KR102389666B1 - Apparatus and method using a decentralized game approach for radio and computing resource allocation in co-located edge computing - Google Patents

Abstract

The present invention relates to a technical idea for managing wireless and computing resources in coexisting edge computing, and a method for allocating wireless and computing resources in coexisting edge computing according to an embodiment has a conflicting relationship in minimizing latency. Formulating the resource allocation problem for two entities as a Generalized Nash Equilibrium Problem (GNEP), converting the formulated GNEP into a Nash Equilibrium Problem (NEP), and allocating resources on a penalty basis for the converted NEP may include steps.

Description

Apparatus and method using a decentralized game approach for radio and computing resource allocation in co-located edge computing

The present invention relates to the technical idea of managing wireless and computing resources in coexistence edge computing, and more particularly, in a new business model described as co-location mobile edge computing, two different entities, Mobile Network Operators (MNO) and CRP (Computing Resource Providers) It relates to the technical idea of efficiently managing the wireless and computing resources owned by it.

Tower sharing between multiple mobile network operators (MNOs) is a popular approach because MNOs can share the physical infrastructure deployed by third-party tower providers to increase coverage to meet growing user demands by reducing capital expenditures.

Wireless network virtualization is the driving force by which radio resources (active infrastructure) and physical devices (passive infrastructure) can be shared between multiple mobile virtual network operators.

Additionally, Computing Resource Providers (CRPs) such as IBM and Vapor can work with tower providers to deploy servers in towers to deliver low-latency applications such as virtual reality and augmented reality to users in real time.

Recently, resource allocation has been an important topic in Multi-access Edge Computing (MEC).

Energy-efficient radio resource allocation and offloading in a multi-cell environment where computing resource allocation in an edge server is not considered is proposed.

The technology proposed by MEC for dynamic task offloading and scheduling IoT service takes into account continuous upload delay. Here, the downlink radio resource allocation is also ignored. Most of the previous work can ignore the queue model in latency calculations. However, according to the prior art, a queue network for providing energy-efficient resources by extending the CPU resources of a server without considering radio resource allocation has been proposed.

Another queuing model for MEC takes into account latency and reliability-aware task offload by controlling the transfer capacity of users and CPU cycles of the MEC server.

Over-the-air transmission and cloud implementation are modeled with the Poisson process in a technique that takes into account the upper bound of delay.

The Generalized Nash Equilibrium Problem (GNEP) has become an important approach in solving resource allocation problems. Formulation of service provisioning in a multi-cloud environment using GNEP can be found in the prior art.

The prior art proposes a GNEP-based algorithm for solving offloading decisions by scheduling users in the MEC.

In order to efficiently solve GNEP, a penalty-based algorithm has been studied in the prior art.

Offloading tasks to the MEC server requires the MNO's radio resources and computing resources to send and receive CRP's data to process the tasks.

In order to solve the problem of difficult resource allocation problem due to the strong coupling between providers, the resource allocation problem is formulated as GNEP, and a penalty-based algorithm is needed to solve the formalized GNEP.

Korean Patent Laid-Open Patent No. 10-2018-0119162 "Platform for Computing at the Mobile Edge" Korean Patent Application Laid-Open No. 10-2018-0112825 "Method and Server for Controlling Relocation of Mec Applications"

An object of the present invention is to efficiently manage wireless and computing resources owned by two different entities, Mobile Network Operators (MNO) and Computing Resource Providers (CRP), in a new business model.

A method of allocating wireless and computing resources in coexistence edge computing according to an embodiment includes formulating a resource allocation problem for two different entities that have conflicting relationships to minimize latency as a Generalized Nash Equilibrium Problem (GNEP); It may include converting the formulated GNEP into a Nash Equilibrium Problem (NEP), and allocating resources based on a penalty for the converted NEP.

The two different entities according to an embodiment are a Mobile Network Operator (MNO) and a Computing Resource Provider (CRP), and the step of formulating it as a GNEP considers the connection relationship between the MNO's radio resource allocation and the CRP's computing resource allocation. Therefore, it may include the step of formulating the resource allocation problem into a GNEP.

