A method and apparatus for single-user task offloading based on downlink NOMA

Abstract

This invention discloses a single-user task offloading method and apparatus based on downlink NOMA. Communication is established between a wireless device with computing tasks and N base stations equipped with edge servers using downlink NOMA. First, the operating range of the wireless device's transmission duration is initialized. A first transmission duration and a second transmission duration are selected from the operating range according to preset rules. Then, for the first and second transmission durations, under the condition of satisfying the maximum transmission power of the wireless device, the corresponding optimal offloading task allocation scheme is searched. The operating range of the wireless device's transmission duration is then updated, and the search continues until the lower limit of the final operating range and its corresponding offloading task allocation scheme are output as the final offloading task allocation scheme. This invention optimizes the communication duration and offloading task allocation, thereby reducing task completion latency.

Description

Technical Field

[0001] This application belongs to the field of task offloading technology, and particularly relates to a single-user task offloading method and apparatus based on downlink NOMA, which is applicable to edge computing networks based on downlink NOMA communication. Background Technology

[0002] With the rapid development of the Internet of Things (IoT), numerous computationally intensive and latency-sensitive computing tasks have emerged. However, many wireless devices in the IoT have limited computing resources and power, resulting in unacceptable latency when processing such tasks. Mobile edge computing (MEC) technology allows local users to offload some or all of their computing load to edge servers, thereby significantly reducing local user latency.

[0003] Most previous research on edge computing networks has employed orthogonal multiple access (OMA) technologies, such as frequency division multiplexing (FDM) and time division multiplexing (TDM). However, due to their orthogonality, these technologies typically have low spectrum utilization efficiency.

[0004] The task completion latency for local users can be divided into three parts: offloading and transmission time, edge server computation time, and local user download time of computation results. Reducing the task completion latency for local users in edge computing networks is one of the important research directions for those skilled in the art. Summary of the Invention

[0005] The purpose of this application is to provide a single-user task offloading method and apparatus based on downlink NOMA, which applies NOMA technology to mobile edge computing networks and optimizes communication duration and offloading task allocation, thereby reducing task completion latency.

[0006] To achieve the above objectives, the technical solution of this application is as follows:

[0007] A single-user task offloading method based on downlink NOMA includes:

[0008] Step 1: A wireless device with computing tasks communicates with N base stations equipped with edge servers using downlink NOMA communication.

[0009] Step 2: Initialize the operating range of wireless device transmission duration, and take the first transmission duration and the second transmission duration from the operating range according to the preset rules;

[0010] Step 3: For the first transmission duration and the second transmission duration, under the condition that the maximum power that the wireless device can transmit is satisfied, search for the optimal offload task allocation scheme corresponding to the first transmission duration and the second transmission duration respectively.

[0011] Step 4: Compare the optimal unloading task allocation schemes corresponding to the first transmission duration and the second transmission duration, update the operating range of the wireless device transmission duration. If the operating range of the wireless device transmission duration after the update is greater than the preset threshold, then take the first transmission duration and the second transmission duration from the operating range according to the preset rules and return to Step 3. Otherwise, output the unloading task allocation scheme corresponding to the lower limit of the current operating range as the final unloading task allocation scheme.

[0012] Furthermore, the initialization of the wireless device transmission duration operating range, which involves taking a first transmission duration and a second transmission duration from the operating range according to a preset rule, includes:

[0013] First, initialize the operating range of the wireless device's transmission duration, using t. min This represents the lower limit of the transmission duration operating range, and is initialized to t. min =0; use t max This represents the upper limit of the transmission duration interval, initialized with t. max =S / μ L S and μ L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively.

[0014] Then, according to preset rules, the first transmission duration t1 and the second transmission duration t2 are taken from the operating interval. In a preferred embodiment, the golden ratio method is used to determine the values, that is:

[0015] t1=t min +0.382×(t max -t min )

[0016] t2=t min +0.618×(t max -t min ).

