g', SE _u,g =0; if g==g', SE _u,g is equal to the number of bits that user u can send per RB in the upstream empty packet/frequency division g (this value is determined by The quality of the channel where user u is located and the upstream empty packet/frequency division g include which users jointly decide).

Let BSR _u denote the number of bits to be sent by the current TTI user u.

Among them, the above convex optimization problem is used to solve RB _g , real-time frequency resource allocation strategy; the objective function of the above convex optimization problem is to maximize the uplink throughput of the current TTI cell; the first constraint of the above convex optimization problem refers to any empty packet of the current TTI /The number of RBs obtained by frequency division should not exceed the total amount of RBs available in the current TTI, and as far as possible meet the minimum RB requirements of all users/frequency division; the second constraint of the above convex optimization problem is that all empty groups of the current TTI / The total number of RBs obtained by frequency division is not greater than the total available RBs of the current TTI.

In S403, a genetic algorithm is used in conjunction with the evaluation mechanism of the air separation user pairing strategy to determine the optimal uplink air separation user pairing and time-frequency resource allocation strategy.

FIG. 7 is a schematic flow chart of determining the optimal space-division user pairing and time-frequency resource allocation strategy using the genetic algorithm and the space-division user pairing strategy evaluation mechanism in the time-frequency-space resource allocation method of the embodiment shown in FIG. 4. As shown in FIG. 7, the time-frequency space resource allocation method includes: S701, setting the population size, the maximum number of iterations, initializing the number of iterations, and population; S702, obtaining the corresponding empty allocation pair strategy for each individual according to the individual chromosome information, thereby obtaining each Each individual corresponds to the value of I _u,g , and finally obtains the value of each individual corresponding to SE _u,g , G _MU , G _SU ; S703, the particle corresponding to I _u,g , SE _u,g , G _MU , G _SU The value of is substituted into the air separation user pairing strategy evaluation mechanism, solving the convex optimization problem to obtain the time-frequency resource allocation strategy and maximum cell throughput corresponding to each particle, and setting the particle fitness to this maximum cell throughput; the number of iterations +1; S704, determine whether the number of iterations reaches the maximum number of iterations, when the result is no, execute S705, otherwise execute S706; S705, select, cross, mutate, and return to S702; S706, output the empty allocation pair strategy and time corresponding to the optimal individual Frequency resource allocation strategy.

In S701, the population size and the maximum number of iterations are set, and the number of iterations and the population are initialized.

In the embodiment of the present application, it is assumed that the cell supports up to three uplink empty packets, and the number of users to be scheduled at the same time is U=16, so there are 4 ¹⁶ = 4,294,967,296=2 ³² possible pairs of empty allocation pairs; therefore, we set the population The individual chromosome length is 32; here, we set the population size to 100, the maximum number of iterations is 100, and initialize the number of iterations to 0. In addition, let Gbestpop denote the global optimal individual chromosome sequence obtained by all individuals during the iteration process, and Gbestvalue denote the global optimal fitness of all particles obtained during the iteration process.

In S702, each individual's corresponding empty allocation pair strategy is obtained according to the individual's chromosome information, thereby obtaining the value of each individual corresponding to I _u,g , and finally the value of each individual corresponding to SE _u,g , G _MU , and G _SU .

For example, an individual chromosome sequence is 11000100011001010011010111111111. First, we convert this chromosome sequence to quaternary 3010121103113333. We let 0 in the quaternary value represent frequency division, 1 represents empty group 1, 2 represents empty group 2, 3 represents empty group 3, so we can get the space division strategy represented by the above individuals: user 2, 4, 9 frequency division ; User 3, 5, 7, 8, 11, 12 enters empty group 1, user 6 enters empty group 2 (empty group 2 has only one user, so it can be regarded as another frequency division user), users 1, 10, 13 , 14, 15, 16 enter the empty packet 3. From this, we get the value of I _u,g (due to space limitations, the value of I _u,g is not given here), G _MU =2, G _SU =4. Finally, the frequency division user SE is obtained according to the channel quality of the frequency division user, and the space division user SE is obtained according to the channel quality of the space division user and the correlation between the users in the empty group (due to space limitations, the value of SE _u,g is not here Given).