According to an embodiment, the step of allocating resources based on the penalty includes receiving initial penalty parameters and resource allocation selections from the two different entities, and predefined self-defined parameters according to the selected initial penalty parameters and resource allocation, respectively. performing the steps of solving an optimization problem, determining whether the self-optimization problem is solved, and if the self-optimization problem is solved as a result of the determination, performing the steps of finding a GNE that terminates the algorithm, and as a result of the determination, the self-optimization problem If not resolved, it may include updating the initial penalty parameters.

The step of determining whether the self-optimization problem is solved according to an embodiment includes determining whether the self-optimization problem is solved by determining whether a coupling constraint on resource allocation is satisfied with respect to the self-optimization problem may include steps.

One of the two different entities according to an embodiment is a Computing Resource Provider (CRP), and the optimization problem for the CRP among the self-optimization problems may be defined by the following [Equation 1].

[Equation 1]

One of the two different entities according to an embodiment is a mobile network operator (MNO), and among the self-optimization problems, the optimization problem for the MNO may be defined by the following [Equation 2].

[Equation 2]

The processor according to an embodiment solves a resource allocation problem for two different entities, a Mobile Network Operator (MNO) and a Computing Resource Provider (CRP), which have a conflicting relationship in minimizing the latency, as a Generalized Nash Equilibrium Problem (GNEP). Formulate, but formulate the resource allocation problem as a GNEP in consideration of the connection relationship between the radio resource allocation of the MNO and the computing resource allocation of the CRP, and transform the formulated GNEP into a Nash Equilibrium Problem (NEP), and the transformed NEP Resources can be allocated on a penalty basis.

In order to allocate resources based on the penalty, the processor receives initial penalty parameters and resource allocations selected from the two different entities, and is predefined according to the selected initial penalty parameters and resource allocations, respectively. solve the self-optimization problem, determine whether the self-optimization problem is solved If not resolved, the initial penalty parameters may be updated.

The processor according to an embodiment determines whether the self-optimization problem is solved by determining whether a coupling constraint on resource allocation is satisfied in relation to the self-optimization problem in order to determine whether the self-optimization problem is solved can be judged

According to an embodiment, wireless and computing resources owned by Mobile Network Operators (MNOs) and Computing Resource Providers (CRPs), which are different entities, may be efficiently managed in a new business model.

1 is a view for explaining a system model according to an embodiment.
2 is a diagram illustrating a method of allocating wireless and computing resources in coexistence edge computing according to an embodiment.
3 is a diagram illustrating a method of allocating resources based on a penalty according to an embodiment.
4 is a diagram illustrating a flowchart of an algorithm for allocating wireless and computing resources in coexistence edge computing.

Specific structural or functional descriptions of the embodiments according to the concept of the present invention disclosed herein are only exemplified for the purpose of explaining the embodiments according to the concept of the present invention, and the embodiment according to the concept of the present invention These may be embodied in various forms and are not limited to the embodiments described herein.

Since the embodiments according to the concept of the present invention may have various changes and may have various forms, the embodiments will be illustrated in the drawings and described in detail herein. However, this is not intended to limit the embodiments according to the concept of the present invention to specific disclosed forms, and includes changes, equivalents, or substitutes included in the spirit and scope of the present invention.

Terms such as first or second may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from other components, for example, without departing from the scope of rights according to the concept of the present invention, a first component may be named a second component, Similarly, the second component may also be referred to as the first component.

When an element is referred to as being “connected” or “connected” to another element, it is understood that it may be directly connected or connected to the other element, but other elements may exist in between. it should be On the other hand, when it is said that a certain element is "directly connected" or "directly connected" to another element, it should be understood that the other element does not exist in the middle. Expressions describing the relationship between elements, for example, “between” and “between” or “directly adjacent to”, etc. should be interpreted similarly.

The terminology used herein is used only to describe specific embodiments, and is not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise. In the present specification, terms such as "comprise" or "have" are intended to designate that the described feature, number, step, operation, component, part, or combination thereof exists, and includes one or more other features or numbers, It should be understood that the possibility of the presence or addition of steps, operations, components, parts or combinations thereof is not precluded in advance.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in a commonly used dictionary should be interpreted as having a meaning consistent with the meaning in the context of the related art, and should not be interpreted in an ideal or excessively formal meaning unless explicitly defined in the present specification. does not

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. However, the scope of the patent application is not limited or limited by these embodiments. Like reference numerals in each figure indicate like elements.