[0017] Furthermore, for the first transmission duration and the second transmission duration, under the condition of satisfying the maximum power that the wireless device can transmit, the optimal offloading task allocation scheme corresponding to the first transmission duration and the second transmission duration are searched respectively, including:

[0018] Step S3.1: Initialize k = 0. Under a given transmission duration t1, calculate the offloading task allocation S1,…,S for the wireless device according to the following formula. N and the transmit power required by the wireless device under the offloading task allocation

[0019]

[0020] Where i equals 1, let 1 / g0 be 0, W is the transmission bandwidth, n0 is the Gaussian white noise power spectral density, and γ i Let S be the computing speed of the i-th edge server, k represent the number of edge servers to which the wireless device offloads tasks, and S and μ represent the computing speed of the i-th edge server. L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively, g i This represents the channel gain between the wireless device and the edge server i. S represents the transmit power required by the wireless device. i The amount of task offloaded by the wireless device to the i-th edge server;

[0021] Step S3.2: Increment the value of k by 1. If P max If the maximum power that the wireless device can transmit is given, then return to step S3.1 to continue calculating the offloading task allocation S1,…,S for the wireless device. N and the transmit power required by the wireless device under the offloading task allocation Output the wireless device offloading task allocation scheme when k=N, until k equals N; or until... Output the offloading task allocation scheme for the wireless device corresponding to k-1.

[0022] Furthermore, for the first transmission duration and the second transmission duration, under the condition of satisfying the maximum power that the wireless device can transmit, the optimal offloading task allocation scheme corresponding to the first transmission duration and the second transmission duration are searched respectively, including:

[0023] Step S4.1: Initialize k = 0. Under a given transmission duration t1, calculate the offloading task allocation S1,…,S for the wireless device according to the following formula. N and the transmit power required by the wireless device under the offloading task allocation

[0024]

[0025] Where i equals 1, let 1 / g0 be 0, W is the transmission bandwidth, n0 is the Gaussian white noise power spectral density, and γ i Let S be the computing speed of the i-th edge server, k represent the number of edge servers to which the wireless device offloads tasks, and S and μ represent the computing speed of the i-th edge server. L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively, g i This represents the channel gain between the wireless device and the edge server i. S represents the transmit power required by the wireless device. i The amount of task offloaded by the wireless device to the i-th edge server;

[0026] Step S4.2: Increment the value of k by 1. If P max If the maximum power that the wireless device can transmit is given, then return to step S4.1 to continue calculating the offloading task allocation S1,…,S for the wireless device. N and the transmit power required by the wireless device under the offloading task allocation Until the value of k equals N, output the wireless device offloading task allocation scheme corresponding to k=N;

[0027] or until at this time and Reduce the offloading workload of the edge server with the worst channel gain by ΔS k Initialize ΔS k The upper and lower limits are taken as ΔS. k The average of the upper and lower limits The offloading task allocation S1,…,S for wireless devices is calculated using the following formula. N And the required transmit power P of the wireless device under the offloading task allocation N ' OMA :

[0028]

[0029] If P N ' OMA >P max Then let: otherwise

[0030] Continue to judge If the value is greater than a preset threshold, then continue taking ΔS. k The average of the upper and lower limits Calculate the offloading task allocation S1,…,S for wireless devices. N And the required transmit power P of the wireless device under the offloading task allocation N ' OMA ,until If the value is not greater than a preset threshold, output the wireless device unloading task allocation scheme.

[0031] Furthermore, the step of comparing the optimal offload task allocation scheme corresponding to the first transmission duration and the second transmission duration, and updating the operating range of the wireless device's transmission duration, includes:

[0032] The total latency D(t,L) for wireless device task completion is calculated using the following formula:

[0033]

[0034] Where, γ i Let S and μ be the computing speed of the i-th edge server. L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively, S i The amount of task offloaded by the wireless device to the i-th edge server, where t represents the transmission duration;

[0035] Compare the total latency of the wireless device task completion corresponding to the first transmission duration and the second transmission duration. If the total latency of the wireless device task completion corresponding to the first transmission duration is greater than the total latency of the wireless device task completion corresponding to the second transmission duration, then set the lower limit of the operating range of the wireless device transmission duration to the first transmission duration; otherwise, set the upper limit of the operating range of the wireless device transmission duration to the second transmission duration.

