Method and system for managing and allocating institutional resources based on cloud computing

Abstract

The present invention discloses a method and system for managing and allocating agency resources based on cloud computing, including: configuring a cloud server based on the agency resource type and structure, encoding and compressing resource data and aggregating the data, storing the data in the cloud server, performing association analysis to establish a topological relationship network, allocating an edge cloud server to a terminal user, configuring a program for adding, deleting, modifying, and checking to obtain a terminal user entity, building a key for the terminal user entity, encrypting the data and index of the terminal user, outsourcing the encrypted data to the cloud server, obtaining a hybrid private cloud, obtaining a resource request, calculating the request priority of the resource index in the hybrid private cloud, setting the index order based on the request priority, and designing a coordination algorithm based on the index order to manage and allocate resources. The method utilizes cloud computing to achieve efficient management and allocation of agency resources, improve security and resource utilization, and has good interpretability.

Description

Technical Field

[0001] The present invention relates to the field of resource management, and more particularly to a method and system for managing and allocating agency resources based on cloud computing. Background Art

[0002] In today's digital age, government agencies face many challenges in digital management and allocation of resources.

[0003] Traditional resource management methods often rely on manual records and paper documents, which have problems such as untimely, inaccurate, and difficult to share information. At the same time, with the continuous expansion and complexity of government business, the types and quantities of resources are also increasing. Traditional management methods can no longer meet the needs of efficient management and reasonable allocation of resources.

[0004] With the rapid development of cloud computing technology, it provides a new solution for the management and allocation of government resources. Through the cloud computing platform, government agencies can digitally integrate various resources, break information silos, and realize resource information sharing and collaborative management. At the same time, the elastic expansion characteristics of cloud computing can dynamically adjust resource allocation according to the actual needs of government agencies, improve resource utilization, and reduce management costs. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for managing and allocating agency resources based on cloud computing.

[0006] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0007] The first aspect of the present invention provides a method for managing and deploying agency resources based on cloud computing, comprising:

[0008] Determine the type of agency resources and configure cloud servers based on the agency resource type and agency structure;

[0009] After data encoding and compression are performed on the resource data, aggregation is performed based on data category and time series to obtain a first aggregation class set and a second aggregation class set, which are stored in the cloud server;

[0010] Based on the association analysis between the first aggregation class set, the second aggregation class set and the terminal user, a topological relationship network is established, and an edge cloud server is allocated to the terminal user in the topological relationship network;

[0011] Configure the edge cloud server with a program for adding, deleting, modifying and checking to obtain a terminal user entity;

[0012] Constructing a key for the end user entity, using the key to encrypt the end user's data and index, and outsourcing it to the cloud server to obtain a hybrid private cloud;

[0013] Obtaining a resource request, setting a space stack for storing a resource index, calculating a request priority of the resource index, and setting an index order in the hybrid private cloud based on the request priority;

[0014] Design a coordination algorithm based on the index sequence to manage and allocate resources.

[0015] As a further method, the method for configuring the cloud server based on the agency architecture and agency resource type includes:

[0016] Design the data warehouse structure. Determine the dimensions of the data warehouse according to the organizational structure. For each dimension, design the dimension table structure, including: primary key, dimension attributes, and hierarchy. If there is an association between dimension tables, use foreign keys to establish connections.

[0017] Deploy a data warehouse on a cloud server, use block storage to store resource data that requires frequent random read and write operations, and use object storage to store unstructured resource data that does not require frequent random read and write operations, and store the resource data in the data warehouse.

[0018] As a further method, the method of performing aggregation based on data sources and time series to obtain a first aggregation class set and a second aggregation class set, and storing them in the cloud server includes:

[0019] Divide the sources of resource data, including but not limited to different departments, different business systems and external cooperation organizations, assign a unique identification code to each source, and combine resource data from the same data source into one category to form multiple aggregation categories, thereby obtaining a first aggregation category set;

[0020] Record the creation time, update time and access time of the resource data, set the time granularity, group the time intervals according to the time granularity, obtain multiple aggregation classes based on the creation time, update time and access time, and obtain a second aggregation class set;

[0021] The storage memory of the cloud server is divided into multiple storage nodes to form a distributed storage network. The hash calculation is performed based on the unique identification code and timestamp of the resource data to determine the node location of the data storage. The expression is:

[0022] Node=(MD5(((SHA256(ID×T 2 )+R)×W)))mod N

[0023] Among them, MD5(·) represents the message digest algorithm, SHA256(·) represents the secure hash algorithm SHA-256, ID is the unique identification code of the resource data, T is the timestamp of the resource data, R is the balance factor, W is the access frequency of the resource data, mod represents the modulus operation, and N is the total number of storage nodes.

[0024] As a further method, the method of establishing a topological relationship network based on association analysis between the first aggregation class set, the second aggregation class set and the terminal user includes:

[0025] For the first aggregated set, an index is created with the data source as the key. For the second aggregated set, the creation time is used as the primary index key, and the update time and access time are used as the secondary index key. Hierarchical indexing is performed according to the time granularity, and the indexes are combined to obtain a multi-dimensional index.

[0026] Analyze the association between the data source in the first aggregation class set and the time feature in the second aggregation class set according to the multi-dimensional index, and analyze the association between the terminal user and the resource data set;

[0027] The data sources, time nodes, and terminal users are taken as nodes in the network. According to the results of association analysis, the connection relationship and weight between the nodes are determined to obtain a topological relationship network.