1 is a view for explaining a system model 100 according to an embodiment.

The system model 100 according to an embodiment may be modeled as a Poisson process in consideration of an upper limit of delay in wireless transmission and cloud execution.

The Generalized Nash Equilibrium Problem (GNEP) becomes a prominent approach in solving resource allocation problems because it captures the coupling between players. The formulation of service provisioning in a multi-cloud environment using GNEP can also be found in the system model 100 according to an embodiment.

In the system model 100 according to one embodiment, a GNEP-based algorithm may be used to resolve offloading decisions by reserving users. To this end, the GNEP and its solution approach can be discussed, and the penalty-based algorithm can be studied to solve the GNEP efficiently.

Offloading tasks to the MEC server requires the MNO's radio resources to send and receive data and the CRP's compute resources for task processing. Such strong coupling between providers may make the resource allocation problem difficult, but the system model 100 according to one embodiment formulates the resource allocation problem as a GNEP and operates as a penalty-based algorithm for solving the formulated GNEP.

The system model 100 according to an embodiment may represent a system model of a co-located mobile edge computing system.

To this end, reference numeral 110 denotes a plurality of Mobile Network Operators (MNOs), and reference numeral 120 denotes a Computing Resource Provider (CRP).

According to the system model 100 , a Mobile Network Operator (MNO) and a Computing Resource Provider (CRP) may be co-located in the same tower station 130 .

The MNO of the system model 100 owns radio resources and physical devices necessary to transmit and receive radio signals.

However, space in the tower station 130 and other facilities such as cooling and electricity from the tower provider may be rented. CRP distributes servers in towers by renting space and facilities from tower providers.

When users offload their work to the MEC server, that is, to use the services provided by the CRP, radio resources managed by the MNO and computing resources owned by the CRP are required. This offloading requires three phases: upload phase, compute phase and download phase, each phase can be modeled as a queue.

User u can create a job of speed λu that follows a homogeneous Poisson process. Also, the input file size of this task follows an exponential distribution with mean bu.

이다.In other words, the service rate of the uplink queue, which is the user queue, is the data rate of the uplink transmission divided by the input file size.

am.

비율로 MEC 서버 풀에 도착한다. 이러한 작업의 필수 CPU주기는 평균 cu를 사용하여 지수분포를 따른다.Offloaded tasks are the sum of the service percentages of all user queues.

It arrives at the MEC server pool at a rate. The required CPU cycles for these tasks follow an exponential distribution using the average cu.

이다.The MEC server pool counts this as in the processor sharing approach. The service speed of the MEC server pool is determined by dividing the CPU cycles per second by the required CPU cycles.

am.

비율로 도착한다. 이 결과는 평균 ou와 함께 지수 분포를 따른다. 대응하는 기지국 j는 결과 파일 크기로 나눈 다운 링크 전송의 데이터 속도 인

의 속도로 사용자에게 결과를 다시 전송한다.After the MEC server pool calculates the user's work, the result is sent to that MNO's base station.

arrive in proportion. This result follows an exponential distribution with the mean ou. The corresponding base station j is the data rate of the downlink transmission divided by the resulting file size

Sends the result back to the user at the speed of

2 is a diagram illustrating a method of allocating wireless and computing resources in coexistence edge computing according to an embodiment.

The method of allocating wireless and computing resources in coexistence edge computing according to an embodiment may formulate a resource allocation problem for two different entities that have conflicting relationships to minimize latency as a Generalized Nash Equilibrium Problem (GNEP). There is (step 201).

For example, two different entities may be interpreted as a Mobile Network Operator (MNO) and a Computing Resource Provider (CRP).

In this case, in order to formulate the GNEP, the resource allocation problem can be formulated as the GNEP by considering the connection relationship between the radio resource allocation of the MNO and the computing resource allocation of the CRP.

In addition, the method of allocating wireless and computing resources in coexistence edge computing according to an embodiment may convert a formulated GNEP into a Nash Equilibrium Problem (NEP) (step 202).

In addition, the method of allocating wireless and computing resources in coexistence edge computing according to an embodiment may allocate resources based on a penalty for the converted NEP (step 203).

A method of allocating resources based on a penalty will be described in more detail with reference to FIG. 3 .