[0036] This application also proposes a single-user task offloading device based on downlink NOMA, including a processor and a memory storing a number of computer instructions, which, when executed by the processor, implement the steps of the single-user task offloading method based on downlink NOMA.

[0037] This application proposes a single-user task offloading method and apparatus based on downlink NOMA, employing an efficient two-layer algorithm to determine the transmission duration and offloading task allocation scheme. The inner layer designs a heuristic algorithm based on properties and combines it with binary search to obtain the edge server offloading task allocation that minimizes the total task completion latency for a given transmission duration. The outer layer applies the golden section search method to obtain the transmission duration that minimizes the total task completion latency. This technical solution optimizes the communication duration and offloading task allocation, thereby reducing task completion latency. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the edge computing network structure of this application;

[0039] Figure 2 This is a flowchart of the single-user task offloading method based on downlink NOMA in this application. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0041] Most edge computing networks currently employ orthogonal multiple access (OMA) technologies, such as frequency division multiplexing (FDM) and time division multiplexing (TDM). However, their spectrum utilization efficiency is generally lower than that of non-orthogonal multiple access (NOMA) technologies. This application introduces power domain NOMA technology, which is widely used in the industry and can provide communication services to multiple users simultaneously on the same frequency band. NOMA technology further reduces latency during data transmission.

[0042] This application refers to the NOMA-based edge computing network as a NOMA-MEC network. In the NOMA-MEC network, the task completion latency for local users can be divided into three parts: offloading transmission time, edge server computation time, and local user download time of computation results. This application omits the latency of each edge server transmitting computation results back, because the size of the computation results is usually much smaller than the computation task, and the transmission power of the edge server is much higher than that of the local node. However, in practical applications, local users are often limited by computing resources and operating power, and the computation rate of the edge server also has an upper limit. Therefore, this application considers that the offloading transmission time for local users and the computation time for edge servers cannot be ignored.

[0043] This application provides a single-user task offloading method based on downlink NOMA, which can be applied to, for example... Figure 1 The edge computing network application environment shown includes a wireless device WD with computing tasks and N base stations BS equipped with edge servers. Each wireless device WD and each base station BS is equipped with a single antenna for communication. The total computing workload required by the user on the wireless device WD is denoted as S. When the user offloads some computing tasks to the edge servers, downlink NOMA communication is used. The edge servers use ECS (Elastic Compute Service). i Let i = 1, 2, ..., N, and represent the edge servers (ECS). i Its computing power is γ i When a wireless device (user) has a total task S that needs to be calculated, the total task can be divided into N+1 parts, which are calculated locally by the wireless device and by N edge servers, with the task allocated to the ECS (Elastic Compute Service) being... i The computational complexity is expressed in S i Therefore, latency can also be divided into N+1 types: local computation latency and the latency of each edge server. The latency of each server includes the same NOMA communication latency and its own computation latency. The maximum value among these N+1 latency values ​​is the total latency of the task.

[0044] In one embodiment, such as Figure 2 As shown, a single-user task offloading method based on downlink NOMA is provided, including:

[0045] Step S1: A wireless device with computing tasks communicates with N base stations equipped with edge servers using downlink NOMA communication.

[0046] For example Figure 1 In the edge computing network shown, the total amount of computing tasks required on the wireless device WD is denoted as S. When the user offloads some computing tasks to the edge server, downlink NOMA communication is used.

[0047] Both the wireless device WD and each base station BS are equipped with a single antenna for communication. The channel gain g between the wireless device WD and the edge server i is... i Calculate according to the following formula:

[0048]

[0049] Among them, A d The product of the antenna gains of both sides, where π is pi, and f is the product of the antenna gains of both sides. c For the carrier frequency, d i Let d be the distance between the wireless device WD and the i-th edge server. e α is the path loss exponent. i The Rayleigh fading factor is independent for each channel, with a mean of 1.

