Computing resource allocation method and apparatus, and electronic device

Abstract

The present disclosure provides a computing resource allocation method and apparatus, and an electronic device. A privacy leakage risk is taken into account, task allocation between local computing of terminal devices and edge servers can be dynamically adjusted, an edge server selection policy is refined to each task, and the requirements of the terminal devices for edge server resources is automatically adjusted, thereby reducing the overall energy consumption of a health monitoring system, improving energy efficiency and the response speed, enhancing privacy protection, and improving user experience.

Description

Computing resource allocation methods and devices, electronic equipment

[0001] This disclosure claims priority to Chinese Patent Application No. 202411189532.4, filed with the China National Intellectual Property Administration on August 28, 2024, entitled "Method and Apparatus for Allocating Computing Resources, Electronic Device, Storage Medium", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This disclosure relates to the field of computing resource allocation technology, and in particular to a computing resource allocation method, apparatus, and electronic device. Background Technology

[0003] The Internet of Things (IoT) in healthcare has become a transformative force, combining medical devices and applications to create a highly personalized and efficient healthcare ecosystem. Health monitoring systems are a key application of IoT in healthcare. In these systems, wearable devices continuously track a patient's biomedical data and transmit it to the patient's terminal device. This terminal device acts as a gateway, forwarding the data to healthcare providers, enabling them to continuously monitor the patient's health at minimal cost. Furthermore, health monitoring systems can expand a patient's range of activities, significantly improving their quality of life.

[0004] Existing health monitoring systems face significant challenges due to limitations in communication resources and the insufficient computing power of patients' own devices, resulting in constraints on system efficiency and responsiveness. Mobile edge computing (MEC) and sixth-generation wireless networks offer solutions to overcome these obstacles. The integration of MEC and sixth-generation wireless networks in the medical IoT ecosystem may become the cornerstone of next-generation medical technologies.

[0005] In health monitoring systems based on the Internet of Things (IoT) for healthcare, the allocation of computing resources is a critical issue that significantly impacts the overall system performance. Existing technologies propose a module placement method using a classification regression tree algorithm, which can effectively allocate computing tasks while optimizing authentication, confidentiality, integrity, availability, capacity, speed, and cost; however, they neglect the standby time of users' mobile devices and device energy consumption. Existing technologies also utilize Stackelberg game theory as an incentive mechanism to influence users' offloading decisions, aiming to balance minimizing the energy consumption of mobile edge computing servers with ensuring user experience quality; however, they ignore the needs of the healthcare monitoring system. Furthermore, existing technologies propose a cooperative bargaining game to optimize the allocation of wireless channel resources, using a decentralized, non-cooperative game for computational offloading decisions to minimize the overall system energy consumption. While considering medical urgency, information timeliness, and the overall system energy consumption including servers during resource allocation, they neglect the risk of user privacy breaches and the standby time of mobile devices. Therefore, how to balance optimizing user device energy efficiency with strict privacy protection during computing resource allocation has become an urgent problem to be solved in this field. Summary of the Invention

[0006] This disclosure aims to address at least one of the problems existing in the prior art by providing a method, apparatus, and electronic device for allocating computing resources.

[0007] One aspect of this disclosure provides a computing resource allocation method applied to a health monitoring system, the health monitoring system including multiple edge servers, multiple terminal devices respectively communicatively connected to each of the edge servers, and a number of wearable devices respectively communicatively connected to each of the terminal devices;

[0008] The computing resource allocation method includes:

[0009] Based on the communication energy consumption between the terminal device and each of the edge servers, as well as the computing energy consumption, privacy leakage risk assessment indicators and their coefficients corresponding to the terminal device, the utility function and its constraints corresponding to the terminal device are determined based on the edge server selection strategy.

[0010] Based on the utility function, the edge server selection strategy corresponding to the maximum utility of each terminal device is determined;

[0011] Determine the utility increment of each terminal device under the edge server selection strategy when its utility is maximized, relative to its utility under the current edge server selection strategy.

[0012] The current edge server selection strategy of the terminal device with the largest utility increment is updated to the edge server selection strategy corresponding to the one with the largest utility, so as to allocate the computing resources of the edge server to the terminal device.

[0013] Optionally, the step of determining the utility function and its constraints corresponding to the terminal device based on the edge server selection strategy, according to the communication energy consumption between the terminal device and each of the edge servers, and the computing energy consumption, privacy leakage risk assessment indicators and their coefficients corresponding to the terminal device, includes:

[0014] The utility function is constructed according to the following formula:

[0015] Where, ρ i This represents the edge server selection strategy corresponding to the i-th terminal device; ρ -i This represents the edge server selection strategy for terminal devices other than the i-th terminal device; Represents the utility function corresponding to the i-th terminal device; Let the cost function corresponding to the i-th terminal device be expressed as:

[0016] Where ρ represents the set of edge server selection strategies corresponding to all terminal devices; S i This represents the privacy leakage risk assessment index corresponding to the i-th terminal device; S represents i The coefficient of I(·); I(·) represents the indicator function, the function value of I(·) is 1 when the expression in parentheses is true, and the function value of I(·) is 0 when the expression in parentheses is false; This represents the energy consumption for communication between the i-th terminal device and the edge server; This represents the total energy consumed by the i-th terminal device in processing the computing task;

[0017] The constraints of the utility function are determined according to the following formula:

[0018] ρ i ∈{0}∪K for i=1, 2,..,N;

[0019] Where N represents the number of terminal devices; K represents the set of all edge servers.