[0028] As a further method, the method for allocating an edge cloud server to a terminal user in the topological relationship network includes:

[0029] A data sharing mechanism is set up between different edge cloud servers, specifically: based on the connection relationship between nodes in the topological network, a distributed storage system is used to store data, and data sharing between different edge cloud servers is ensured through a data synchronization protocol;

[0030] Set up a firewall security system for the edge cloud server, specifically: when there is abnormal behavior, block the IP address of the edge cloud server and suspend the terminal user's permissions;

[0031] Add user permissions for each edge cloud server to access the central cloud server, and authenticate each edge server on the central cloud server. The expression for batch authentication of the central cloud server is:

[0032]

[0033] Among them, N * is the total number of edge servers, Auth(ESC i ,C) is the authentication function. When there is c in the authentication information set C of the central cloud server k, making the user permission set UP i With c k If the matching degree is greater than or equal to T, it returns true, indicating that the i-th edge cloud server has passed the identity authentication, otherwise it returns false, ESC i is the ith edge cloud server, C is the authentication information set of the central cloud server, c k is an authentication information in C, UP i is the user permission set of the ith edge cloud server, t is the time point of identity authentication, t0 is the time point of the last update of the edge server, and T1 is the matching standard threshold;

[0034] A buffer mechanism is set between the edge cloud server and the central cloud server. Specifically, when the communication transmission is completed, the edge cloud server and the central cloud server interrupt the transmission control protocol by closing the program.

[0035] As a further method, the method of configuring the add, delete, modify and query program to the edge cloud server includes:

[0036] When adding, deleting, and modifying resource data, after the terminal user submits the operation request, the edge cloud server performs the corresponding operation. After completion, the edge cloud server transmits it to the central cloud server for verification, storage, and index update, and records the operation log in the data warehouse and forwards it to the central cloud server;

[0037] When querying resource data, the user initiates a request and enters keywords and filter conditions. The central cloud server uses the inverted index to query and returns the results.

[0038] The difference storage mechanism is used to manage the historical versions of resource data and store them in the data warehouse, where the expression for difference calculation is:

[0039]

[0040] Where, ΔD v is the difference set between the vth version and the v-1th version of the resource data, pos is the position index of the change in the resource data, change is the change content at the pos position, and w p (pos) is the position weight function, It is represented as the edit distance between the lth attribute in the resource data in version v-1 and version v. It is expressed as the Manhattan distance between the lth attribute in the resource data in version v-1 and version v, thr is the standard deviation of the change of all data points between version v-1 and version v, m and n are the maximum number of record rows and the maximum number of record columns in the dimension table in the data warehouse, respectively.

[0041] As a further method, the method of constructing a key for the end-user entity comprises:

[0042] Construct the initial key based on the elliptic curve signature algorithm;

[0043] The complexity measure for calculating the initial key is:

[0044]

[0045] Among them, β is the information entropy weight coefficient, k is the number of different characters in the initial key, and f j is the frequency of the jth character in the initial key, γ is the weight coefficient of character distribution uniformity, is the average frequency of occurrence of different characters, α is the variance weight coefficient;

[0046] Simulate the side channel attack and calculate the difficulty coefficient of the successful attack on the initial key. The expression is:

[0047]

[0048] Among them, λ is the resource weight coefficient, and C cpu The maximum CPU load that can be used for side channel attacks and the current CPU load used for side channel attacks, respectively. and C gpu are the maximum GPU load that can be used for side channel attacks and the current GPU load used for side channel attacks, respectively. and M memory are the maximum memory size that can be used for side channel attacks and the current memory size used for side channel attacks, μ is the time weight coefficient, T attack The time required to successfully attack the initial key;

[0049] To strengthen the key strength of the initial key, the expression is:

[0050]

[0051] Among them, Q0 is the initial key strength, E(·) is the encryption algorithm of the key derivation function, and ξ represents the initial key.

[0052] As a further method, the method of calculating the request priority of the resource index and setting the index order based on the request priority includes:

[0053] Calculate the request priority of the resource index. The expression is:

[0054]

[0055] Among them, I is the resource index set, g is the number of features of priority in the time dimension, T cur is the current time point of the hybrid private cloud, T req (I) is the resource request time of resource index I, S t is the time scale of the t-th time dimension feature, m is the number of features in the resource data size dimension of priority, D size (I) is the size of the resource data corresponding to resource index I, S d is the data size scale of the d-th resource data size dimension feature;

[0056] Sort resource indexes in descending order of request priority. For resource indexes with the same priority, further sort them according to the time when the requests were submitted.

[0057] In the hybrid private cloud, an idle stack is set up to store the index of sorted resources. As resources are allocated and new requests arrive, the idle stack and request priority are continuously updated. If a resource is released or returned, its index is added back to the idle stack.

[0058] As a further method, the method for managing and allocating resources by designing a coordination algorithm based on the index sequence includes:

[0059] Set the objective function, the expression is:

[0060]

[0061] Among them, I is the resource index set, I i ,I j and I k are the indexes of the i-th, j-th, and k-th resources in the resource index set, respectively. i ) and P(I j ) are resource index I i and Resource Index I j The request priority, S(I i ) and S(I j ) are resource index I i and Resource Index I j The index order of |I| is the size of resource index set I, Used to measure the differences between different resource indexes;

[0062] Combined with the simulated annealing algorithm, the design coordination algorithm is constructed as follows: take the balanced allocation of resources as the current state S, and generate the neighborhood state S from the current state S according to the resource index order * , calculate the change of the objective function ΔF=F(S * )-F(S), where F(S *) is the objective function value of the neighborhood state, F(S) is the objective function value of the initial state, if ΔF<0, then the neighborhood state S is accepted * As the new current state, if ΔF>0, then The probability of accepting the neighborhood state S * , where T2 is the temperature parameter, and the calculation formula of the temperature parameter is:

[0063]

[0064] in, is the balance coefficient of the cooling strategy, λ is the cooling control coefficient, k is the current iteration number, T0 is the initial temperature of the algorithm, F c is the objective function value of the current iteration, F p is the objective function value of the previous iteration;

[0065] The coordination algorithm is iterated. When the change of the objective function in 10 consecutive iterations is less than the initial threshold, the algorithm terminates and outputs the resource coordination plan of the current state as the optimal solution.