[0050] In this embodiment, it is assumed that the channel gain between wireless device WD and edge server 1 to edge server N is decreasing, i.e., satisfying g1≥g2≥…≥g N .

[0051] Step S2: Initialize the operating range of wireless device transmission duration, and take the first transmission duration and the second transmission duration from the operating range according to the preset rules.

[0052] In this embodiment, N channels have gains g = [g1, g2, ..., g...]. N The edge server calculates the corresponding wireless device transmission duration t and wireless device task offloading allocation L, where L = {S1, S2, ..., S...} N} represents the set of tasks offloaded by wireless devices to N edge servers. S i The amount of task offloaded by the wireless device to the i-th edge server.

[0053] First, initialize the operating range of the wireless device's transmission duration, using t. min This represents the lower limit of the transmission duration operating range, and is initialized to t. min =0; use t max This represents the upper limit of the transmission duration interval, initialized with t. max =S / μ L S and μ LThese represent the total workload that the wireless device needs to process and the local computing speed, respectively.

[0054] Then, according to preset rules, the first transmission duration t1 and the second transmission duration t2 are taken from the operating interval. In a preferred embodiment, the golden ratio method is used to determine the values, that is:

[0055] t1=t min +0.382×(t max -t min )

[0056] t2=t min +0.618×(t max -t min )

[0057] It should be noted that using the golden ratio to determine the value is only one specific implementation method. Those skilled in the art can use other methods to determine the value, such as arbitrarily choosing two values ​​that differ by a certain duration, which will not be elaborated here. This embodiment uses the example of the second duration being greater than the first duration for illustration. It is easy to understand that, conversely, the subsequent update running interval is adjusted accordingly, which does not affect the execution of the technical solution of this application.

[0058] Step S3: For the first transmission duration and the second transmission duration, under the condition that the maximum power that the wireless device can transmit is satisfied, search for the optimal offload task allocation scheme corresponding to the first transmission duration and the second transmission duration respectively.

[0059] This step aims to find the offload task allocation scheme that is closest to the maximum power that the wireless device can transmit, while satisfying the maximum power that the wireless device can transmit, as the optimal offload task allocation scheme for the first transmission duration and the second transmission duration respectively.

[0060] In a preferred embodiment, the optimal offloading task allocation scheme corresponding to the first transmission duration and the second transmission duration is searched, including:

[0061] Step S3.1: Initialize k = 0. Under a given transmission duration t1, calculate the offloading task allocation S1,…,S for the wireless device according to the following formula. N and the transmit power required by the wireless device under the offloading task allocation

[0062]

[0063] Where i equals 1, let 1 / g0 be 0, W is the transmission bandwidth, n0 is the Gaussian white noise power spectral density, and γ iLet S be the computation rate of the i-th edge server, and k represent the number of edge servers to which the wireless device offloads tasks. S and μ L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively, g i This represents the channel gain between the wireless device and edge server i. Non-orthogonal multiple access (NOMA) technology allows the wireless device to share bandwidth W with all edge servers. This indicates the transmit power required by the wireless device. i The amount of task offloaded by the wireless device to the i-th edge server.

[0064] Step S3.2: Increment the value of k by 1. If P max If the maximum power that the wireless device can transmit is given, then return to step S3.1 to continue calculating the offloading task allocation S1,…,S for the wireless device. N and the transmit power required by the wireless device under the offloading task allocation Output the wireless device offloading task allocation scheme when k=N, until k equals N; or until... Output the offloading task allocation scheme for the wireless device corresponding to k-1.