[0020] Optionally, the utility increment of the i-th terminal device under the edge server selection strategy when its utility is maximized, relative to its utility under the current edge server selection strategy, is expressed as:

[0021] Where, Δi (t) represents the utility increment corresponding to the i-th terminal device at time t; Let ρ′ represent the edge server selection strategy for the i-th terminal device when its utility is maximized. i Meanwhile, the utility function of other terminal devices while keeping their edge server selection strategy unchanged; This indicates that the selection strategy for the i-th terminal device on its corresponding edge server is ρ. i Meanwhile, the utility function of other terminal devices while keeping their edge server selection strategy unchanged.

[0022] Optionally, the privacy leakage risk assessment index S corresponding to the i-th terminal device i Represented as:

[0023] Where j represents the number of the wearable device communicating with the i-th terminal device, j = 1, 2, ..., M, and M represents the total number of wearable devices communicating with the i-th terminal device; g i，j This represents the calculated location corresponding to the j-th wearable device that is communicatively connected to the i-th terminal device. A value of 0 indicates that the calculated location is local to the terminal device, and a value of 1 indicates that the calculated location follows the selection of the terminal device; S i，j This represents the computing task t corresponding to the j-th wearable device that is communicatively connected to the i-th terminal device. i,j Privacy leakage risk assessment index.

[0024] Optionally, the communication energy consumption between the i-th terminal device and the edge server Represented as:

[0025] Where, p i This represents the transmission power of the i-th terminal device; The transmission delay corresponding to the i-th terminal device is represented as:

[0026] in, Let represent the total task size uploaded by the i-th terminal device to the edge server, and express it as:

[0027] Where, d i，j Represents the computation task t i,j Data size; Represents the computation task t i,j The size of the calculated data;

[0028] R i Let represent the transmission rate between the i-th terminal device and its selected edge server, and express it as:

[0029] Where k represents the edge server number and k = 1, 2, ..., K, and K represents the total number of edge servers; h i，k ρ represents the channel gain between the i-th terminal device and the k-th edge server; a represents the terminal device number, and the value of a is a value from 1 to N that is different from i; a This represents the edge server selection strategy corresponding to the a-th terminal device; p a h represents the transmission power of the a-th terminal device; a，k σ represents the channel gain between the i-th terminal device and the k-th edge server; B represents the channel bandwidth; σ 2 This indicates the background noise power.

[0030] Optionally, the total energy consumed by the i-th terminal device in processing the computing task. Represented as:

[0031] in, This represents the computing power of the i-th terminal device; This represents the computing power of the i-th terminal device; Let represent the number of CPU cycles required for the i-th terminal device to execute the computation task locally on its terminal device, and express it as:

[0032] Among them, f i,j Indicates the processing of computational task t i，j The number of CPU cycles required.

[0033] Optionally, the computing resource allocation method further includes:

[0034] For each terminal device, it is determined whether the number of times it selects the terminal device to execute computing tasks locally exceeds a preset threshold. If so, the computing location corresponding to the wearable device with the highest security level in its communication connection is set to the terminal device local.

[0035] Another aspect of this disclosure provides a computing resource allocation device for use in a health monitoring system, the health monitoring system including multiple edge servers, multiple terminal devices respectively communicatively connected to each of the edge servers, and a number of wearable devices respectively communicatively connected to each of the terminal devices;

[0036] The computing resource allocation device includes:

[0037] The construction module is used to determine the utility function and its constraints corresponding to the terminal device based on the communication energy consumption between the terminal device and each of the edge servers, as well as the computing energy consumption, privacy leakage risk assessment indicators and their coefficients corresponding to the terminal device, and the edge server selection strategy.

[0038] The first determining module is used to determine the edge server selection strategy corresponding to each terminal device when its utility is maximized, based on the utility function.

[0039] The second determining module is used to determine the utility increment of each terminal device under the edge server selection strategy when its utility is maximized, relative to its utility under the current edge server selection strategy.

[0040] The update module is used to update the current edge server selection policy of the terminal device with the largest utility increment to the edge server selection policy corresponding to the maximum utility, so as to allocate the computing resources of the edge server to the terminal device.

[0041] Another aspect of this disclosure provides an electronic device comprising:

[0042] At least one processor; and,

[0043] A memory that is communicatively connected to at least one processor; wherein,

[0044] The memory stores instructions that can be executed by at least one processor, which enables the at least one processor to perform the computing resource allocation method described above.

