A multi-layer computing task offloading method and device assisted by reconfigurable intelligent surface

Abstract

The present invention provides a method and device for offloading multi-layer computing tasks assisted by a reconfigurable intelligent surface, and relates to the field of communication technology. The method comprises: constructing a multi-layer mobile edge computing network model assisted by a reconfigurable intelligent surface; designing a two-stage offloading strategy, including: offloading from a user device to a decoding forward repeater in the first stage, and offloading from a decoding forward repeater to a base station in the second stage; calculating offloading energy loss and computing energy loss; constructing an optimization problem, minimizing offloading energy loss and computing energy loss by optimizing subcarrier allocation, task volume allocation, time allocation, transmission power allocation, and reconfigurable intelligent surface beamforming vectors; decomposing the optimization problem into three sub-problems, and obtaining a local optimal solution by an alternating iterative method. Simulation results show that the present invention can significantly reduce system energy consumption.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method and device for offloading multi-layer computing tasks assisted by a reconfigurable intelligent surface. Background Art

[0002] Mobile Edge Computing (MEC) provides a way to process and analyze data through edge servers near mobile user equipment (UE), rather than relying on cloud computing capabilities in centralized data centers. The introduction of MEC effectively liberates mobile devices from heavy computing tasks, and is particularly suitable for emerging computing-intensive and delay-sensitive applications for fifth-generation mobile networks. As a key component in MEC systems, offloading strategies transfer data processing tasks from user equipment to network edge services or even cloud servers through wireless communications. However, considering the limitations of offloading and computing, energy resources are relatively limited. Therefore, reducing energy consumption becomes a vital task while meeting constraints such as computing power and task latency.

[0003] At present, researchers have proposed some offloading strategies to minimize the energy consumption of mobile electronic devices. These studies not only focus on the scheduling strategies of single users, but also explore multi-user scheduling strategies to reduce computing energy consumption and latency. To further improve performance, multi-layer edge computing introduces an additional computing layer consisting of network edge relay nodes to process distributed data from the cloud to IoT terminals more flexibly and efficiently. In addition, the deployment of reconfigurable intelligent surfaces (RIS) can help improve offloading efficiency. Recently, a study proposed a transmissive reconfigurable metasurface transceiver to enable a multi-layer computing network architecture to reduce energy consumption. However, according to the search, there is no report on the use of RIS to assist multi-layer computing systems. Summary of the invention

[0004] In view of the above problems, the object of the present invention is to provide a multi-layer computing task offloading method and device assisted by a reconfigurable intelligent surface, which offloads the computing task bits from a user equipment (UE) to a decode-and-forward (DF) repeater, and then further offloads them to a base station (BS). Both stages of wireless communication are assisted by RIS, which can significantly reduce energy consumption.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] On the one hand, a multi-layer computing task offloading method assisted by a reconfigurable smart surface is provided, the method comprising the following steps:

[0007] S1. Construct a multi-layer mobile edge computing network model assisted by reconfigurable smart surface;

[0008] S2. Design a two-stage offloading strategy, including: the first stage is offloading from the user equipment to the decoding forward repeater, and the second stage is offloading from the decoding forward repeater to the base station;

[0009] S3, calculating unloading energy loss and calculating energy loss;

[0010] S4. Construct an optimization problem to minimize the offloading energy loss and computing energy loss by optimizing subcarrier allocation, task allocation, time allocation, transmit power allocation and reconfigurable smart surface beamforming vector;

[0011] S5. Decompose the optimization problem into three sub-problems and obtain the local optimal solution through alternating iteration method.

[0012] Optionally, the step S1 specifically includes:

[0013] A multi-layer mobile edge computing MEC network model assisted by a reconfigurable intelligent surface RIS is constructed, which includes K user equipments UE, a decoding forward repeater (DF relay), and a base station BS; wherein the DF relay, BS and each UE are equipped with a single antenna, and a RIS consists of M antenna arrays to assist in computing, wherein the channel bandwidth B is divided into N subcarriers, and the bandwidth of each subcarrier is W = B / N.

[0014] Optionally, the step S2 specifically includes:

[0015] Assume that the DF relay and the computing node CN are located at the same location, and the BS and MEC server are also located at the same location; the data volume of the kth user is D k , the amount of data is successfully transmitted within the time length T; the time length T is divided into three time periods, where the first time period is denoted as t I , the amount of data uninstalled by the kth user To the DF relay, the DF relay is in the remaining Tt I The amount of data executed in time is task; in the second time period DF relay unloaded data volume to BS; in the third time period In the MEC server, the computing offloading task is performed; the offloading strategy includes two stages: the first stage is offloading to the DF relay, and the second stage is offloading to the BS.