In S703, the space division user pairing strategy evaluation mechanism is used to obtain the throughput of the cell corresponding to each particle as its fitness; the number of iterations is +1.

In S702, we get the values of I _u,g , SE _u,g , G _MU , and G _SU corresponding to each particle. In S703, we divide each particle I _u,g , SE _u,g , G The values of _MU and G _SU are substituted into the air separation user pairing strategy evaluation mechanism:

Solve the above convex optimization problem to obtain the optimal time-frequency resource allocation strategy corresponding to each individual and the uplink throughput of each individual corresponding to the current TTI cell

Use it as the individual fitness; after all individual fitness calculations are completed, update the global optimal individual chromosome sequence information Gbestpop and the corresponding global optimal individual fitness Gbestvalue. After completing the above work, the number of iterations will be +1.

In S704, whether the maximum number of iterations is reached.

Determine whether the current number of iterations is equal to the maximum number of iterations, if not, go to S705; if yes, go to S706.

In S705, select, cross, mutate, and return to S702.

The selection, crossover and mutation processes are shown in the pseudocode below:

% Selection process

totalfit=sum(fitvalue)%fitvalue represents the set of fitness values of all individuals, and totalfit represents the sum of fitness of all individuals

p_fitvalue=fitvalue/totalfit% p_fitvalue represents the ratio of each individual’s fitness to totalfit

p_fitvalue=cumsum(p_fitvalue)%sumsum() represents the accumulation function

ms=sort(rand(P))% generates P 0 to 1 random decimals, and assigns the P random numbers from small to large order to ms, P is the number of individuals in the population

fitin=1

newin=1

While newin<=P% selects P individuals to enter the next iteration, pop represents the current iteration individual set, and newpop represents the next iteration individual set. It can be seen that whether an individual can enter the next iteration is proportional to its fitness. In addition, it can be seen that an individual may be eliminated in this iteration, that is, unable to enter the next iteration; at the same time, an individual may also be copied multiple times into the next iteration

If `ms(newin)<p_fitvalue(fitin)

Newpop(newin)=pop(fitin)

Newin+=1

else fitin+=1

End If

End While

% Cross process

newpop＝pop

pointer=randperm(P)% randperm(P) randomly sorts P integers (1 to P) and outputs

For i=1:2: P-1% randomly cross P individuals and cross

Ifrand()<pc%rand() randomly generates a random decimal from 0 to 1, pc is the crossover probability

cpoint=round(rand()*(chromlength-1))% randomly find a cross point to cross the chromosome sequence of two paired individuals, chromlength is the length of the chromosome sequence, round() function is a rounding function

newpop(pointer(i),:)=[pop(pointer(i),1:cpoint+1),pop(pointer(i+1),cpoint+2:chromlength)]

newpop(pointer(i+1),:)=[pop(pointer(i+1),1:cpoint+1),pop(pointer(i),cpoint+2:chromlength)]

End If

End For

% Variation process

newpop＝pop

For i=1: P% mutates P individuals sequentially

If rand()<pm%pm is the crossover probability

Mpoint=round(rand()*(chromlength-1))% find a mutation point randomly

Newpop(i,mpoint+1)=1-newpop(i,mpoint+1)% reverse the mutation point chromosome information (0 to 1,1 to 0)

End If

End For

In S706, the empty allocation pair strategy and the time-frequency resource allocation strategy corresponding to the optimal individual are output.

The individual chromosome sequence information with the highest fitness in the previous iterations is output, so as to obtain its corresponding empty allocation pair strategy and time-frequency resource allocation strategy.

4. Use EPF (Enhanced Proportional Fairness Enhanced Proportional Fairness Algorithm) to screen out the current TTI uplink users to be scheduled and use adaptive particle swarm algorithm to determine the pairing strategy of air separation users.