[0066] The second aspect of the present invention provides an agency resource management and deployment system based on cloud computing, comprising:

[0067] The central cloud server module is used to configure the central cloud server based on the agency architecture and agency resource type;

[0068] A resource data aggregation module is used to perform data encoding and compression on the resource data, and then aggregate the resource data based on the data source and time series to obtain a first aggregation class set and a second aggregation class set, and store the obtained data in the central cloud server;

[0069] The edge cloud server module is used to perform association analysis based on the first aggregation class set, the second aggregation class set and the terminal users, establish a topological relationship network, and allocate edge cloud servers to the terminal users in the topological relationship network;

[0070] A user entity acquisition module, used to configure the add, delete, modify and query program to the edge cloud server to obtain the terminal user entity;

[0071] A hybrid private cloud module is used to construct a key for the terminal user entity, encrypt the terminal user's data and index using the key, and outsource it to the central cloud server to obtain a hybrid private cloud;

[0072] A request analysis module, configured to obtain a resource request, find a corresponding resource index in the hybrid private cloud, calculate a request priority of the resource index, and set an index order based on the request priority;

[0073] The resource scheduling module is used to design a coordination algorithm based on the index sequence to manage and allocate resources.

[0074] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0075] (1) The present invention can effectively reduce data storage requirements and improve the storage efficiency of cloud servers by encoding resource data and aggregating based on categories and time series.

[0076] (2) The present invention improves the privacy of resource data by increasing the strength of the key, wherein the complexity measurement comprehensively considers factors such as information entropy and uniformity of character distribution of resources, and the difficulty coefficient of successful attack comprehensively considers factors such as computing resources and time required for the attack.

[0077] (3) The present invention can effectively protect the security and privacy of agency resources by constructing keys for end-user entities and outsourcing the encrypted data and indexes to the cloud server. It also involves security mechanisms such as firewalls between edge cloud servers and central cloud servers, further enhancing the security of the system, preventing attacks from external networks, and protecting agency resources from malware and hackers.

[0078] (4) The present invention calculates resource priorities and sets the allocation index order based on the priorities, making resource allocation more scientific and reasonable, and can give priority to meeting urgent and important resource requests, thereby improving the emergency response capabilities and rule efficiency of the agency. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 The present invention is a flowchart of the steps of the cloud computing-based agency resource management and deployment method in an embodiment of the present invention. DETAILED DESCRIPTION

[0080] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0081] Reference Figure 1 As shown, the present invention provides a method for managing and allocating agency resources based on cloud computing, including:

[0082] S100 determines the type of agency resources and configures the cloud server based on the agency resource type and agency structure;

[0083] In a specific embodiment, the specific experimental environment is: server, including: processor: Intel XeonGold 6348 processor, memory: 64GB DDR4 ECC memory, storage: 4TB SSD array; network equipment, including: high-speed Ethernet switch, supporting 10GbE and above network connection, firewall, professional hardware firewall; software environment: operating system, Linux distribution, virtualization technology KVM and VMware ESXi, cloud management platform, OPENStack, database, PostgreSQL, programming language and tools, Python3.11 and Pycharm Pro; network environment: internal network, using VLAN technology, internal network, using Ethernet dedicated line to access cloud services;

[0084] In the actual evaluation, a large municipal government agency, covering multiple departments, including the Planning Bureau, Housing and Construction Bureau, Environmental Protection Bureau, etc., handles a large amount of government work and data on a daily basis. With the continuous expansion of business, the traditional resource management method can no longer meet the needs of efficient office work, so it decided to introduce a resource management and allocation method based on cloud computing;

[0085] In the actual assessment, according to the agency structure, the dimensions of the data warehouse are determined to be departmental, business type, time and region. For the departmental dimension, taking the Planning Bureau as an example, the dimension table structure includes attributes such as department number (primary key), department name, department responsibilities, personnel structure and departmental hierarchical structure. The business type dimension table includes business type number, business name and business process. The time dimension table includes timestamp (primary key), year, month, day and hour attributes. The region dimension table includes region number, region name and administrative level. Connections are established between dimension tables through foreign keys. The data warehouse is deployed on the central cloud server. For business data that requires frequent random read and write operations, such as planning files under review, block storage is used for storage. For some unstructured data such as historical documents and statistical reports, object storage is used for storage. Various resource data of the agency are stored in the data warehouse. The storage capacity of the central cloud server is 1000TB, and 300TB has been used so far.

[0086] S200 performs data encoding and compression on the resource data, and then performs aggregation based on data category and time series to obtain a first aggregation class set and a second aggregation class set, and stores them in the cloud server;

[0087] It needs to be explained that data encoding and compression is to reduce the storage space and transmission cost of resource data and improve processing and storage efficiency. Based on data category and time series aggregation, the former facilitates centralized management and analysis of specific types of data, and the latter can analyze the changing trend of data over time. The obtained first aggregation class set and second aggregation class set organize data from the data category and time dimension respectively to meet different analysis needs. Storing in a cloud server can ensure secure data storage and convenient access, and realize shared collaboration. Its high reliability and scalability can protect data, facilitate access and collaboration for different users, and ensure data security.