[0065] It is easy to understand that if Then continuously calculate the offloading task allocation S1,…,S for the wireless devices. N and the transmit power required by the wireless device under the offloading task allocation If k equals N, or... The output will be:

[0066]

[0067] In this process, once If the value of k is less than N, it means that selecting k edge servers has exceeded the maximum power that the wireless device can transmit. In this case, the offloading task allocation scheme of the wireless device corresponding to k-1 is output.

[0068] Considering that in the previous embodiment, in and Even with these conditions, there is still unused power, so further optimization is possible.

[0069] In another preferred embodiment, the optimal offloading task allocation scheme corresponding to the first transmission duration and the second transmission duration is searched, including:

[0070] Step S4.1: Initialize k = 0. Under a given transmission duration t1, calculate the offloading task allocation S1,…,S for the wireless device according to the following formula. Nand the transmit power required by the wireless device under the offloading task allocation

[0071]

[0072] Where i equals 1, let 1 / g0 be 0, W is the transmission bandwidth, n0 is the Gaussian white noise power spectral density, and γ i Let S be the computing speed of the i-th edge server, k represent the number of edge servers to which the wireless device offloads tasks, and S and μ represent the computing speed of the i-th edge server. L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively, g i This represents the channel gain between the wireless device and the edge server i. S represents the transmit power required by the wireless device. i The amount of task offloaded by the wireless device to the i-th edge server.

[0073] Step S4.2: Increment the value of k by 1. If P max If the maximum power that the wireless device can transmit is given, then return to step S4.1 to continue calculating the offloading task allocation S1,…,S for the wireless device. N and the transmit power required by the wireless device under the offloading task allocation Until the value of k equals N, output the wireless device offloading task allocation scheme corresponding to k=N;

[0074] or until at this time and Reduce the offloading workload of the edge server with the worst channel gain by ΔS k Initialize ΔS k The upper and lower limits are taken as ΔS. k The average of the upper and lower limits The offloading task allocation S1,…,S for wireless devices is calculated using the following formula. N And the required transmit power P of the wireless device under the offloading task allocation N ' OMA :

[0075]

[0076] If P N ' OMA >P max Then let: otherwise

[0077] Continue to judge If the value is greater than a preset threshold, then continue taking ΔS. kThe average of the upper and lower limits Calculate the offloading task allocation S1,…,S for wireless devices. N And the required transmit power P of the wireless device under the offloading task allocation N ' OMA ,until If the value is not greater than a preset threshold, output the wireless device unloading task allocation scheme.

[0078] That is, in and At that time, the offloading workload of the edge server with the worst channel gain is reduced by ΔS. k ,use Indicates ΔS k The lower bound value during binary search is initialized. use Indicates ΔS k The upper bound value during binary search is initialized.

[0079] In ΔS k Within the search range, take the average of the upper and lower limits. Its value is shown below:

[0080]

[0081] If that For the value P, we have N ' OMA >P max ,but otherwise Where P N ' OMA This indicates the transmit power required by the wireless device corresponding to the offloading task allocation. If Then take the average of the upper and lower limits again. Calculate the offloading task allocation S1,…,S for wireless devices. N And the required transmit power P of the wireless device under the offloading task allocation N ' OMA ,until If the error is not greater than a preset threshold, output the wireless device offloading task allocation scheme. δ is the preset threshold, which is the error tolerance.

[0082] Using the method described in this step, we can obtain the corresponding offloading task allocation scheme for transmission durations t1 and t2. and

[0083] Step S4: Compare the optimal offload task allocation schemes corresponding to the first transmission duration and the second transmission duration, update the operating range of the wireless device transmission duration. If the operating range of the wireless device transmission duration after the update is greater than the preset threshold, then take the first transmission duration and the second transmission duration from the operating range according to the preset rules and return to step S3. Otherwise, output the lower limit of the current operating range and its corresponding offload task allocation scheme as the final offload task allocation scheme.

[0084] This step compares the optimal offload task allocation schemes corresponding to the first and second transmission durations. The comparison can be made based on the transmit power required by each allocation scheme, or by calculating the total latency for the wireless device to complete the task.