The basic idea of this embodiment is to use EPF to screen out the current TTI uplink users to be scheduled; determine the air separation user pairing strategy evaluation mechanism; use adaptive genetic algorithm with the air separation user pairing strategy evaluation mechanism to determine the uplink optimal air separation user pairing And time-frequency resource allocation strategy.

In S401, the current TTI uplink to-be-scheduled users are screened using EPF.

According to the weighted PF factor of each user (weighted PF factor refers to the value obtained by summing the PF factor and QoS, etc.), the U users with the largest weighted PF factor are selected from the current TTI uplink active users as the current TTI uplink wait Scheduling users; where the value of U is determined by the base station's uplink scheduling capability per TTI and the current number of TTI uplink active users: if the base station's uplink scheduling capability per TTI is not less than the current number of TTI uplink active users, no user screening is performed, that is, U equals The number of current TTI uplink active users; if the base station's uplink scheduling capability per TTI is less than the current TTI uplink active users, user screening is performed, U is equal to the maximum number of uplink scheduled users per TTI supported by the base station; in the embodiments of this application, it is assumed that the current TTI uplink The number of active users is 24, and the uplink scheduling capability of the cell is 32, that is, U=24.

Let u∈{1,...,U} denote the current TTI uplink user index to be scheduled.

Let SE _u,g denote the number of bits that the current TTI user u can send per RB in the upstream empty packet/frequency division g; where, let g'denote the upstream empty packet assigned to user u under the current TTI given empty allocation pair strategy/ Frequency division index, if g≠g', SE _u,g =0; if g==g', SE _u,g is equal to the number of bits that user u can send per RB in the upstream empty packet/frequency division g (this value is determined by The quality of the channel where user u is located and the upstream empty packet/frequency division g include which users jointly decide).

Let BSR _u denote the number of bits to be sent by the current TTI user u.

In S403, an adaptive genetic algorithm is used in conjunction with the space-separation user pairing strategy evaluation mechanism to determine the optimal uplink space-separation user pairing and time-frequency resource allocation strategy.

FIG. 8 is a schematic flowchart of determining the optimal space-division user pairing and time-frequency resource allocation strategy using the adaptive genetic algorithm and the space-division user pairing strategy evaluation mechanism in the time-frequency-space resource allocation method of the embodiment shown in FIG. 4. As shown in FIG. 8, the time-frequency space resource allocation method includes: S801, setting the population size, the maximum number of iterations, initializing the number of iterations, and population; S802, obtaining the corresponding empty allocation pair strategy for each individual according to the individual chromosome information, thereby obtaining each Each individual corresponds to the value of I _u,g , and finally obtains the value of each individual corresponding to SE _u,g , G _MU , G _SU ; S803, the corresponding I _u,g , SE _u,g , G _MU , G _SU The value of is substituted into the air separation user pairing strategy evaluation mechanism, solving the convex optimization problem to obtain the time-frequency resource allocation strategy and maximum cell throughput corresponding to each particle, and setting the particle fitness to this maximum cell throughput; the number of iterations +1; S804, determine whether the number of iterations reaches the maximum number of iterations; when the result is no, execute S805, otherwise execute S806; S805, select, adaptive crossover, adaptive mutation, and return to S802; S806, output the empty allocation corresponding to the optimal individual Assign strategies to strategies and time-frequency resources.

In S801, set the population size, the maximum number of iterations, initialize the number of iterations, the population.

In this embodiment, it is assumed that the cell supports up to three uplink and empty packets, and the number of users to be scheduled at the same time is U=24, so there are 4 ²⁴ = 281,474,976,710,656=2 ⁴⁸ possible pairs of empty allocation pairs; therefore, we set the population The individual chromosome length is 48; here, we set the population size to 100, the maximum number of iterations is 100, and initialize the number of iterations to 0. In addition, let Gbestpop denote the global optimal individual chromosome sequence obtained by all individuals during the iteration process, and Gbestvalue denote the global optimal fitness of all particles obtained during the iteration process.