[0045] Another aspect of this disclosure provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the computing resource allocation method described above.

[0046] Compared to existing technologies, this disclosure can automatically adjust the terminal device's demand for edge server resources, reducing the overall energy consumption of the health monitoring system. It prioritizes reducing terminal device energy consumption as a primary objective, effectively extending the terminal device's standby time. For sensitive user health and medical information, it takes into account the risk of privacy leaks, effectively reducing the risk of user privacy breaches and ensuring the security of user information. Furthermore, it can dynamically adjust the task allocation between the terminal device's local computing and the edge server, refining the edge server selection strategy to each task to achieve an optimal balance between reducing energy consumption and protecting privacy. This improves the overall system performance and user experience, enhances the response speed of medical services, and increases user trust and satisfaction with the health monitoring system. Attached Figure Description

[0047] Figure 1 is a schematic diagram of the communication connection relationship of a health monitoring system provided in one embodiment of this disclosure;

[0048] Figure 2 is a flowchart of a computing resource allocation method provided in another embodiment of this disclosure;

[0049] Figure 3 is a schematic diagram of the strategy selection result in Example 1 provided by another embodiment of this disclosure;

[0050] Figure 4 is a schematic diagram of the dynamic changes in the terminal device utility in Example 2 provided by another embodiment of this disclosure;

[0051] Figure 5 is a schematic diagram of a computing resource allocation device provided in another embodiment of this disclosure;

[0052] Figure 6 is a schematic diagram of the structure of an electronic device provided in another embodiment of this disclosure. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the various embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been provided in the various embodiments of this disclosure to facilitate a better understanding of the disclosure. However, the technical solutions claimed in this disclosure can be implemented even without these technical details and with various variations and modifications based on the following embodiments. The division of the various embodiments below is for ease of description and should not constitute any limitation on the specific implementation of this disclosure. The various embodiments can be combined with and referenced by each other without contradiction.

[0054] One embodiment of this disclosure relates to a computing resource allocation method applied to a health monitoring system. As shown in Figure 1, the health monitoring system includes multiple edge servers, multiple terminal devices communicatively connected to each edge server, and several wearable devices communicatively connected to each terminal device. Each wearable device can communicate with the terminal devices via a wireless link, thereby forming an intra-WBAN network. The multiple terminal devices can also form an extra-WBAN network with their corresponding edge servers, enabling the edge servers to send the wearable device monitoring data collected by the terminal devices to a healthcare provider via a cloud server. The multiple terminal devices can communicate with their corresponding edge servers via wireless links, each edge server can communicate with the cloud server via a wired link, and the cloud server can communicate with the healthcare provider via a wired link.

[0055] Wearable devices are used to continuously monitor key health indicators such as heart rate, blood pressure, blood sugar, and other biomarkers. Terminal devices store information, process data, and act as gateways. They can decide whether to perform computational tasks locally or offload them to edge servers. In other words, terminal devices can aggregate and process monitoring data collected by the user's wearable devices and send relevant medical reports or data required by the edge servers. The edge servers then forward these reports or monitoring data to cloud servers and healthcare providers, enabling healthcare providers to continuously monitor the user's biomedical data at minimal cost.

[0056] Leveraging the high speed and low latency of 6th generation mobile networks, terminal devices can efficiently transmit collected user health data to nearby edge servers. These edge servers are strategically positioned to perform real-time data processing, significantly reducing latency compared to traditional cloud-based health monitoring systems. The edge servers can also immediately use the processed data to gain health insights or send it to healthcare providers for further analysis and action. The health monitoring system provided in this embodiment not only responds quickly to users' healthcare needs but also improves the overall responsiveness and accuracy of the health monitoring system, enhancing the responsiveness and accuracy of medical interventions.

[0057] Resource allocation for edge servers presents significant challenges. Therefore, the computing resource allocation method provided in this embodiment focuses on minimizing energy consumption at the terminal device level, recognizing the critical energy sensitivity of terminal devices. Furthermore, biomedical data collected by wearable devices raises significant privacy concerns, necessitating a strategy that cleverly integrates privacy risk management. Therefore, when allocating resources for edge servers, a delicate balance must be struck between optimizing terminal device energy efficiency and rigorous privacy protection.

[0058] As shown in Figure 2, the resource allocation calculation method includes steps S210 to S240.

[0059] Step S210: Based on the communication energy consumption between the terminal device and each edge server, as well as the computing energy consumption, privacy leakage risk assessment indicators and their coefficients corresponding to the terminal device, determine the utility function and its constraints corresponding to the terminal device based on the edge server selection strategy.

[0060] Specifically, the information processed and transmitted by the health monitoring system is essentially medical information, raising significant concerns about user privacy. Data collected by wearable devices provides in-depth insights into an individual's health status. From a privacy and security perspective alone, users prefer to process tasks locally on the terminal device rather than uploading the collected data to an edge server. Local computing on the terminal device allows users to reduce the risk of data leakage collected by wearable devices while directly transmitting necessary medical reports to healthcare providers. Therefore, this implementation establishes a privacy leakage risk assessment index to systematically evaluate the possibility of privacy leakage during task transmission.