3 is a diagram illustrating a method of allocating resources based on a penalty according to an embodiment.

In order to allocate resources based on a penalty, in the method of allocating resources based on a penalty according to an embodiment, initial penalty parameters and resource allocation may be selected from two different entities (step 301).

Next, the method of allocating resources based on a penalty according to an embodiment may solve a predefined self-optimization problem according to the selected initial penalty parameter and resource allocation, respectively (step 302).

Also, in the method of allocating resources based on a penalty according to an embodiment, it may be determined whether a self-optimization problem is solved (step 303).

For example, in order to determine whether the self-optimization problem is solved, it can be determined whether the self-optimization problem is solved by determining whether a coupling constraint on resource allocation is satisfied with respect to the self-optimization problem.

Specifically, among the self-optimization problems used by the method of allocating resources based on a penalty according to an embodiment, the optimization problem for CRP may be defined as the following [Equation 1].

[Equation 1]

In [Equation 1], the components of the delay function can be calculated as follows

In addition, the meaning of each variable in [Equation 1] is as follows.

는 모든 사용자의 종단 간 지연시간인 CRP의 목적함수(objective function)을 의미한다.

denotes the objective function of CRP, which is the end-to-end delay time of all users.

는 업링크 자원 할당 벡터를 의미한다.

denotes an uplink resource allocation vector.

는 기지국(Base Station) j를 의미한다.

denotes a base station j.

는 업링크 및 컴퓨팅 리소스 할당과 관련된 컴퓨팅 리소스 대기열의 안정성을 의미한다.

Means the stability of the computing resource queues related to the uplink and computing resource allocation.

는 CRP의 컴퓨팅 리소스의 한계를 정하는 CRP의 전략을 의미한다.

means CRP's strategy to set the limits of CRP's computing resources.

p has a CRP of 0 and an MNO of 1,… , may indicate the index of the player who is N.

Among the self-optimization problems used by the method of allocating resources based on a penalty according to an embodiment, the optimization problem for the MNO may be defined as the following [Equation 2].

Among the self-optimization problems, the optimization problem for the MNO may be defined by the following [Equation 2].

[Equation 2]

In [Equation 2], the components of the delay function can be calculated as follows

p has a CRP of 0 and an MNO of 1,… , may indicate the index of the player who is N.

In the method of allocating resources based on a penalty according to an embodiment, when the self-optimization problem is resolved as a result of the determination, the step of finding a GNE that terminates the algorithm is performed, and if the self-optimization problem is not resolved as a result of the determination, the initial penalty parameter can be updated (step 304).

4 is a diagram illustrating a flowchart of an algorithm for allocating radio and computing resources in coexistence edge computing.

Each player, such as CRP and MNO, chooses initial penalty parameters and resource allocation. Based on the selected parameters, the self-optimization problem can be solved as defined in [Equation 1] and [Equation 2].

If the joint constraint for optimal resource allocation is satisfied, then the GNE that terminates the algorithm is found. Otherwise, the player must update the penalty parameter and solve the optimization problem until the coupling constraint is met.

According to an embodiment of the present invention, resource allocation of Mobile Edge Computing includes resource management of two different entities, several service providers having conflicting interests.

Efficiently manage wireless and computing resources to provide services with minimal latency with conflicting interests between mobile network operators and computing resource providers.

For the commercialization case, the proposed scheme can be implemented for a new business model of co-located mobile edge computing, where the mobile network operator is responsible for allocating radio resources and the computing resource provider controls the computing resources.

The algorithm of FIG. 4 may be operated by a processor of a wireless and computing resource allocating apparatus operating at least one of a computer program, code, and instruction.

In addition, the algorithm can be divided into operation 410 in CRP and operation 420 in MNO, and operation 410 in CRP and operation 420 in MNO are not dependent on each other and can be processed independently. there is.

For convenience of explanation, first looking at the operation in the CRP, the CRP may select an initial penalty parameter and resource allocation (411).

For example, the initial penalty parameter may be expressed as follows.

와 같이 표현될 수 있다.In addition, an initialization point for resource allocation is

can be expressed as

Next, in CRP, the penalty problem can be solved by using the optimization problem for CRP (412).

As an example, the optimization problem for CRP may be defined through [Equation 1] described above.