[0085] In a preferred embodiment, comparing the optimal offload task allocation schemes corresponding to the first transmission duration and the second transmission duration, and updating the operating range of the wireless device's transmission duration includes:

[0086] The total latency D(t,L) for wireless device task completion is calculated using the following formula:

[0087]

[0088] Where, γ i Let S and μ be the computing speed of the i-th edge server. L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively, S i The amount of task offloaded by the wireless device to the i-th edge server, where t represents the transmission duration;

[0089] Compare the total latency of the wireless device task completion corresponding to the first transmission duration and the second transmission duration. If the total latency of the wireless device task completion corresponding to the first transmission duration is greater than the total latency of the wireless device task completion corresponding to the second transmission duration, then set the lower limit of the operating range of the wireless device transmission duration to the first transmission duration; otherwise, set the upper limit of the operating range of the wireless device transmission duration to the second transmission duration.

[0090] Specifically, with This represents the total latency for the wireless device to complete the task corresponding to the first transmission duration. This indicates the total latency for the wireless device to complete the task corresponding to the second transmission duration. Then t min =t1, otherwise t max =t2.

[0091] Finally, calculate the operating interval t of the updated wireless device transmission duration. max -t min If t max -t minIf the value exceeds the preset threshold, the first transmission duration and the second transmission duration are taken from the running interval according to the preset rules, and then the process returns to step S3. Otherwise, the lower limit of the current running interval and its corresponding unloading task allocation scheme are output as the final unloading task allocation scheme.

[0092] This application employs an efficient two-layer algorithm to determine the transmission duration and offloading task allocation scheme. The inner layer designs a heuristic algorithm based on properties and combines it with binary search to obtain the edge server offloading task allocation that minimizes the total task completion latency for a given transmission duration value. The outer layer applies the golden section search method to obtain the transmission duration that minimizes the total task completion latency.

[0093] In one embodiment, this application also provides a single-user task offloading apparatus based on downlink NOMA, including a processor and a memory storing a plurality of computer instructions, wherein the computer instructions, when executed by the processor, implement the steps of the single-user task offloading method based on downlink NOMA.

[0094] Specific limitations regarding the downlink NOMA-based single-user task offloading device can be found in the limitations of the downlink NOMA-based single-user task offloading method described above, and will not be repeated here. The aforementioned downlink NOMA-based single-user task offloading device can be implemented wholly or partially through software, hardware, or a combination thereof. It can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations described above.

[0095] The memory and processor are electrically connected directly or indirectly to enable data transmission or interaction. For example, these components can be electrically connected to each other via one or more communication buses or signal lines. The memory stores a computer program that can run on the processor, which implements the network topology layout method in this embodiment of the invention by running the computer program stored in the memory.

[0096] The memory may be, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), etc. The memory stores the program, and the processor executes the program upon receiving an execution instruction.

[0097] The processor may be an integrated circuit chip with data processing capabilities. The aforementioned processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor.