[0088] In the actual assessment, the sources of resource data are divided into different departments, different business systems (planning approval system, construction project management system, environmental monitoring system, etc.) and external cooperation agencies, and each source is assigned a unique identification code, where the identification code of the Planning Bureau is 001 and the planning approval system is 002. The resource data from the same data source are combined into one category to form multiple aggregation classes, and the first aggregation class set is obtained. There are 10 different data sources in total, and each source contains an average of 1,000 resource data; the creation time, update time and access time of the resource data are recorded, and the time granularity is set to hours. A day is divided into 24 time intervals to obtain the second aggregation class set. Within a month, the creation time of resource data is distributed in different time periods, and an average of 500 new resource data are created every day;

[0089] In the actual evaluation, the storage memory of the central cloud server is divided into 200 storage nodes. For one of the resource data, its unique identification code is 001, the timestamp is 202410151000 (indicating 10:00 on October 15, 2024), the resource data access frequency is 20 times a day, and the balance factor is 0.6. The storage node position can be obtained using the hash calculation formula: Hash(MD5(001||SHA256(202410151000))||0.6||20)%200;

[0090] S300 performs association analysis based on the first aggregation class set, the second aggregation class set and the terminal user, establishes a topological relationship network, and allocates an edge cloud server to the terminal user in the topological relationship network;

[0091] It should be explained that the association analysis of the first aggregation class set, the second aggregation class set and the terminal user is aimed at finding the connection and rules between them, establishing a topological relationship network, and intuitively presenting the connection structure between elements such as terminal users and data sets. Edge cloud servers are allocated to terminal users in the network, which can provide computing and storage resources close to users, improve data processing speed and response time, reduce network latency, and enhance service experience. It can also effectively manage and allocate resources and improve the overall efficiency and performance of the system.

[0092] In the actual evaluation, for the first aggregation class set, an index is established with the data source as the key, where the data index of the Planning Bureau is PJ_INDEX. For the second aggregation class set, the creation time is used as the main index key, and the update time and access time are used as the auxiliary index keys. The hierarchical index is performed according to the time granularity, and the indexes are combined to obtain a multi-dimensional index. The obtained multi-dimensional index contains 10,000 entries; the association between the data source in the first aggregation class set and the time characteristics in the second aggregation class set, as well as the association between the terminal users (staff members of various departments) and the resource data set are analyzed. The data source, time node, and terminal user are used as nodes in the network. According to the results of the association analysis, the connection relationship and weight between the nodes are determined. For example, if a staff member of the Planning Bureau accesses a planning file at 10 am, a connection is established between the Planning Bureau, the time node 202410151000, and the staff member. The connection weight is determined to be 0.8 based on factors such as access frequency and business relevance, and the topological relationship network has 500 nodes and 1,000 connections.

[0093] In the actual evaluation, a data sharing mechanism is set up between different edge cloud servers, and a distributed storage system is used to store data. The data sharing between different edge cloud servers is ensured through a data synchronization protocol. The edge cloud servers of the Planning Bureau and the Housing and Construction Bureau share some data related to urban planning and construction. User permissions for accessing the central cloud server are added to each edge cloud server. It is assumed that the central cloud server authenticates the 8 edge servers, and the authentication function is

[0094] S400 configures the edge cloud server with a program for adding, deleting, modifying and checking to obtain a terminal user entity;

[0095] In the actual evaluation, the staff of the Planning Bureau added a planning plan. The edge cloud server recorded the operation and transmitted it to the central cloud server. The central cloud server recorded the operation log in the data warehouse and updated the index. When querying resource data, the user initiated a request and entered keywords and filter conditions. The central cloud server used the inverted index to query and returned the result. The differential storage mechanism is used to manage the historical version of resource data. There are two versions of resource data, version 1 and version 2. The maximum number of record rows is 200 and the maximum number of record columns is 30. For the change of a certain attribute, the differential set is calculated. The content with position index 8 changes, and the changed content is "new data content". The position weight function is 0.7. The edit distance of this attribute between version 1 and version 2 is 4, and the Manhattan distance is 6. The standard deviation of the change range of all data points between version 1 and version 2 is 3, so the differential set is {8, "new data content", 0.7, 4, 6, 3};

[0096] S500 builds a key for the terminal user entity, uses the key to encrypt the terminal user's data and index, and outsources them to the cloud server to obtain a hybrid private cloud;

[0097] It is important to explain that the purpose of building a key for the end-user entity is to protect the data security of the end-user. Using the key to encrypt the user's data and index can prevent the data from being illegally obtained during transmission and storage. The encrypted content is then outsourced to the cloud server to form a hybrid private cloud, which not only utilizes the storage and computing capabilities of the cloud server, but also ensures the privacy and security of the end-user data through encryption.