[0061] According to IEEE standard 802.15.6, wearable devices will select a security level when transmitting messages. There are three security levels: Level 0 (unencrypted), Level 1 (authenticated but not encrypted), and Level 2 (authenticated and encrypted). All tasks of wearable devices can be classified into the above three security levels.

[0062] Define an indicator function I(·), where the function value of I(·) is 1 when the expression within its parentheses is true, and 0 when the expression within its parentheses is false. Let t be the computational task corresponding to the j-th wearable device that is communicatively connected to the i-th terminal device. i,j The privacy leakage risk assessment index is S. i，j Then S i，j It can be calculated using the following formula: Where s = 0, 1, 2 represents the security level, α i，j，s ∈[0,∞) represents the coefficient of security level s, v i，j Represents the computation task t i，j The amount of information. α i，j，s The value of increases as the value of s increases. j represents the number of the wearable device that communicates with the i-th terminal device, j = 1, 2, ..., M, where M represents the total number of wearable devices that communicate with the i-th terminal device.

[0063] In an external wireless body domain network, each terminal device can only communicate with one edge server at a time. Let ρ denote the edge server selection strategy corresponding to the i-th terminal device. i When the number of terminal devices is N (N is a positive integer), the set ρ consisting of the edge server selection strategies corresponding to all terminal devices can be represented as ρ = {ρ1, ρ2, ..., ρ...} N The set ρ contains the strategic choices made by each terminal device when selecting an edge server for communication purposes. i=k indicates that the i-th terminal device selected the k-th edge server for task offloading. Here, k represents the edge server number (k = 1, 2, ..., K), and K represents the total number of edge servers (K is a positive integer). ρ i =0 indicates that the i-th terminal device prefers to process tasks locally. The transmission rate R between the i-th terminal device and its selected edge server is... i It can be calculated in the following ways: Where, p i h represents the transmission power of the i-th terminal device. i，k Let represent the channel gain between the i-th terminal device and the k-th edge server. 'a' represents the terminal device number, and its value is from 1 to N, different from i. a This represents the edge server selection strategy corresponding to the a-th terminal device. a h represents the transmission power of the a-th terminal device. a,k σ represents the channel gain between the i-th terminal device and the k-th edge server. B represents the channel bandwidth. 2 This indicates the background noise power.

[0064] In practical applications, low-power operation of wearable devices is crucial, requiring designs that can accommodate both minimal battery size and infrequent charging cycles. Low energy consumption is essential for ensuring the long-term operation and effectiveness of wearable devices deployed in various medical monitoring scenarios. In the health monitoring system provided in this embodiment, although the energy consumption of the wearable device for information collection and task generation remains constant, the transmission energy consumption generated when it sends data to the terminal device makes the overall energy consumption of the wearable device variable. Energy consumption within the external wireless body domain network is primarily characterized by two key elements: communication energy consumption between the terminal device and the edge server, and computing energy consumption.

[0065] The total task size uploaded by the i-th terminal device to the edge server It can be calculated using the following formula: Where, d i，j Represents the computation task t i，j The data size. Represents the computation task t i,j The size of the calculated data. g i，j It is a binary variable representing the calculated location corresponding to the j-th wearable device that is connected to the i-th terminal device. Its value of 0 indicates that the calculated location is local to the terminal device, and its value of 1 indicates that the calculated location follows the selection of the terminal device.

[0066] Transmission delay corresponding to the i-th terminal device It can be calculated using the following formula: The energy consumption of communication between the i-th terminal device and the edge server It can be calculated using the following formula:

[0067] In the health monitoring system provided in this embodiment, computational tasks can be processed by terminal devices or edge servers. After the computational tasks are processed, the resulting data can be transmitted to a cloud server and forwarded to the healthcare provider. This workflow ensures that computational tasks crucial for health monitoring and analysis can be effectively executed, making full use of the distributed computing resources available within the network.

[0068] The computation task t will be processed i,j The required number of CPU cycles is denoted as f. i，j Then, the number of CPU cycles required for the i-th terminal device to execute the computing task locally on its terminal device is... It can be calculated using the following formula: Among them, f i，j Represents the computation task t i,j The number of CPU cycles required. The number of CPU cycles required for the i-th terminal device to offload computing tasks to the edge server for processing. It can be calculated using the following formula:

[0069] The computing power of the i-th terminal device and the computing power of the k-th edge server can be respectively derived from... and This is represented as follows: Assuming the edge server allocates an equal amount of computing resources to each terminal device, the available computing power that the i-th terminal device obtains from the edge server is... It can be calculated using the following formula: Where a takes the values ​​1, 2, ..., N.

[0070] The total energy consumed by the i-th terminal device in processing the computing task It can be calculated using the following formula: in, This represents the computational power of the i-th terminal device.