[0098] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A single-user task offloading method based on downlink NOMA, characterized in that, The single-user task offloading method based on downlink NOMA includes: Step 1: A wireless device with computing tasks communicates with N base stations equipped with edge servers using downlink NOMA communication. Step 2: Initialize the operating range of wireless device transmission duration, and take the first transmission duration and the second transmission duration from the operating range according to the preset rules; Step 3: For the first transmission duration and the second transmission duration, under the condition that the maximum power that the wireless device can transmit is satisfied, search for the optimal offload task allocation scheme corresponding to the first transmission duration and the second transmission duration respectively. Step 4: Compare the optimal unloading task allocation schemes corresponding to the first transmission duration and the second transmission duration, update the operating range of the wireless device transmission duration. If the operating range of the wireless device transmission duration after the update is greater than the preset threshold, then take the first transmission duration and the second transmission duration from the operating range according to the preset rules and return to Step 3. Otherwise, output the unloading task allocation scheme corresponding to the lower limit of the current operating range as the final unloading task allocation scheme. The initialization of the wireless device transmission duration within an operating range, wherein a first transmission duration and a second transmission duration are selected from the operating range according to a preset rule, includes: First, initialize the operating range of the wireless device's transmission duration, using t. min This represents the lower limit of the transmission duration operating range, and is initialized to t. min =0; use t max This represents the upper limit of the transmission duration interval, initialized with t. max =S / μ L S and μ L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively. Then, according to preset rules, the first transmission duration t1 and the second transmission duration t2 are taken from the operating interval, and the golden ratio method is used to determine the values, that is: t1=t min +0.382×(t max -t min ); t2=t min +0.618×(t max -t min ); For the first transmission duration and the second transmission duration, under the condition of satisfying the maximum power that the wireless device can transmit, the optimal offloading task allocation scheme corresponding to the first transmission duration and the second transmission duration are searched respectively, including: Step S3.1: Initialize k = 0. Under a given transmission duration t1, calculate the offloading task allocation S1,…,S for the wireless device according to the following formula. N and the transmit power required by the wireless device under the offloading task allocation Where i equals 1, let 1 / g0 be 0, W is the transmission bandwidth, n0 is the Gaussian white noise power spectral density, and γ i Let S be the computing speed of the i-th edge server, k represent the number of edge servers to which the wireless device offloads tasks, and S and μ represent the computing speed of the i-th edge server. L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively, g i This represents the channel gain between the wireless device and the edge server i. S represents the transmit power required by the wireless device. i The amount of task offloaded by the wireless device to the i-th edge server; Step S3.2: Increment the value of k by 1. If P max If the maximum power that the wireless device can transmit is given, then return to step S3.1 to continue calculating the offloading task allocation S1,…,S for the wireless device. N and the transmit power required by the wireless device under the offloading task allocation Output the wireless device offloading task allocation scheme when k=N, until k equals N; or until... Output the offload task allocation scheme for the wireless device corresponding to k-1; The step of comparing the optimal offload task allocation schemes corresponding to the first transmission duration and the second transmission duration, and updating the operating range of the wireless device's transmission duration, includes: The total latency D(t,L) for wireless device task completion is calculated using the following formula: Where, γ i Let S and μ be the computing speed of the i-th edge server. L These represent the total workload required to be processed by the wireless device and the local computing speed, respectively, S i The amount of task offloaded by the wireless device to the i-th edge server, where t represents the transmission duration; Compare the total latency of the wireless device task completion corresponding to the first transmission duration and the second transmission duration. If the total latency of the wireless device task completion corresponding to the first transmission duration is greater than the total latency of the wireless device task completion corresponding to the second transmission duration, then set the lower limit of the operating range of the wireless device transmission duration to the first transmission duration; otherwise, set the upper limit of the operating range of the wireless device transmission duration to the second transmission duration.

2. The single-user task offloading method based on downlink NOMA according to claim 1, characterized in that, The step of searching for the optimal offloading task allocation scheme for the first transmission duration and the second transmission duration, respectively, under the condition of satisfying the maximum power that the wireless device can transmit, also includes: or until at this time and Reduce the offloading workload of the edge server with the worst channel gain by ΔS k Initialize ΔS k The upper and lower limits are taken as ΔS. k The average of the upper and lower limits The offloading task allocation S1,…,S for wireless devices is calculated using the following formula. N And the required transmit power P of the wireless device under the offloading task allocation N ' OMA : At this point, if P′ NOMA >P max Then let: otherwise Continue to judge If the value is greater than a preset threshold, then continue taking ΔS. k The average of the upper and lower limits Calculate the offloading task allocation S1,…,S for wireless devices. N And the required transmit power P′ of the wireless device under this offloading task allocation. NOMA ,until If the value is not greater than a preset threshold, output the wireless device unloading task allocation scheme.

3. A single-user task offloading device based on downlink NOMA, comprising a processor and a memory storing a plurality of computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the method according to any one of claims 1 to 2.

Images