[0098] In the actual evaluation, the initial key is constructed based on the signature algorithm of the elliptic curve, where an initial key is “hW&@qE8z%2sD(,B7|JZb}dS6iU;?g~C*yI+o9aF{@R1lTtV0mK34A$^.uQN<f H> xXc]p / LP)r_(:Yj#5kG!", use the key derivation function to calculate the complexity measure of the initial key calculated and the difficulty coefficient of the successful attack on the initial key to strengthen the initial key;

[0099] S600 obtains a resource request, sets a space stack for storing a resource index, calculates a request priority of the resource index, and sets an index order in the hybrid private cloud based on the request priority;

[0100] It needs to be explained that after obtaining the resource request, a space stack is set to store the resource index. The space stack can conveniently manage and organize the resource index, and then calculate the request priority of the resource index. This is to determine the importance and urgency of different resource requests. The index order is set in the hybrid private cloud based on the request priority, which can give priority to high-priority resource requests and improve the efficiency of resource acquisition. This can ensure that key resources can be retrieved and obtained faster to meet the urgent needs of users. At the same time, the retrieval order of resources is reasonably arranged to optimize the resource management and response capabilities of the entire system;

[0101] In the actual evaluation, there are three resource requests A, B, and C. The data size corresponding to resource A is 50MB, and the request time is 1 hour before the current time point; the data size of resource B is 30MB, and the request time is 30 minutes before the current time point; the data size of resource C is 40MB, and the request time is 45 minutes before the current time point. After calculation, the request priority of resource B is the highest, followed by A, and C is the lowest. In the hybrid private cloud, the index order is set according to this priority, and the resource B index is ranked first;

[0102] S700 designs a coordination algorithm based on the index sequence to manage and allocate resources.

[0103] Specifically, the hybrid private cloud is used in government agencies, and an effective comparison before and after the application is obtained, as shown in Table 1:

[0104] Table 1 Comparison of experimental results

[0105]

[0106] In this embodiment, the method for configuring a cloud server based on an agency architecture and an agency resource type includes:

[0107] Design the data warehouse structure. Determine the dimensions of the data warehouse according to the organizational structure. For each dimension, design the dimension table structure, including: primary key, dimension attributes, and hierarchy. If there is an association between dimension tables, use foreign keys to establish connections.

[0108] Deploy a data warehouse on a cloud server, use block storage to store resource data that requires frequent random read and write operations, and use object storage to store unstructured resource data that does not require frequent random read and write operations, and store the resource data in the data warehouse.

[0109] In this embodiment, the method of performing aggregation based on data sources and time series to obtain a first aggregation class set and a second aggregation class set, and storing them in the cloud server includes:

[0110] Divide the sources of resource data, including but not limited to different departments, different business systems and external cooperation organizations, assign a unique identification code to each source, and combine resource data from the same data source into one category to form multiple aggregation categories, thereby obtaining a first aggregation category set;

[0111] Record the creation time, update time and access time of the resource data, set the time granularity, group the time intervals according to the time granularity, obtain multiple aggregation classes based on the creation time, update time and access time, and obtain a second aggregation class set;

[0112] The storage memory of the cloud server is divided into multiple storage nodes to form a distributed storage network. The hash calculation is performed based on the unique identification code and timestamp of the resource data to determine the node location of the data storage. The expression is:

[0113] Node=(MD5(((SHA256(ID×T 2 )+R)×W)))mod N

[0114] Among them, MD5(·) represents the message digest algorithm, SHA256(·) represents the secure hash algorithm SHA-256, ID is the unique identification code of the resource data, T is the timestamp of the resource data, R is the balance factor, W is the access frequency of the resource data, mod represents the modulus operation, and N is the total number of storage nodes.

[0115] In this embodiment, the method for establishing a topological relationship network based on association analysis between the first aggregation class set, the second aggregation class set and the terminal user includes:

[0116] For the first aggregated set, an index is created with the data source as the key. For the second aggregated set, the creation time is used as the primary index key, and the update time and access time are used as the secondary index key. Hierarchical indexing is performed according to the time granularity, and the indexes are combined to obtain a multi-dimensional index.

[0117] Analyze the association between the data source in the first aggregation class set and the time feature in the second aggregation class set according to the multi-dimensional index, and analyze the association between the terminal user and the resource data set;

[0118] The data sources, time nodes, and terminal users are taken as nodes in the network. According to the results of association analysis, the connection relationship and weight between the nodes are determined to obtain a topological relationship network.

[0119] In this embodiment, the method for allocating an edge cloud server to a terminal user in the topological relationship network includes:

[0120] A data sharing mechanism is set up between different edge cloud servers, specifically: based on the connection relationship between nodes in the topological network, a distributed storage system is used to store data, and data sharing between different edge cloud servers is ensured through a data synchronization protocol;

[0121] Set up a firewall security system for the edge cloud server, specifically: when there is abnormal behavior, block the IP address of the edge cloud server and suspend the terminal user's permissions;

[0122] Add user permissions for each edge cloud server to access the central cloud server, and authenticate each edge server on the central cloud server. The expression for batch authentication of the central cloud server is:

[0123]

[0124] Among them, N * is the total number of edge servers, Auth(ESC i ,C) is the authentication function. When there is c in the authentication information set C of the central cloud server k , making the user permission set UPi With c k If the matching degree is greater than or equal to T, it returns true, indicating that the i-th edge cloud server has passed the identity authentication, otherwise it returns false, ESC i is the ith edge cloud server, C is the authentication information set of the central cloud server, c k is an authentication information in C, UP i is the user permission set of the ith edge cloud server, t is the time point of identity authentication, t0 is the time point of the last update of the edge server, and T1 is the matching standard threshold;

[0125] A buffer mechanism is set between the edge cloud server and the central cloud server. Specifically, when the communication transmission is completed, the edge cloud server and the central cloud server interrupt the transmission control protocol by closing the program.

[0126] In this embodiment, the method for configuring the add, delete, modify and query program to the edge cloud server includes:

[0127] When adding, deleting, and modifying resource data, after the terminal user submits the operation request, the edge cloud server performs the corresponding operation. After completion, the edge cloud server transmits it to the central cloud server for verification, storage, and index update, and records the operation log in the data warehouse and forwards it to the central cloud server;

[0128] When querying resource data, the user initiates a request and enters keywords and filter conditions. The central cloud server uses the inverted index to query and returns the results.