[0071] The total energy consumed by the edge server to process computing tasks from the i-th terminal device. It can be calculated using the following formula: in, This indicates the computing power of the edge server.

[0072] After collecting all relevant data, the terminal device faces a critical decision: whether to offload the task to an edge server or process it locally on the terminal device. Most existing technologies focus on reducing the overall energy consumption of the system, but the computing resource allocation method provided in this embodiment shifts towards reducing the operating expenses of the terminal device and incorporates considerations of privacy leakage risks. This embodiment considers two factors: First, given the widespread use of user-carried terminal devices, prioritizing the energy consumption of the terminal device itself outweighs the focus on the energy requirements of the edge server; second, the inherent confidentiality of user health data, and the processing of a large amount of user personal information for each task, makes processing tasks directly on the edge server a less desirable option. In contrast, processing tasks locally on the terminal device and then transmitting the relevant reports to healthcare providers can effectively enhance user privacy protection. Therefore, this embodiment must prioritize user-related privacy impacts in the decision-making process regarding task processing.

[0073] Since one terminal device corresponds to one user, the cost function corresponding to the i-th terminal device is... It can be represented as Among them, S i The privacy leakage risk assessment index corresponding to the i-th terminal device can be represented as: S represents i The coefficient.

[0074] Based on the above, step S210, for example, includes: constructing a utility function according to the following formula:

[0075] in, Let ρ represent the utility function corresponding to the i-th terminal device. -i This represents the edge server selection strategy for terminal devices other than the i-th terminal device.

[0076] For example, step S210 further includes: determining the constraints of the utility function according to the following formula:

[0077] ρ i ∈{0}∪K for i=1.2,...,N.

[0078] Where K represents the set of all edge servers.

[0079] Step S220: Based on the utility function, determine the edge server selection strategy corresponding to each terminal device when its utility is maximized.

[0080] Specifically, after obtaining the utility function corresponding to the i-th terminal device... Afterwards, according to Determine the edge server selection strategy corresponding to the i-th terminal device when its utility is maximized, and then obtain the edge server selection strategy corresponding to each terminal device when its utility is maximized.

[0081] Step S230: Determine the utility increment of each terminal device under the edge server selection policy when its utility is maximized, relative to its current edge server selection policy.

[0082] Specifically, to address the optimization challenges related to edge server resource allocation, the computational resource allocation method provided in this embodiment is actually a dynamic edge server resource allocation algorithm. The algorithm first sets a utility function for each terminal device, assuming that all terminal devices initially execute all tasks locally, thus establishing a baseline for system operation. Then, the algorithm begins an iterative process until the system reaches a Nash equilibrium. During the algorithm's iteration, each iteration includes a policy evaluation, where the incremental utility gain from potential edge server resource reallocation is compared with the existing edge server selection policy. In this iteration, the terminal device that obtains the maximum utility gain updates its edge server selection policy accordingly. Once the incremental utility gain of a terminal device stabilizes at zero, i.e., a Nash equilibrium is reached, the algorithm converges. Therefore, step S230 requires determining the utility increment of each terminal device under its edge server selection policy at its maximum utility, relative to its current edge server selection policy, to identify the terminal device with the largest utility increment.

[0083] For example, the utility increment of the i-th terminal device under the edge server selection policy when its utility is maximized, relative to its utility under the current edge server selection policy, is expressed as:

[0084] Where, Δ i (t) represents the utility increment corresponding to the i-th terminal device at time t. ρ′ i This represents the edge server selection strategy for the i-th terminal device when its utility is maximized. Let ρ′ represent the edge server selection strategy for the i-th terminal device when its utility is maximized. i Meanwhile, the utility function of other terminal devices while keeping their edge server selection strategy unchanged. This indicates that the selection strategy for the i-th terminal device on its corresponding edge server is ρ. i Meanwhile, the utility function of other terminal devices while keeping their edge server selection strategy unchanged.

[0085] Step S240: Update the current edge server selection policy of the terminal device with the largest utility increment to the edge server selection policy corresponding to the one with the largest utility, so as to allocate the computing resources of the edge server to the terminal device.

[0086] By updating the current edge server selection strategy for the terminal device with the largest utility increment to the edge server selection strategy corresponding to the one with the largest utility, the resource utilization of the edge server can be effectively improved, thereby effectively improving the computing efficiency.

[0087] For example, the computing resource allocation method further includes: for each terminal device, determining whether the number of times it selects the terminal device to execute computing tasks locally exceeds a preset threshold; if so, setting the computing location corresponding to the wearable device with the highest security level of its communication connection to the terminal device local.

[0088] Specifically, to optimize local computing costs for terminal devices while protecting user privacy, the computing resource allocation method provided in this embodiment also includes a dynamic adjustment mechanism. When the number of times the terminal device performs computing tasks locally exceeds a preset threshold, this dynamic adjustment mechanism automatically modifies the terminal device's requests for edge server computing, isolating tasks with a higher risk of privacy leakage for local processing on the terminal device. This is manifested in setting the computing location corresponding to the wearable device with the highest security level communicating with the terminal device as local to the terminal device and excluding it from the decision-making process regarding whether edge computing is needed. Conversely, when the terminal device continuously obtains computing resources from the edge server within the preset threshold, this dynamic adjustment mechanism encourages more tasks to be included in the scope of potential outsourcing to the edge server. That is, this strategic algorithm balances privacy protection and computing efficiency, guiding the system to achieve optimal operational equilibrium.

