A joint allocation algorithm for computing and storage resources in satellite communication networks

Abstract

The present invention discloses a joint allocation algorithm for computing and storage resources for satellite communication networks, and relates to the field of satellite mobile communication and edge computing technology. In the present invention, the problem of limited computing power and battery capacity of satellite terminals and inability to perform large-scale computing tasks is solved through the joint allocation of communication, computing and storage resources, and the service capability of satellites to terminals is further improved. The three-dimensional resources including communication resources, storage capacity and server computing frequency are reasonably utilized to provide support for communication satellite systems in major national plans such as space-ground integration and satellite Internet; the problem of joint allocation of communication, computing and storage resources of mobile edge computing networks that integrate satellites and 6G is solved, which can effectively reduce the execution delay of terminal computing tasks, while taking into account the fairness of terminal service performance, and ensuring that computing services are completed within the tolerance range of energy consumption.

Description

Technical Field

[0001] The present invention relates to the field of satellite mobile communications and edge computing technology, and in particular to a joint allocation algorithm for computing and storage resources for satellite communication networks. Background Art

[0002] Driven by the trend of high throughput, high computing volume and global connectivity of future mobile communication networks, the demand for integrated applications of high-performance computing, storage and transmission is increasing. In order to provide services to remote areas and sea and air terminals that are not covered by ground cellular networks, satellite communication networks can serve as a supplement to ground mobile communication systems and provide seamless services to terminals around the world. Due to the limited information processing capabilities of mobile terminals, it is challenging to carry out high-computing applications locally in the terminals. In order to cope with this problem, mobile edge computing has become a powerful technical approach, which migrates computing tasks to computing servers at the edge, thereby reducing energy loss of mobile terminals and reducing task execution latency.

[0003] Although there have been some studies on the application of mobile edge computing technology to satellite communication networks, the joint allocation of communication, computing and storage resources in satellite mobile edge computing networks has not been studied. Since data transmission performance is no longer the main focus of future 6G networks, future information services include diversified demands for communication, computing and storage resources, and the three dimensions of resources are coupled with each other, which will jointly affect the user's service experience. In addition, with the continuous development of chip and manufacturing technology, the capabilities of satellite platforms are rapidly improving. In addition to data communication capabilities, in the future, satellites will also have computing and storage capabilities, which can provide comprehensive services of communication, computing and storage for terminals that are not accessible by ground networks. Therefore, it is very necessary to design a joint allocation algorithm for communication, computing and storage resources for satellite mobile edge computing networks. Summary of the invention

[0004] The purpose of the present invention is to propose a joint allocation algorithm for computing and storage resources in a satellite communication network in order to solve the above problems.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A satellite communication network-oriented computing and storage resource joint allocation algorithm comprises the following steps:

[0007] S1: Initialize the parameter configuration of the terminal, satellite edge computing server, and mission model;

[0008] S2: Establish communication, computing, and storage models to quantify the costs of local computing and edge computing, including time and energy costs;

[0009] S3: build cache constraints;

[0010] S4: Construct the objective function of resource allocation based on the communication, computing and storage models;

[0011] S5: According to the objective function, a heuristic algorithm is designed to obtain the computing task offloading and database downloading strategies of each terminal in steps, and the database cache strategy is obtained through further calculation. Finally, the communication resource allocation and computing resource allocation strategies between the terminal and the satellite are obtained based on the convex optimization algorithm.

[0012] Preferably, the parameter configuration of the initialization terminal, the satellite edge computing server, and the task model in step S1 includes:

[0013] A satellite mobile edge computing network consists of J terminals and N visible satellites. Each satellite is deployed with a satellite mobile edge computing server to execute computing tasks from the terminals.