[0129] The difference storage mechanism is used to manage the historical versions of resource data and store them in the data warehouse, where the expression for difference calculation is:

[0130]

[0131] Where, ΔD v is the difference set between the vth version and the v-1th version of the resource data, pos is the position index of the change in the resource data, change is the change content at the pos position, and w p (pos) is the position weight function, It is represented as the edit distance between the lth attribute in the resource data in version v-1 and version v. It is expressed as the Manhattan distance between the lth attribute in the resource data in version v-1 and version v, thr is the standard deviation of the change of all data points between version v-1 and version v, m and n are the maximum number of record rows and the maximum number of record columns in the dimension table in the data warehouse, respectively.

[0132] In this embodiment, the method for constructing a key for the terminal user entity includes:

[0133] Construct the initial key based on the elliptic curve signature algorithm;

[0134] The complexity measure for calculating the initial key is:

[0135]

[0136] Among them, β is the information entropy weight coefficient, k is the number of different characters in the initial key, and f j is the frequency of the jth character in the initial key, γ is the weight coefficient of character distribution uniformity, is the average frequency of occurrence of different characters, α is the variance weight coefficient;

[0137] Simulate the side channel attack and calculate the difficulty coefficient of the successful attack on the initial key. The expression is:

[0138]

[0139] Among them, λ is the resource weight coefficient, and C cpu The maximum CPU load that can be used for side channel attacks and the current CPU load used for side channel attacks, respectively. and C gpu are the maximum GPU load that can be used for side channel attacks and the current GPU load used for side channel attacks, respectively. and M memory are the maximum memory size that can be used for side channel attacks and the current memory size used for side channel attacks, μ is the time weight coefficient, T attack The time required to successfully attack the initial key;

[0140] To strengthen the key strength of the initial key, the expression is:

[0141]

[0142] Among them, Q0 is the initial key strength, E(·) is the encryption algorithm of the key derivation function, and ξ represents the initial key.

[0143] In this embodiment, the method of calculating the request priority of the resource index and setting the index order based on the request priority includes:

[0144] Calculate the request priority of the resource index. The expression is:

[0145]

[0146] Among them, I is the resource index set, g is the number of features of priority in the time dimension, T cur is the current time point of the hybrid private cloud, T req (I) is the resource request time of resource index I, S t is the time scale of the t-th time dimension feature, m is the number of features in the resource data size dimension of priority, D size (I) is the size of the resource data corresponding to resource index I, S d is the data size scale of the d-th resource data size dimension feature;

[0147] Sort resource indexes in descending order of request priority. For resource indexes with the same priority, further sort them according to the time when the requests were submitted.

[0148] In the hybrid private cloud, an idle stack is set up to store the index of sorted resources. As resources are allocated and new requests arrive, the idle stack and request priority are continuously updated. If a resource is released or returned, its index is added back to the idle stack.

[0149] In this embodiment, the method for managing and allocating resources by designing a coordination algorithm based on the index sequence includes:

[0150] Set the objective function, the expression is:

[0151]

[0152] Among them, I is the resource index set, I i ,I j and I k are the indexes of the i-th, j-th, and k-th resources in the resource index set, respectively. i ) and P(I j ) are resource index I i and Resource Index I j The request priority, S(I i ) and S(I j ) are resource index I i and Resource Index I j The index order of |I| is the size of resource index set I, Used to measure the differences between different resource indexes;

[0153] Combined with the simulated annealing algorithm, the design coordination algorithm is constructed as follows: take the balanced allocation of resources as the current state S, and generate the neighborhood state S from the current state S according to the resource index order * , calculate the change of the objective function ΔF=F(S * )-F(S), where F(S *) is the objective function value of the neighborhood state, F(S) is the objective function value of the initial state, if ΔF<0, then the neighborhood state S is accepted * As the new current state, if ΔF>0, then The probability of accepting the neighborhood state S * , where T2 is the temperature parameter, and the calculation formula of the temperature parameter is:

[0154]

[0155] in, is the balance coefficient of the cooling strategy, λ is the cooling control coefficient, k is the current iteration number, T0 is the initial temperature of the algorithm, F c is the objective function value of the current iteration, F p is the objective function value of the previous iteration;

[0156] The coordination algorithm is iterated. When the change of the objective function in 10 consecutive iterations is less than the initial threshold, the algorithm terminates and outputs the resource coordination plan of the current state as the optimal solution.

[0157] The second aspect of the present invention also provides an agency resource management and deployment system based on cloud computing, including:

[0158] A cloud server configuration module is used to determine the type of agency resources and configure cloud servers based on the agency resource type and agency structure;

[0159] A resource data aggregation module, used to perform data encoding and compression on the resource data, and then aggregate based on data category and time series to obtain a first aggregation class set and a second aggregation class set, and store them in the cloud server;

[0160] The edge cloud server module is used to perform association analysis based on the first aggregation class set, the second aggregation class set and the terminal users, establish a topological relationship network, and allocate edge cloud servers to the terminal users in the topological relationship network;

[0161] A user entity acquisition module, used to configure the add, delete, modify and query program to the edge cloud server to obtain the terminal user entity;

[0162] A hybrid private cloud module is used to construct a key for the end user entity, encrypt the end user's data and index using the key, and outsource it to the cloud server to obtain a hybrid private cloud;

[0163] A request analysis module, configured to obtain resource requests, set a space stack for storing resource indexes, calculate request priorities of resource indexes, and set an index order in the hybrid private cloud based on the request priorities;

[0164] The resource scheduling module is used to design a coordination algorithm based on the index sequence to manage and allocate resources.