[0089] The computing resource allocation method provided in this disclosure, compared to existing technologies, can automatically adjust the terminal device's demand for edge server resources, reducing the overall energy consumption of the health monitoring system. It prioritizes reducing terminal device energy consumption as a primary objective, effectively extending the terminal device's standby time. Regarding sensitive user health and medical information, the computing resource allocation method provided in this disclosure takes privacy leakage risks into account, effectively reducing the risk of user privacy leaks and ensuring the security of user information. Furthermore, the computing resource allocation method provided in this disclosure can dynamically adjust the task allocation between the terminal device's local computing and the edge server, refining the edge server selection strategy to each task to achieve an optimal balance between reducing energy consumption and protecting privacy. This improves the overall system performance and user experience, enhances the response speed of medical services, and increases user trust and satisfaction with the health monitoring system.

[0090] To enable those skilled in the art to better understand the above implementation methods, a specific example is provided below to illustrate the dynamic allocation algorithm for edge server resources involved in the computing resource allocation method.

[0091] The algorithm for dynamic resource allocation on edge servers is as follows:

[0092] Input: A set of N terminal devices, a set of wearable devices corresponding to each terminal device, and a set of edge servers.

[0093] Output: The set of edge server selection strategies for each terminal device.

[0094] initialization:

[0095] 1. Set the initial time slot t = 0.

[0096] 2. For each terminal device i:

[0097] -Initialize the edge server selection strategy ρ i It is 0.

[0098] - Initialize counter σ i It is 0.

[0099] - For the j-th wearable device of each terminal device i, the location g will be calculated. i，j Initialize to 1.

[0100] Strategy Update:

[0101] 3. For each time slot t:

[0102] - Initialize the edge server selection strategy candidate set Ω to be empty.

[0103] -For each terminal device:

[0104] - Find the utility function that maximizes terminal device i. Edge server selection strategy ρ′ i ,

[0105] -Calculate the utility increment Δ of the strategy change i (t),

[0106] -will(Δ i (t), ρ′ i Add to the edge server selection strategy candidate set Ω, Ω←Ω∪{(Δ i (t), ρ′ i )}.

[0107] -When Ω is not empty:

[0108] - Find the utility increment Δ i The terminal device n of (t),

[0109] n←argmax n∈N Δ i (t).

[0110] - Update the edge server selection policy for terminal device n, ρ n (t+1)←ρ′ n .

[0111] - Remove (Δ) from the edge server selection strategy candidate set Ω i (t), ρ′ i Remove(Δ) n (t), ρ′ n )from Ω.

[0112] Mission strategy update:

[0113] 4. For each time slot t:

[0114] -For each terminal device i:

[0115] -If the edge server of terminal device i selects strategy ρ i =0:

[0116] -Counter δ i Increase by 1.

[0117] -If the counter δ i Exceeding the preset threshold δ threshold ,but:

[0118] - Find the index S with the highest privacy breach risk assessment index i，j Wearable devices n.

[0119] -g i,n Set to 0.

[0120] Otherwise, the counter δ i Reset to 0.

[0121] 5. Increase time slot t by 1.

[0122] The beneficial effects of the computing resource allocation method provided in this disclosure will be illustrated below through examples.

[0123] Taking 30 terminal devices, each equipped with 1 to 10 wearable devices through a random selection process, as an example. The task data size generated by the wearable devices ranges from 100KB to 300KB, with the specific data size determined randomly. The required CPU cycles for processing the tasks are between 10 Mbps and 100 Mbps. The health monitoring system operates on a 5MHz bandwidth, with a standard data transmission power of 0.1W. Edge servers are randomly distributed, and the radio coverage range is 50 meters. The channel gain is set to... Among them, 1 i，j Let represent the distance between the i-th terminal device and the j-th wearable device it communicates with, and 3 represent the path loss factor. The computing power of the edge server and the terminal device is configured to 30,000MHz and 2,000MHz, respectively. Both the terminal device and the edge server are allocated a uniform transmission and computing power of 0.1W. Medical criticality and privacy leakage risk assessment indicators are both set in the range of 0 to 3. The privacy leakage risk assessment indicator S corresponding to the i-th terminal device is... i coefficient The value is 0.012.

[0124] Example 1: A simulation experiment involving 30 terminal devices and 5 edge servers was conducted to observe the dynamic changes in the edge server selection strategy of the terminal devices over time. Initially, all terminal devices selected local computing. Figure 3 shows the strategy selection results of this experiment, which demonstrates a clear trend towards equilibrium over time.