[0016] Optionally, in step S3, the calculation of the unloading energy loss is as follows:

[0017] When unloading to the DF relay in the first stage, the amount of data unloaded by the kth user is To DF relay, the transmission power is On the nth subcarrier, the channel gain from the kth user to the DF relay is defined as h k,n ; The channel gain from the kth user to the RIS is g k,n ; The channel gain from RIS to DF relay is h n ; Therefore, the achievable data rate from the kth user to the DF relay is described as:

[0018]

[0019] in, is the power of Gaussian white noise, diag(h n ) is h n The diagonal matrix of represents the RIS beamforming matrix; v I =θ I 1 M×1 represents the RIS beamforming vector; definition Is the nth subcarrier allocated to the kth user? When the nth subcarrier is allocated to the kth user, otherwise,

[0020] Therefore, the amount of tasks that the kth user offloads to the DF relay is The unloading energy loss in the first time period is:

[0021]

[0022] When unloading to the BS in the second phase, in the second time period When the received signal is successfully decoded, the DF relay unloads the task to BS; at this time, on the nth subcarrier, assuming θ II and f n denote the channel gains from BS to DF relay, from DF relay to RIS, and from RIS to BS respectively; let represents the transmission power of the DF relay on the nth subcarrier, let represents the subcarrier allocation variable;

[0023] Therefore, in the second period The achievable data rate of the received signal at the BS is expressed as:

[0024]

[0025] in, And v II =θ II 1 M×1 , A H Denotes the conjugate transpose of matrix A. The amount of task data offloaded from the kth user to the BS is

[0026] MEC performs computation offloading tasks in the remaining time, and the transmission energy loss in the second time period is:

[0027]

[0028] Therefore, the total unloading energy loss is expressed as:

[0029]

[0030] Optionally, in step S3, the energy loss is calculated as follows:

[0031] The computing process is divided into three categories: the first category is local computing, the second category is DF relay computing, and the third category is MEC computing;

[0032] In the first type of local computation, the kth user performs tasks in the entire time period T Use CPU frequency; let c t represents the number of MEC CPU cycles in the first type of local computation to calculate the task bits for each user, so there is where f t,k and f t,max They represent the CPU frequency and maximum CPU frequency required by the local server to calculate the data of the kth user, respectively. t represents the effective capacitance factor of the local server; the total local computing energy loss of all users is when hour, Taking the minimum value, the local calculated energy loss is:

[0033]

[0034] In the second type of DF relay calculation, after decoding the offload task received from the kth user, the DF relay performs the task amount in the remaining time Assume that DF relay can execute the offloading tasks of different users in parallel, let c r represents the number of CPU cycles of MEC in the second type of DF relay calculation to calculate the task bits of each user, a r represents the effective capacitance factor of the DF relay computing server; let fr,k and f r,max They represent the CPU frequency and the maximum CPU frequency required by CN to calculate the data of the kth user. Since the DF relay performs the offloading task for all users, we get ∑ k f r，k ≤f r，max , Therefore, the energy loss of DF relay is calculated as:

[0035]

[0036] In the third type of MEC calculation, in the third time period, the MEC server performs the calculation offloading task; it is assumed that MEC can execute the offloading tasks of different users in parallel, and let c m represents the number of MEC CPU cycles in the third type of MEC calculation to calculate the task bits of each user; let f m,k and f m,max It represents the CPU frequency and maximum CPU frequency required by the MEC server to calculate the data of the kth user; thus, ∑ k f m，k ≤f m，max and

[0037] Optionally, in step S4, the optimization problem is expressed as:

[0038]

[0039] in, represents the subcarrier allocation matrix, represents the task allocation matrix, represents the time allocation matrix, represents the transmit power allocation matrix, p max represents the maximum transmission power; |a| represents the absolute value of element a, v s [m] represents vector v s The mth element of I 、v II They represent the RIS beamforming vectors of the two stages respectively; C1 and C2 represent the input task constraints; C3 and C4 represent the transmission power constraints of the k-th user and DF relay respectively; C5, C6, and C7 represent the subcarrier allocation constraints at different time points; C8 and C9 represent the computing power constraints of the k-th user, CN, and MEC server; C10 and C11 represent the time allocation constraints; C12 represents the RIS beamforming matrix constraints.