[0165] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. A method for managing and deploying agency resources based on cloud computing, characterized in that: The following steps are involved: Configure the central cloud server based on the agency architecture and agency resource type; After data encoding and compression are performed on the resource data, aggregation is performed based on the data source and time series to obtain a first aggregation class set and a second aggregation class set, which are stored in the central cloud server; Based on the association analysis between the first aggregation class set, the second aggregation class set and the terminal user, a topological relationship network is established, and an edge cloud server is allocated to the terminal user in the topological relationship network; Configure the edge cloud server with a program for adding, deleting, modifying and checking to obtain a terminal user entity; Constructing a key for the end user entity, using the key to encrypt the end user's data and index, and outsourcing it to the central cloud server to obtain a hybrid private cloud; Obtain a resource request, find a corresponding resource index in the hybrid private cloud, calculate a request priority of the resource index, and set an index order based on the request priority; Design a coordination algorithm based on the index sequence to manage and allocate resources.

2. The cloud computing-based agency resource management and deployment method according to claim 1, characterized in that: The method for configuring a central cloud server based on an agency architecture and an agency resource type includes: Design the data warehouse structure, determine the dimensions of the data warehouse according to the organizational structure, and for each dimension, design the dimension table structure, including: primary key, dimension attributes, and hierarchy. If there is an association between dimension tables, use foreign keys to establish connections; A data warehouse is deployed on the central cloud server. Block storage is used to store resource data that requires frequent random read and write operations, and object storage is used to store unstructured resource data that does not require frequent random read and write operations. The resource data is stored in the data warehouse.

3. The cloud computing-based agency resource management and deployment method according to claim 1, characterized in that: The method of performing aggregation based on data sources and time series to obtain a first aggregation class set and a second aggregation class set, and storing them in the central cloud server includes: Divide the sources of resource data, including but not limited to different departments, different business systems and external cooperation organizations, assign a unique identification code to each source, combine resource data from the same data source into one category, form multiple aggregation categories, and obtain a first aggregation category set; Record the creation time, update time and access time of the resource data, set the time granularity, group the time intervals according to the time granularity, obtain multiple aggregation classes based on the creation time, update time and access time, and obtain a second aggregation class set; The storage memory of the central cloud server is divided into multiple storage nodes to form a distributed storage network. The hash calculation is performed based on the unique identification code and timestamp of the resource data to determine the node location of the data storage. The expression is: Node=(MD5(((SHA256(ID×T 2 )+R)×W)))mod N Among them, MD5(·) represents the message digest algorithm, SHA256(·) represents the secure hash algorithm SHA-256, ID is the unique identification code of the resource data, T is the timestamp of the resource data, R is the balance factor, W is the access frequency of the resource data, mod represents the modulus operation, and N is the total number of storage nodes.

4. The cloud computing-based agency resource management and deployment method according to claim 1, characterized in that: The method for establishing a topological relationship network based on association analysis between the first aggregation class set, the second aggregation class set and the terminal user includes: For the first aggregated set, an index is created with the data source as the key. For the second aggregated set, the creation time is used as the primary index key, and the update time and access time are used as the secondary index key. Hierarchical indexing is performed according to the time granularity, and the indexes are combined to obtain a multi-dimensional index. Analyze the association between the data source in the first aggregation class set and the time feature in the second aggregation class set according to the multi-dimensional index, and analyze the association between the terminal user and the resource data set; The data sources, time nodes, and terminal users are taken as nodes in the network. According to the results of association analysis, the connection relationship and weight between the nodes are determined to obtain a topological relationship network.

5. The cloud computing-based agency resource management and deployment method according to claim 1, characterized in that: The method for allocating an edge cloud server to a terminal user in the topological relationship network includes: A data sharing mechanism is set up between different edge cloud servers, specifically: based on the connection relationship between nodes in the topological network, a distributed storage system is used to store data, and data sharing between different edge cloud servers is ensured through a data synchronization protocol; Set up a firewall security system for the edge cloud server, specifically: when there is abnormal behavior, block the IP address of the edge cloud server and suspend the terminal user's permissions; Add user permissions for each edge cloud server to access the central cloud server, and authenticate each edge server on the central cloud server. The expression for batch authentication of the central cloud server is: Among them, N * is the total number of edge servers, Auth(ESC i ,C) is the authentication function. When there is c in the authentication information set C of the central cloud server k , making the user permission set UP i With c k If the matching degree is greater than or equal to T, it returns true, indicating that the i-th edge cloud server has passed the identity authentication, otherwise it returns false, ESC i is the ith edge cloud server, C is the authentication information set of the central cloud server, c k is an authentication information in C, UP i is the user permission set of the ith edge cloud server, t is the time point of identity authentication, t0 is the time point of the last update of the edge server, and T1 is the matching standard threshold; A buffer mechanism is set between the edge cloud server and the central cloud server. Specifically, when the communication transmission is completed, the edge cloud server and the central cloud server interrupt the transmission control protocol by closing the program.