In S802, each individual's corresponding empty allocation pair strategy is obtained according to the individual's chromosome information, so as to obtain the value of each individual corresponding to I _u,g , and finally the value of each individual corresponding to SE _u,g , G _MU , and G _SU .

For example, an individual chromosome sequence is 101001010000110011101111100001011100000000000000. First, we convert this chromosome sequence to quaternary 221100303233201130000000. We let 0 in the quaternary value represent frequency division, 1 represents empty group 1, 2 represents empty group 2, 3 represents empty group 3, so we can get the above-mentioned individual represents the air separation strategy: namely users 5, 6, 8, 14 , 18, 19, 20, 21, 22, 23, 24; users 3, 4, 15, 16 enter null group 1, users 1, 2, 10, 13 enter null group 2, users 7, 9, 11, 12.17 Enter empty group 3. From this, we get the value of I _u,g (due to space limitations, the value of I _u,g is not given here), G _MU =3, G _SU =11. Finally, the frequency division user SE is obtained according to the channel quality of the frequency division user, and the space division user SE is obtained according to the channel quality of the space division user and the correlation between the users in the empty group (due to space limitations, the value of SE _u,g is not here Given).

In S803, the space division user pairing strategy evaluation mechanism is used to obtain the throughput of the cell corresponding to each particle as its fitness; the number of iterations is +1.

In S802, we get the values of I _u,g , SE _u,g , G _MU , and G _SU corresponding to each particle. In S803, we divide each particle I _u,g , SE _u,g , G The values of _MU and G _SU are substituted into the air separation user pairing strategy evaluation mechanism:

In S804, whether the maximum number of iterations is reached.

Determine whether the current number of iterations is equal to the maximum number of iterations, if not, go to S805; if yes, go to S806.

In S805, select, adaptive crossover, adaptive mutation, and return to S802.

% Selection process

p_fitvalue=cumsum(p_fitvalue)%sumsum() represents the accumulation function

fitin=1

newin=1

If ms(newin)<p_fitvalue(fitin)

Newpop(newin)=pop(fitin)

Newin+ = 1

Elsefitin+=1

End If

End While

% Adaptive Cross Process

pc1 = 0.9; pc2 = 0.6% given adaptive crossover probability range: from 60% to 90%

newpop＝pop

pointer=randperm(P)% randperm(P) randomly sorts P integers (1 to P) and outputs

favg=mean(fitvalue)% to obtain the average fitness of P individuals

For i=1:2: P-1% randomly cross P individuals and cross

f_slash=max(fitvalue(pointer(i)), fitvalue(pointer(i+1)))% to obtain the maximum fitness between two individuals

If `f_slash>=favg% get cross probability pc

pc=pc1-(pc1-pc2)*(f_slash-favg)/(Gbestvalue-favg)

Else

pc=pc1

End If

End For

% Adaptive mutation process

pm1 = 0.1; pm2 = 0.001% given adaptive mutation probability range: from 0.1% to 10%

newpop＝pop

favg=mean(fitvalue)% to obtain the average fitness of P individuals

For i=1: P% mutates P individuals sequentially

If fitvalue(i)>=favg% obtain the mutation probability pm

pm＝pm1-(pm1-pm2)*(fitvalue(i)-favg)/(Gbestvalue-favg)

Else

pm＝pm1

End If

If rand()<pm%pm is the crossover probability

Mpoint=round(rand()*(chromlength-1))% find a mutation point randomly

End If

End For

After completing population selection, adaptive crossover, and adaptive mutation, return to S802.

In S806, the empty allocation pair strategy and the time-frequency resource allocation strategy corresponding to the optimal individual are output.