[0040] Optionally, the step S5 specifically includes:

[0041] Decompose the optimization problem into three sub-problems: sub-problem 1, sub-problem 2 and sub-problem 3;

[0042] In subproblem 1, fix variables A, P, and v I and v II , jointly optimize D and T; in subproblem 2, fix the variables T, D, v I and v II , jointly optimize A and P; in subproblem 3, jointly optimize v I and v II , and keep other variables fixed; then, obtain the local optimal solution by alternately iterating subproblems 1, 2, and 3;

[0043] For subproblem 1, the task allocation matrix and the time allocation matrix are jointly optimized; therefore, subproblem 1 is reformulated as:

[0044]

[0045] The unloading energy loss is an affine function of the time distribution. About The convex function of About and Convex function, constraints C1, C2, C8, C10, C11 are affine, C9 is non-convex, which makes sub-problem 1 a non-convex function; use SCA to calculate the first-order Taylor expansion of the left-hand side of the two equations in C9, and get the upper term of these two terms respectively; for the first inequality of C9, the left side The previous term was:

[0046] C12:

[0047] in yes The value at the i-th generation, t I(i) t I The value in the i-th generation; for the left side of the second inequality of C9 For the last term, it is represented by:

[0048] C13:

[0049] in, yes The value at generation i, for The value at the i-th generation; thus, the approximate convex form of subproblem 1 is obtained:

[0050]

[0051] The above problems are solved directly through the CVX tool library;

[0052] Subproblem 2: Jointly optimize the subcarrier allocation matrix A and the transmit power allocation matrix P; the binary variable A makes the C4-C7 constraints non-convex. To avoid the binary attribute of C5, the binary variable is relaxed to:

[0053] C14:

[0054] Then, by introducing an auxiliary variable Decouple A and P in unloading energy loss; Subproblem 2 is reformulated as:

[0055]

[0056] In subproblem 2, the objective function is affine, constraints C4, C6, C7, C15, C16 are affine, and C1 is convex; therefore, subproblem 2 is a standard convex function and can be solved directly using the CVX tool;

[0057] Subproblem 3 performs RIS beamforming, and the optimization variable is v I and v II , and keep the variables A, D, T, and P unchanged; design a beamforming vector optimization problem whose goal is to maximize the total data rate in each time period, expressed as:

[0058]

[0059] Subproblem 3 is non-convex. To solve this problem, define and make and rank() and tr() represent the rank and trace of the matrix respectively; thus we get:

[0060]

[0061] The equivalent form of subproblem 3 is expressed as:

[0062]

[0063] in Representation Matrix The existence of the non-convex rank 1 constraint makes the problem non-convex. The rank 1 constraint can be transformed into:

[0064]

[0065] At this time, the penalty function framework is used to solve this problem.

[0066] On the other hand, a reconfigurable intelligent surface-assisted multi-layer computing task offloading device is provided, which is used to implement any of the above methods, and the device includes:

[0067] A model building module for building a multi-layer mobile edge computing network model assisted by a reconfigurable smart surface;

[0068] A strategy design module, used for designing a two-stage offloading strategy, including: offloading from the user equipment to the decoding forward repeater in the first stage, and offloading from the decoding forward repeater to the base station in the second stage;

[0069] A calculation module, used for calculating unloading energy loss and calculating energy loss;

[0070] Optimization problem building module, used to build optimization problems, minimizing offloading energy loss and computing energy loss by optimizing subcarrier allocation, task allocation, time allocation, transmit power allocation and reconfigurable smart surface beamforming vector;

[0071] The solving module is used to decompose the optimization problem into three sub-problems and obtain the local optimal solution through the alternating iterative method.

[0072] In another aspect, an electronic device is provided, the electronic device comprising:

[0073] processor;

[0074] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are loaded and executed by the processor, the steps of the multi-layer computing task offloading method as described above are implemented.

[0075] On the other hand, a computer-readable storage medium is provided, in which program code is stored. The program code can be called by a processor to execute the steps of the multi-layer computing task offloading method as described above.