6. The cloud computing-based agency resource management and deployment method according to claim 1, characterized in that: The method for configuring a program for adding, deleting, modifying and checking the edge cloud server includes: When adding, deleting, and modifying resource data, after the terminal user submits the operation request, the edge cloud server performs the corresponding operation. After completion, the edge cloud server transmits it to the central cloud server for verification, storage, and index update, and records the operation log in the data warehouse and forwards it to the central cloud server; When querying resource data, the user initiates a request and enters keywords and filter conditions. The central cloud server uses the inverted index to query and returns the results. The difference storage mechanism is used to manage the historical versions of resource data and store them in the data warehouse, where the expression for difference calculation is: Where, ΔD v is the difference set between the vth version and the v-1th version of the resource data, pos is the position index of the change in the resource data, change is the change content at the pos position, and w p (pos) is the position weight function, It is represented as the edit distance between the lth attribute in the resource data in version v-1 and version v. It is expressed as the Manhattan distance between the lth attribute in the resource data in version v-1 and version v, thr is the standard deviation of the change of all data points between version v-1 and version v, m and n are the maximum number of record rows and the maximum number of record columns in the dimension table in the data warehouse, respectively.

7. The cloud computing-based agency resource management and deployment method according to claim 1, characterized in that: The method of constructing a key for the end-user entity comprises: Construct the initial key based on the elliptic curve signature algorithm; The complexity measure for calculating the initial key is: Among them, β is the information entropy weight coefficient, k is the number of different characters in the initial key, and f j is the frequency of the jth character in the initial key, γ is the weight coefficient of character distribution uniformity, is the average frequency of occurrence of different characters, α is the variance weight coefficient; Simulate the side channel attack and calculate the difficulty coefficient of the successful attack on the initial key. The expression is: Among them, λ is the resource weight coefficient, and C cpu The maximum CPU load that can be used for side channel attacks and the current CPU load used for side channel attacks, respectively. and C gpu are the maximum GPU load that can be used for side channel attacks and the current GPU load used for side channel attacks, respectively. and M memory are the maximum memory size that can be used for side channel attacks and the current memory size used for side channel attacks, μ is the time weight coefficient, T attack The time required to successfully attack the initial key; To strengthen the key strength of the initial key, the expression is: Among them, Q0 is the initial key strength, E(·) is the encryption algorithm of the key derivation function, and ξ represents the initial key.

8. The cloud computing-based agency resource management and deployment method according to claim 1, characterized in that: The method of calculating the request priority of the resource index and setting the index order based on the request priority includes: Calculate the request priority of the resource index. The expression is: Among them, I is the resource index set, g is the number of features of priority in the time dimension, T cur is the current time point of the hybrid private cloud, T req (I) is the resource request time of resource index I, S t is the time scale of the t-th time dimension feature, m is the number of features in the resource data size dimension of priority, D size (I) is the size of the resource data corresponding to resource index I, S d is the data size scale of the d-th resource data size dimension feature; Sort resource indexes in descending order of request priority. For resource indexes with the same priority, further sort them according to the time when the requests were submitted. In the hybrid private cloud, an idle stack is set up to store the index of sorted resources. As resources are allocated and new requests arrive, the idle stack and request priority are continuously updated. If a resource is released or returned, its index is added back to the idle stack.

9. The cloud computing-based agency resource management and deployment method according to claim 1, characterized in that: The method for managing and allocating resources by designing a coordination algorithm based on the index sequence includes: Set the objective function, the expression is: Among them, I is the resource index set, I i ,I j and I k are the indexes of the i-th, j-th, and k-th resources in the resource index set, respectively. i ) and P(I j ) are resource index I i and Resource Index I j The request priority, S(I i ) and S(I j ) are resource index I i and Resource Index I j The index order of |I| is the size of resource index set I, Used to measure the differences between different resource indexes; Combined with the simulated annealing algorithm, the design coordination algorithm is constructed as follows: take the balanced allocation of resources as the current state S, and generate the neighborhood state S from the current state S according to the resource index order * , calculate the change of the objective function ΔF=F(S * )-F(S), where F(S * ) is the objective function value of the neighborhood state, F(S) is the objective function value of the initial state, if ΔF<0, then the neighborhood state S is accepted * As the new current state, if ΔF>0, then The probability of accepting the neighborhood state S * , where T2 is the temperature parameter, and the calculation formula of the temperature parameter is: in, is the balance coefficient of the cooling strategy, λ is the cooling control coefficient, k is the current iteration number, T0 is the initial temperature of the algorithm, F c is the objective function value of the current iteration, F p is the objective function value of the previous iteration; The coordination algorithm is iterated. When the change of the objective function in 10 consecutive iterations is less than the initial threshold, the algorithm terminates and outputs the resource coordination plan of the current state as the optimal solution.

10. A cloud computing-based agency resource management and deployment system, used to execute the cloud computing-based agency resource management and deployment method according to any one of claims 1 to 9, characterized in that: The system comprises: The central cloud server module is used to configure the central cloud server based on the agency architecture and agency resource type; A resource data aggregation module is used to perform data encoding and compression on the resource data, and then aggregate the resource data based on the data source and time series to obtain a first aggregation class set and a second aggregation class set, and store the obtained data in the central cloud server; The edge cloud server module is used to perform association analysis based on the first aggregation class set, the second aggregation class set and the terminal users, establish a topological relationship network, and allocate edge cloud servers to the terminal users in the topological relationship network; A user entity acquisition module, used to configure the add, delete, modify and query program to the edge cloud server to obtain the terminal user entity; A hybrid private cloud module is used to construct a key for the terminal user entity, encrypt the terminal user's data and index using the key, and outsource it to the central cloud server to obtain a hybrid private cloud; A request analysis module, configured to obtain a resource request, find a corresponding resource index in the hybrid private cloud, calculate a request priority of the resource index, and set an index order based on the request priority; The resource scheduling module is used to design a coordination algorithm based on the index sequence to manage and allocate resources.

Images