In a second aspect, the present application provides a computer device 9, as shown in FIG. 9, including a memory 91, a processor 92, and a computer program stored on the memory 91 and executable on the processor, and the processor 92 executes the computer program The steps of the method for allocating space-time resources according to any one of the first aspect of the present application are realized.

The computer device 9 proposed in the second aspect of the present application includes a memory 91, a processor 92, and a computer program stored on the memory 91 and executable on the processor 92. When the processor 92 executes the computer program, it is implemented as the first In any aspect, the steps of the time-frequency-space resource allocation method. Therefore, it has all the beneficial effects of the time-frequency space resource allocation method of any of the above embodiments.

A third aspect of the present application proposes a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, a time-frequency-space resource allocation method as in any of the foregoing embodiments is implemented. Therefore, it has all the beneficial effects of the time-frequency space resource allocation method of any of the above embodiments.

It should be noted that in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply these There is any such actual relationship or order between entities or operations. Moreover, the terms "include", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also those not explicitly listed Or other elements that are inherent to this process, method, article, or equipment. In the absence of more restrictions, the elements defined by the sentence "include one..." do not exclude that there are other identical elements in the process, method, article or equipment that includes the elements.

The above are only specific implementation manners of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown in this document, but should conform to the widest scope consistent with the principles and novel features invented in this document.

Claims

A method for allocating space-time resources includes:

Obtain users to be scheduled;

Determine the pairing strategy of the user to be scheduled according to a preset algorithm;

Obtain the evaluation mechanism of the matching strategy;

Obtain the optimal time-frequency resource allocation method corresponding to the matching strategy according to the evaluation mechanism, and score the matching strategy.
The method for allocating space-time resources according to claim 1, the step of obtaining users to be scheduled includes:

Determine the target number of users to be scheduled according to the number of users that can be scheduled in the current transmission time interval and the number of activated users in the current transmission time interval;

The target number of users to be scheduled are selected from the active users in the current transmission time interval according to a preset filtering rule.
The time-frequency-space resource allocation method according to claim 2, the step of determining the target number of users to be scheduled according to the number of users that can be scheduled in the current transmission time interval and the number of active users in the current transmission time interval ,include:

When the number of users that can be scheduled in the current transmission time interval is greater than or equal to the number of active users in the current transmission time interval, the target number is equal to the number of active users in the current transmission time interval;

When the number of users that can be scheduled in the current transmission time interval is less than the number of active users in the current transmission time interval, the target number is equal to the maximum number of users that can be scheduled in the current transmission time interval.
According to the time-frequency-space resource allocation method of claim 2, the preset screening rule is one of the following rules: polling rule, proportional fairness rule, enhanced proportional fairness maximum load-to-interference ratio rule.
The method for allocating space-time resources according to any one of claims 1 to 4,

The preset algorithm is one of the following algorithms: particle swarm optimization, particle swarm optimization with Gaussian mutation, genetic algorithm, and adaptive genetic algorithm.
The method for allocating space-time resources according to any one of claims 1 to 4, the step of obtaining an evaluation mechanism of a pairing strategy includes:

Obtain the optimal time-frequency resource allocation method corresponding to the pairing strategy at the current transmission time interval, and the corresponding maximum available throughput;

Let the maximum throughput be the score of the pairing strategy.
The time-frequency-space resource allocation method according to claim 6, the step of obtaining an optimal time-frequency resource allocation method corresponding to the pairing strategy at the current transmission time interval and corresponding to the maximum throughput obtainable includes:

After the current transmission time interval is given the pairing strategy, use the evaluation mechanism to obtain a time-frequency resource allocation method that maximizes the throughput of the current transmission time interval;

The maximum throughput is obtained according to the time-frequency resource allocation strategy that can maximize the throughput of the current transmission time interval.
A computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, when the processor executes the computer program, any one of claims 1 to 7 is implemented The steps of the time-frequency space resource allocation method described in the item.
A computer-readable storage medium on which a computer program is stored, which when executed by a processor implements the steps of the time-frequency-space resource allocation method according to any one of claims 1 to 7.