[0014] Preferably, the specific method for quantifying the overhead in step S2 is:

[0015] For local computing, the required database needs to be downloaded from the satellite mobile edge computing server of the satellite. Considering that the downlink transmission between the satellite and the ground terminal adopts orthogonal multiple access, the data transmission rate of terminal j downloading the database from satellite n is:

[0016]

[0017] where y jn represents the normalized downlink communication resources allocated by satellite n to terminal j, B d Indicates the total downlink communication resources, represents the channel gain, σ 2 represents the noise power, so the time required for terminal j to complete the local calculation process can be calculated as:

[0018]

[0019] in In addition, the energy consumed by terminal j is:

[0020]

[0021] where f j represents the communication capability of terminal j, κ represents the energy consumption coefficient, which mainly depends on the chip structure. For edge computing, the data transmission rate of terminal j uploading its mission data to satellite n can be expressed as:

[0022]

[0023] where x jnrepresents the normalized uplink wireless resources allocated by satellite n to terminal j. Since the amount of data of the calculation result is usually small, the time for terminal j to complete the entire edge calculation is:

[0024]

[0025] in represents the computing resources allocated by satellite n to terminal j, It represents the one-way propagation delay. In addition, the energy consumption required for edge computing at terminal j can be expressed as:

[0026]

[0027] Preferably, the method for constructing the cache constraint in step S3 is:

[0028] Since the storage capacity of the satellite is limited, the total amount of database deployed on the satellite mobile edge computing server and the task data migrated by the terminal cannot exceed the storage capacity of the satellite. For satellite n, there are the following constraints:

[0029]

[0030] In addition, only when the satellite deploys the database of a certain task, the terminal can migrate the computing task to the satellite or download the database from the satellite. Therefore, for user j, task k and satellite n, there are the following constraints:

[0031]

[0032] Preferably, the objective function construction method of resource allocation in step S4 is:

[0033] In order to minimize the computing delay of all terminals, the joint optimization problem of communication, computing and storage resources is modeled as:

[0034]

[0035]

[0036]

[0037]

[0038]

[0039]

[0040]

[0041]

[0042]

[0043]

[0044]

[0045] In formula (9), C1-C3 are energy consumption, computation and cache constraints respectively. is the battery capacity reserved for the computing task of terminal j, and They represent the computing and caching capabilities of the nth satellite respectively. C4 means that each device can only be associated with one satellite for computing offloading or database downloading. C5 represents the database placement constraint. C6-C10 are common resource allocation constraints.

[0046] Preferably, the specific method of resource allocation in step S5 is:

[0047] The objective function in formula (9) can be reformulated as:

[0048]

[0049] in From formula (10), we can see that the allocation of wireless resources and satellite computing resources is Computation offloading and database download decisions (α jn ,β jn ) are coupled with each other. In this case, the low-complexity heuristic algorithm is divided into three parts. First, the computation offloading and database downloading decisions are obtained by assuming average resource allocation, then the database deployment is determined, and finally the optimal wireless resource and satellite computing resource allocation is obtained.

[0050] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0051] 1. This application solves the problem of joint allocation of communication, computing and storage resources in the mobile edge computing network that integrates satellite and 6G. It can effectively reduce the execution delay of terminal computing tasks, while taking into account the fairness of terminal service performance and ensuring that computing services are completed within the tolerance range of energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A schematic diagram of the execution process of a satellite communication network-oriented general computing and storage resource joint allocation algorithm according to an embodiment of the present invention is shown;

[0053] Figure 2 A schematic diagram of a system model provided according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] See also Figure 1-2 , the present invention provides a technical solution:

[0056] A satellite communication network-oriented computing and storage resource joint allocation algorithm comprises the following steps:

[0057] S1: Initialize the parameter configuration of the terminal, satellite edge computing server, and mission model;

[0058] S2: Establish communication, computing, and storage models to quantify the costs of local computing and edge computing, including time and energy costs;

[0059] S3: build cache constraints;

[0060] S4: Construct the objective function of resource allocation based on the communication, computing and storage models;

[0061] S5: According to the objective function, a heuristic algorithm is designed to obtain the computing task offloading and database downloading strategies of each terminal in steps, and the database cache strategy is obtained through further calculation. Finally, the communication resource allocation and computing resource allocation strategies between the terminal and the satellite are obtained based on the convex optimization algorithm.