[0125] Example 2: A simulation experiment involving 10 terminal devices and 5 edge servers was conducted to observe the dynamic changes in the utility of the terminal devices over time. Figure 4 shows the dynamic changes in the utility of the terminal devices in this experiment. The results show that the utility of these 10 terminal devices converges and stops changing within a finite number of iterations, and the utility of all terminal devices increases over time.

[0126] Another embodiment of this disclosure relates to a computing resource allocation device applied to a health monitoring system. The health monitoring system includes multiple edge servers, multiple terminal devices communicatively connected to each edge server, and several wearable devices communicatively connected to each terminal device. As shown in FIG5, the computing resource allocation device includes a construction module 510, a first determination module 520, a second determination module 530, and an update module 540.

[0127] The construction module 510 is used to determine the utility function and its constraints for the terminal device based on the edge server selection strategy, according to the communication energy consumption between the terminal device and each edge server, as well as the computing energy consumption, privacy leakage risk assessment indicators and their coefficients of the terminal device.

[0128] The first determining module 520 is used to determine the edge server selection strategy corresponding to each terminal device when its utility is maximized, based on the utility function.

[0129] The second determining module 530 is used to determine the utility increment of each terminal device under the edge server selection strategy when its utility is maximized, relative to its current edge server selection strategy.

[0130] The update module 540 is used to update the current edge server selection policy of the terminal device with the largest utility increment to the edge server selection policy corresponding to the one with the largest utility, so as to allocate the computing resources of the edge server to the terminal device.

[0131] For example, the update module 540 is further configured to: for each terminal device, determine whether the number of times it selects the terminal device to perform computing tasks locally exceeds a preset threshold; if so, set the computing location corresponding to the wearable device with the highest security level of its communication connection to the terminal device local.

[0132] For a detailed implementation of the computing resource allocation device provided in this disclosure, please refer to the computing resource allocation method provided in this disclosure, which will not be repeated here.

[0133] The computing resource allocation device provided in this disclosure, compared with the prior art, can automatically adjust the terminal device's demand for edge server resources, reduce the overall energy consumption of the health monitoring system, and take reducing the terminal device's energy consumption as one of its main objectives, effectively extending the terminal device's standby time; for sensitive user health and medical information, it takes into account the risk of privacy leakage, effectively reducing the risk of user privacy leakage and ensuring the security of user information; it can dynamically adjust the task allocation between the terminal device's local computing and the edge server, and refine the edge server selection strategy to each task to achieve the optimal balance between reducing energy consumption and protecting privacy, improving the overall system performance and user experience, increasing the response speed of medical services, and increasing users' trust and satisfaction with the health monitoring system.

[0134] Another embodiment of this disclosure relates to an electronic device, as shown in FIG6, comprising:

[0135] At least one processor 601; and,

[0136] Memory 602 is communicatively connected to at least one processor 601; wherein,

[0137] The memory 602 stores instructions that can be executed by at least one processor 601, which enables the at least one processor 601 to perform the computing resource allocation method described in the above embodiments.

[0138] The memory and processor are connected via a bus, which can include any number of interconnecting buses and bridges, connecting various circuits of one or more processors and memories. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0139] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0140] Another embodiment of this disclosure relates to a computer-readable storage medium storing a computer program that, when executed by a processor, implements the computing resource allocation method described in the above embodiments.

[0141] That is, those skilled in the art will understand that all or part of the steps in the methods described in the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0142] Those skilled in the art will understand that the above embodiments are specific implementations of this disclosure, and in practical applications, various changes can be made to them in form and detail without departing from the spirit and scope of this disclosure.

Claims

1. A method for allocating computing resources, characterized in that, The system is applied to a health monitoring system, which includes multiple edge servers, multiple terminal devices that are communicatively connected to each of the edge servers, and a number of wearable devices that are communicatively connected to each of the terminal devices. The computing resource allocation method includes: Based on the communication energy consumption between the terminal device and each of the edge servers, as well as the computing energy consumption, privacy leakage risk assessment indicators and their coefficients corresponding to the terminal device, the utility function and its constraints corresponding to the terminal device are determined based on the edge server selection strategy. Based on the utility function, the edge server selection strategy corresponding to the maximum utility of each terminal device is determined; Determine the utility increment of each terminal device under the edge server selection strategy when its utility is maximized, relative to its current edge server selection strategy. The current edge server selection strategy of the terminal device with the largest utility increment is updated to the edge server selection strategy corresponding to the one with the largest utility, so as to allocate the computing resources of the edge server to the terminal device.