Meanwhile, the processor may determine whether the self-optimization problem is solved through the algorithm ( 430 ).

To this end, the processor may determine that the self-optimization problem is solved when both conditions of [Equation 3] are satisfied.

[Equation 3]

In [Equation 3]

는 최적의 다운링크 및 컴퓨팅 자원 할당에 대한 다운링크 통신 자원 대기열의 안정성을 의미하는 것이고,

는 최적의 업링크 및 컴퓨팅 자원 할당에 대한 컴퓨팅 자원 대기열의 안정성을 의미한다.

is the stability of the downlink communication resource queue for optimal downlink and computing resource allocation,

Means the stability of the computing resource queue for optimal uplink and computing resource allocation.

If the self-optimization problem is not resolved as a result of determination at reference numeral 430 , the initial penalty parameter may be updated ( 413 ).

In addition, if the self-optimization problem is solved as a result of determination at reference numeral 430 , a step of finding a GNE for terminating the algorithm may be performed ( 440 ).

와 같이 결정될 수 있다.If [Equation 3] is satisfied, GNE is

can be determined as

Next, first looking at the operation of the MNOs, the MNO may be selected for an initial penalty parameter and resource allocation ( 421 ).

Next, the MNOs may solve the penalty problem by using the optimization problem for the MNO ( 422 ).

Meanwhile, the processor may determine whether the self-optimization problem is solved through the algorithm ( 430 ).

If the self-optimization problem is not resolved as a result of determination at reference numeral 430 , the initial penalty parameter may be updated ( 423 ).

Specifically, the CRP may update the initial penalty parameter as shown in [Equation 4], and the MNO may update the initial penalty parameter as shown in [Equation 5].

[Equation 4]

[Equation 5]

와 같이 결정될 수 있다.In addition, if the self-optimization problem is solved as a result of determination at reference numeral 430, a step of finding a GNE for terminating the algorithm may be performed (440). Similarly, when [Equation 3] is satisfied, GNE is

can be determined as

As a result, by using the present invention, wireless and computing resources owned by different entities, Mobile Network Operators (MNOs) and Computing Resource Providers (CRPs), can be efficiently managed in a new business model.

The device described above may be implemented as a hardware component, a software component, and/or a combination of the hardware component and the software component. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For convenience of understanding, although one processing device is sometimes described as being used, one of ordinary skill in the art will recognize that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that can include For example, the processing device may include a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

Software may comprise a computer program, code, instructions, or a combination of one or more thereof, which configures a processing device to operate as desired or is independently or collectively processed You can command the device. The software and/or data may be any kind of machine, component, physical device, virtual equipment, computer storage medium or apparatus, to be interpreted by or to provide instructions or data to the processing device. , or may be permanently or temporarily embody in a transmitted signal wave. The software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored in one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floppy disks. - includes magneto-optical media, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

As described above, although the embodiments have been described with reference to the limited drawings, various modifications and variations are possible by those skilled in the art from the above description. For example, the described techniques are performed in an order different from the described method, and/or the described components of the system, structure, apparatus, circuit, etc. are combined or combined in a different form than the described method, or other components Or substituted or substituted by equivalents may achieve an appropriate result.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (11)

A method of allocating wireless and computing resources in coexistence edge computing, the method comprising:
In a wireless and computing resource allocation device operating as a processor, formulating a resource allocation problem for two different entities as a Generalized Nash Equilibrium Problem (GNEP);
converting the formulated GNEP into a Nash Equilibrium Problem (NEP) in the device for allocating wireless and computing resources; and
Allocating, in the device for allocating wireless and computing resources, resources on a penalty basis for the converted NEP
including,
Allocating resources based on the penalty comprises:
receiving initial penalty parameters and resource allocation selections from the two different entities;
solving each predefined self-optimization problem according to the selected initial penalty parameter and resource allocation;
determining whether the self-optimization problem is solved; and
If the self-optimization problem is solved as a result of the determination, performing the step of finding a GNE that terminates the algorithm, and if the determination result does not solve the self-optimization problem, updating the initial penalty parameter;
A method of allocating wireless and computing resources in coexisting edge computing, comprising: According to claim 1,
The two different entities are
MNO (Mobile Network Operator) and CRP (Computing Resource Provider),
The steps to formalize as GNEP are:
Formulating the resource allocation problem as a GNEP by considering the link between MNO's radio resource allocation and CRP's computing resource allocation
A method of allocating wireless and computing resources in coexisting edge computing, comprising: delete According to claim 1,
The step of determining whether the self-optimization problem is solved is,
In relation to the self-optimization problem, determining whether the self-optimization problem is solved by determining whether a coupling constraint on resource allocation is satisfied
A method of allocating wireless and computing resources in coexisting edge computing, comprising: According to claim 1,
One of the two different entities,
CRP (Computing Resource Provider),
Among the self-optimization problems, the optimization problem for the CRP is a method of allocating wireless and computing resources in coexistence edge computing, characterized in that the following [Equation 1] is defined.