[0062] Specifically, Figure 1 and Figure 2 As shown, the parameter configuration of the initialization terminal, satellite edge computing server, and task model in step S1 includes:

[0063] A satellite mobile edge computing network consists of J terminals and N visible satellites. Each satellite deploys a satellite mobile edge computing server to execute computing tasks from the terminals. Assume that the task model is expressed as (a k ,b k ,ε k ), where a k Indicates data from the terminal, b k represents the required database, which is deployed in the satellite mobile edge computing server, ε k represents the number of CPU cycles required to compute task k, ξ jk ∈{0,1} indicates whether task k is generated by terminal j. Assuming that each terminal currently has only one computing task, that is, Let α jn ∈{0,1} indicates whether the task from terminal j is computed locally, β jn ∈{0,1} indicates whether the database required for the task of terminal j is downloaded via satellite n.

[0064] The specific method for quantifying the overhead in step S2 is:

[0065] For local computing, the required database needs to be downloaded from the satellite mobile edge computing server of the satellite. Considering that the downlink transmission between the satellite and the ground terminal adopts orthogonal multiple access, the data transmission rate of terminal j downloading the database from satellite n is:

[0066]

[0067] where y jn represents the normalized downlink communication resources allocated by satellite n to terminal j, B d Indicates the total downlink communication resources, represents the channel gain, σ 2 represents the noise power, so the time required for terminal j to complete the local calculation process can be calculated as:

[0068]

[0069] in In addition, the energy consumed by terminal j is:

[0070]

[0071] where f j represents the communication capability of terminal j, κ represents the energy consumption coefficient, which mainly depends on the chip structure. For edge computing, the data transmission rate of terminal j uploading its mission data to satellite n can be expressed as:

[0072]

[0073] where x jn represents the normalized uplink wireless resources allocated by satellite n to terminal j. Since the amount of data of the calculation result is usually small, the time for terminal j to complete the entire edge calculation is:

[0074]

[0075] in represents the computing resources allocated by satellite n to terminal j, It represents the one-way propagation delay. In addition, the energy consumption required for edge computing at terminal j can be expressed as:

[0076]

[0077] The method for constructing the cache constraint in step S3 is:

[0078] Since the storage capacity of the satellite is limited, the total amount of database deployed on the satellite mobile edge computing server and the task data migrated by the terminal cannot exceed the storage capacity of the satellite. For satellite n, there are the following constraints:

[0079]

[0080] In addition, only when the satellite deploys the database of a certain task, the terminal can migrate the computing task to the satellite or download the database from the satellite. Therefore, for user j, task k and satellite n, there are the following constraints:

[0081]

[0082] The objective function construction method of resource allocation in step S4 is:

[0083] In order to minimize the computing delay of all terminals, the joint optimization problem of communication, computing and storage resources is modeled as:

[0084]

[0085]

[0086]

[0087]

[0088]

[0089]

[0090]

[0091]

[0092]

[0093]

[0094]

[0095] In formula (9), C1-C3 are energy consumption, computation and cache constraints respectively. is the battery capacity reserved for the computing task of terminal j, and They represent the computing and caching capabilities of the nth satellite, respectively. C4 means that each device can only be associated with one satellite for computing offloading or database downloading. C5 represents the database placement constraint. C6-C10 are common resource allocation constraints.

[0096] The specific method of resource allocation in step S5 is:

[0097] The objective function in formula (9) can be reformulated as:

[0098]

[0099] in From formula (10), we can see that the allocation of wireless resources and satellite computing resources is Computation offloading and database download decisions (α jn ,β jn ) are coupled with each other. In this case, the low-complexity heuristic algorithm is divided into three parts. First, the computation offloading and database downloading decisions are obtained by assuming average resource allocation, then the database deployment is determined, and finally the optimal wireless resource and satellite computing resource allocation is obtained;

[0100] Specifically, we first assume an average wireless resource and satellite computing resource allocation, i.e. in is defined as the number of devices, except device j, that decide to offload tasks to the nth satellite or download the database from the nth satellite. Therefore, in order to minimize the objective function while considering the load balancing among satellites, the computation offloading and database downloading decisions of each device are obtained sequentially, namely:

[0101]

[0102]

[0103] And update it in each round and

[0104] Then, considering the limitation of satellite storage capacity, the database corresponding to the kth task should be cached only on the satellite that provides edge computing or database download services for task k, that is:

[0105]

[0106] Finally, substituting the obtained (α', β', s') into formula (9), we can get:

[0107]

[0108] Since formula (13) is a convex problem, it can be solved using a standard convex optimization algorithm.