2. The computing resource allocation method according to claim 1, characterized in that, The step of determining the utility function and its constraints corresponding to the terminal device based on the communication energy consumption between the terminal device and each of the edge servers, as well as the computing energy consumption, privacy leakage risk assessment indicators and their coefficients corresponding to the terminal device, and the edge server selection strategy, includes: The utility function is constructed according to the following formula: Where, ρ i This represents the edge server selection strategy corresponding to the i-th terminal device; ρ -i This represents the edge server selection strategy for terminal devices other than the i-th terminal device; Represents the utility function corresponding to the i-th terminal device; Let the cost function corresponding to the i-th terminal device be expressed as: Where ρ represents the set of edge server selection strategies corresponding to all terminal devices; S i This represents the privacy leakage risk assessment index corresponding to the i-th terminal device; S represents i The coefficient of I(·); I(·) represents the indicator function, the function value of I(·) is 1 when the expression in parentheses is true, and the function value of I(·) is 0 when the expression in parentheses is false; This represents the energy consumption for communication between the i-th terminal device and the edge server; This represents the total energy consumed by the i-th terminal device in processing the computing task; The constraints of the utility function are determined according to the following formula: ρ i ∈{0}∪K for i=1,2,...,N; Where N represents the number of terminal devices; K represents the set of all edge servers.

3. The computing resource allocation method according to claim 2, characterized in that, The utility increment of the i-th terminal device under the edge server selection strategy when its utility is maximized, relative to its current edge server selection strategy, is expressed as: Where, Δ i (t) represents the utility increment corresponding to the i-th terminal device at time t; Let ρ′ represent the edge server selection strategy for the i-th terminal device when its utility is maximized. i Meanwhile, the utility function of other terminal devices while keeping their edge server selection strategy unchanged; This indicates that the selection strategy for the i-th terminal device on its corresponding edge server is ρ. i Meanwhile, the utility function of other terminal devices while keeping their edge server selection strategy unchanged.

4. The computing resource allocation method according to claim 2, characterized in that, The privacy leakage risk assessment index S corresponding to the i-th terminal device i Represented as: Where j represents the number of the wearable device communicating with the i-th terminal device, j = 1, 2, ..., M, and M represents the total number of wearable devices communicating with the i-th terminal device; g i，j This represents the calculated location corresponding to the j-th wearable device that is communicatively connected to the i-th terminal device. A value of 0 indicates that the calculated location is local to the terminal device, and a value of 1 indicates that the calculated location follows the selection of the terminal device; S i，j This represents the computing task t corresponding to the j-th wearable device that is communicatively connected to the i-th terminal device. i,j Privacy leakage risk assessment index.

5. The computing resource allocation method according to claim 4, characterized in that, The energy consumption of communication between the i-th terminal device and the edge server Represented as: Where, p i This represents the transmission power of the i-th terminal device; The transmission delay corresponding to the i-th terminal device is represented as: in, Let represent the total task size uploaded by the i-th terminal device to the edge server, and express it as: Where, d i，j Represents the computation task t i，j Data size; Represents the computation task t i，j The size of the calculated data; R i Let represent the transmission rate between the i-th terminal device and its selected edge server, and express it as: Where k represents the edge server number and k = 1, 2, ..., K, and K represents the total number of edge servers; h i，k ρ represents the channel gain between the i-th terminal device and the k-th edge server; a represents the terminal device number, and the value of a is a value from 1 to N that is different from i; a This represents the edge server selection strategy corresponding to the a-th terminal device; p a h represents the transmission power of the a-th terminal device; a，k σ represents the channel gain between the i-th terminal device and the k-th edge server; B represents the channel bandwidth; σ 2 This indicates the background noise power.

6. The computing resource allocation method according to claim 4, characterized in that, The total energy consumed by the i-th terminal device in processing the computing task Represented as: in, This represents the computing power of the i-th terminal device; This represents the computing power of the i-th terminal device; Let represent the number of CPU cycles required for the i-th terminal device to execute the computation task locally on its terminal device, and express it as: Among them, f i,j Indicates the processing of computational task t i,j The number of CPU cycles required.

7. The computing resource allocation method according to claim 1, characterized in that, The computing resource allocation method further includes: For each terminal device, it is determined whether the number of times it selects the terminal device to execute computing tasks locally exceeds a preset threshold. If so, the computing location corresponding to the wearable device with the highest security level in its communication connection is set to the terminal device local.

8. A computing resource allocation device, characterized in that, The system is applied to a health monitoring system, which includes multiple edge servers, multiple terminal devices that are communicatively connected to each of the edge servers, and a number of wearable devices that are communicatively connected to each of the terminal devices. The computing resource allocation device includes: The construction module is used to determine the utility function and its constraints corresponding to the terminal device based on the communication energy consumption between the terminal device and each of the edge servers, as well as the computing energy consumption, privacy leakage risk assessment indicators and their coefficients corresponding to the terminal device, and the edge server selection strategy. The first determining module is used to determine the edge server selection strategy corresponding to each terminal device when its utility is maximized, based on the utility function. The second determining module is used to determine the utility increment of each terminal device under the edge server selection strategy when its utility is maximized, relative to its utility under the current edge server selection strategy. The update module is used to update the current edge server selection policy of the terminal device with the largest utility increment to the edge server selection policy corresponding to the maximum utility, so as to allocate the computing resources of the edge server to the terminal device.

9. An electronic device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the computing resource allocation method of claim 1.

Images