[Equation 1]

In [Equation 1],

denotes the objective function of CRP, which is the end-to-end delay time of all users,

is the uplink resource allocation vector,

means a base station j,

means the stability of the compute resource queues related to uplinks and compute resource allocation,

denotes CRP's strategy to set the limits of CRP's computing resources, p is CRP is 0 and MNO is 1,… , indicating the index of the player with N According to claim 1,
One of the two different entities,
MNO (Mobile Network Operator),
Among the self-optimization problems, the optimization problem for the MNO is a method of allocating wireless and computing resources in coexistence edge computing, characterized in that the following [Equation 2] is defined.

[Equation 2]

In [Equation 2],

denotes the objective function of CRP, which is the end-to-end delay time of all users,

is the uplink resource allocation vector,

means a base station j,

means the stability of the compute resource queues related to uplinks and compute resource allocation,

denotes CRP's strategy to set the limits of CRP's computing resources, p is CRP is 0 and MNO is 1,… , indicating the index of the player with N
In the wireless and computing resource allocation apparatus comprising a processor operating as a computer program (computer program) consisting of a plurality of codes (code) or instructions (instruction),
The processor is
The resource allocation problem for two different entities, the Mobile Network Operator (MNO) and the Computing Resource Provider (CRP), is formulated as the Generalized Nash Equilibrium Problem (GNEP), but the connection between the radio resource allocation of the MNO and the computing resource allocation of the CRP Formulating a resource allocation problem into a GNEP in consideration of the relationship, transforming the formulated GNEP into a Nash Equilibrium Problem (NEP), and allocating resources based on a penalty for the transformed NEP,
The processor is
In order to allocate resources based on the penalty,
receive initial penalty parameters and resource allocations selected from the two different entities, solve predefined self-optimization problems according to the selected initial penalty parameters and resource allocations, respectively, and determine whether the self-optimization problems are resolved; When the self-optimization problem is solved as a result of the determination, the step of finding a GNE that terminates the algorithm is performed, and if the self-optimization problem is not resolved as a result of the determination, the initial penalty parameter is updated. allocation device. delete 8. The method of claim 7,
The processor is
To determine whether the self-optimization problem has been resolved,
In relation to the self-optimization problem, the wireless and computing resource allocation apparatus, characterized in that it is determined whether the self-optimization problem is solved by determining whether a coupling constraint on resource allocation is satisfied. 8. The method of claim 7,
Among the self-optimization problems, the optimization problem for the CRP is a wireless and computing resource allocation device, characterized in that defined by the following [Equation 1].

[Equation 1]

In [Equation 1],

denotes the objective function of CRP, which is the end-to-end delay time of all users,

is the uplink resource allocation vector,

means a base station j,

means the stability of the compute resource queues related to uplinks and compute resource allocation,

denotes CRP's strategy to set the limits of CRP's computing resources, p is CRP is 0 and MNO is 1,… , indicating the index of the player with N 8. The method of claim 7,
Among the self-optimization problems, the optimization problem for the MNO is a wireless and computing resource allocation device, characterized in that defined by the following [Equation 2].

[Equation 2]

In [Equation 2],

denotes the objective function of CRP, which is the end-to-end delay time of all users,

is the uplink resource allocation vector,

means a base station j,

means the stability of the compute resource queues related to uplinks and compute resource allocation,

denotes CRP's strategy to set the limits of CRP's computing resources, p is CRP is 0 and MNO is 1,… , indicating the index of the player with N

Images