[0109] This application solves the problem of joint allocation of communication, computing and storage resources in the mobile edge computing network that integrates satellite and 6G. It can effectively reduce the execution delay of terminal computing tasks, while taking into account the fairness of terminal service performance and ensuring that computing services are completed within the tolerance range of energy consumption.

[0110] The above description of the embodiments enables those skilled in the art to implement or use the present invention. 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 invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A joint allocation algorithm for computing and storage resources in satellite communication networks, characterized in that: The following steps are involved: S1: Initialize the parameter configuration of the terminal, satellite edge computing server, and mission model; S2: Establish communication, computing, and storage models to quantify the costs of local computing and edge computing, including time and energy costs; S3: build cache constraints; S4: Construct the objective function of resource allocation based on the communication, computing and storage models; The method for constructing the objective function of resource allocation in step S4 is: In order to minimize the computing delay of all terminals, the joint optimization problem of communication, computing and storage resources is modeled as: In formula (9), C1-C3 are energy consumption, computation and cache constraints respectively. is the battery capacity reserved for the computing task of terminal j, and They represent the computing and caching capabilities of the nth satellite, respectively. C4 means that each device can only be associated with one satellite for computing offloading or database downloading. C5 represents the database placement constraint. C6-C10 are common resource allocation constraints. S5: According to the objective function, a heuristic algorithm is designed to obtain the computing task offloading and database downloading strategies of each terminal in steps, and the database cache strategy is obtained through further calculation. Finally, the communication resource allocation and computing resource allocation strategies between the terminal and the satellite are obtained based on the convex optimization algorithm. The specific method of resource allocation in step S5 is: The objective function in formula (9) is reformulated as: in From formula (10), we can see that the allocation of wireless resources and satellite computing resources is Computation offloading and database download decisions (α jn ,β jn ) are coupled with each other. In this case, the low-complexity heuristic algorithm is divided into three parts. First, the computation offloading and database downloading decisions are obtained by assuming average resource allocation, then the database deployment is determined, and finally the optimal wireless resource and satellite computing resource allocation is obtained.

2. The satellite communication network-oriented computing and storage resource joint allocation algorithm according to claim 1, characterized in that: The parameter configuration of the initialization terminal, the satellite edge computing server, and the task model in step S1 includes: A satellite mobile edge computing network consists of J terminals and N visible satellites. Each satellite is deployed with a satellite mobile edge computing server to execute computing tasks from the terminals.

3. The satellite communication network-oriented computing and storage resource joint allocation algorithm according to claim 1, characterized in that: The specific method for quantifying the overhead in step S2 is: For local computing, the required database needs to be downloaded from the satellite mobile edge computing server of the satellite. Considering that the downlink transmission between the satellite and the ground terminal adopts orthogonal multiple access, the data transmission rate of terminal j downloading the database from satellite n is: where y jn represents the normalized downlink communication resources allocated by satellite n to terminal j, B d Indicates the total downlink communication resources, represents the channel gain, σ 2 represents the noise power, so the time required for terminal j to complete the local calculation process can be calculated as: in In addition, the energy consumed by terminal j is: where f j represents the communication capability of terminal j, κ represents the energy consumption coefficient, which mainly depends on the chip structure. For edge computing, the data transmission rate of terminal j uploading its mission data to satellite n can be expressed as: where x jn represents the normalized uplink wireless resources allocated by satellite n to terminal j. Since the amount of data of the calculation result is usually small, the time for terminal j to complete the entire edge calculation is: in represents the computing resources allocated by satellite n to terminal j, It represents the one-way propagation delay. In addition, the energy consumption required for edge computing at terminal j can be expressed as:

4. The satellite communication network-oriented computing and storage resource joint allocation algorithm according to claim 1, characterized in that: The method for constructing the cache constraint in step S3 is: Since the storage capacity of the satellite is limited, the total amount of database deployed on the satellite mobile edge computing server and the task data migrated by the terminal cannot exceed the storage capacity of the satellite. For satellite n, there are the following constraints: In addition, only when the satellite deploys the database of a certain task, the terminal can migrate the computing task to the satellite or download the database from the satellite. Therefore, for user j, task k and satellite n, there are the following constraints:

Images