[0076] The beneficial effects brought about by the technical solution provided by the present invention include at least:

[0077] The RIS-assisted multi-layer computing task offloading method and device proposed in the present invention provides an algorithm for minimizing the total energy consumption of the system, in which subcarrier allocation, power allocation, time allocation, task bit allocation and RIS beamforming vector are jointly optimized. Simulation results show that compared with traditional TL-MEC and MLC systems, the present invention has more advantages and can significantly reduce system energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0079] Figure 1 It is a flow chart of a multi-layer computing task offloading method assisted by a reconfigurable intelligent surface provided by an embodiment of the present invention;

[0080] Figure 2 It is a schematic diagram of a multi-layer MEC network model assisted by RIS provided in an embodiment of the present invention;

[0081] Figure 3 is a schematic diagram of time planning of user k within a time period T provided by an embodiment of the present invention;

[0082] Figure 4 is a comparative schematic diagram of the relationship between the total energy consumption and the number of RIS units provided by an embodiment of the present invention;

[0083] Figure 5 is a schematic diagram comparing changes in total energy consumption under different task input data amounts provided by an embodiment of the present invention;

[0084] Figure 6 It is a structural schematic diagram of a reconfigurable intelligent surface-assisted multi-layer computing task offloading device provided by an embodiment of the present invention;

[0085] Figure 7 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0086] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0087] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, sometimes subscripts such as W 1It may be written in non-subscript form such as W1. When the difference is not emphasized, the meaning is the same.

[0088] The embodiment of the present invention provides a multi-layer computing task offloading method assisted by a reconfigurable intelligent surface, such as Figure 1 As shown, the method comprises the following steps:

[0089] S1. Construct a multi-layer mobile edge computing network model assisted by reconfigurable smart surfaces.

[0090] In this step, a multi-layer mobile edge computing MEC network model assisted by a reconfigurable intelligent surface RIS is constructed, such as Figure 2 As shown, it includes K user equipments UE, a decoding forward repeater, i.e., DF relay, and a base station BS; wherein, the DF relay, BS and each UE are equipped with a single antenna, and a RIS consists of M antenna arrays to assist in calculation, wherein the channel bandwidth B is divided into N subcarriers, each subcarrier bandwidth is W=B / N, and the cyclic prefix code is large enough to ignore inter-symbol interference.

[0091] S2. Design a two-stage offloading strategy, including: the first stage is offloading from the user equipment to the decoding forward repeater, and the second stage is offloading from the decoding forward repeater to the base station.

[0092] In this step, the DF relay and the computing node CN are assumed to be located at the same location and connected by a high-throughput, low-latency optical fiber; the BS and MEC server are also located at the same location and connected by a high-throughput, low-latency optical fiber, so the data transmission between them can be considered to be delay-free. The data volume of the kth user is D k , the amount of data is successfully transmitted within the time length T. The time length T is divided into three time periods, such as Figure 3 As shown, the first time period is denoted as t I , the amount of data uninstalled by the kth user To the DF relay, the DF relay is in the remaining Tt I The amount of data executed in time is task; in the second time period DF relay unloaded data volume to BS; in the third time period In the MEC server, the computation offloading task is performed. Since the computation result bits are usually much smaller than the computation input bits, the download time of downlink communication can be ignored. The offloading strategy includes two stages: the first stage is offloading to the DF relay, and the second stage is offloading to the BS.

[0093] S3, calculating the unloading energy loss and calculating the energy loss.

[0094] The calculation of unloading energy loss is as follows:

[0095] When unloading to the DF relay in the first stage, the amount of data unloaded by the kth user is To DF relay, the transmission power is On the nth subcarrier, the channel gain from the kth user to the DF relay is defined as h k,n ; The channel gain from the kth user to the RIS is g k,n ; The channel gain from RIS to DF relay is h n ; Therefore, the achievable data rate from the kth user to the DF relay is described as:

[0096]

[0097] in, is the power of Gaussian white noise, diag(h n ) is h n The diagonal matrix of represents the RIS beamforming matrix; v I =θ I 1 M×1 represents the RIS beamforming vector; definition Is the nth subcarrier allocated to the kth user? When the nth subcarrier is allocated to the kth user, otherwise, Note that each subcarrier is assigned to only one UE, and each UE is assigned at least one subcarrier.

[0098] Therefore, the amount of tasks that the kth user offloads to the DF relay is The unloading energy loss in the first time period is:

[0099]

[0100] When unloading to the BS in the second phase, in the second time period When the received signal is successfully decoded, the DF relay unloads the task to BS; at this time, on the nth subcarrier, assuming θ II and f n denote the channel gains from BS to DF relay, from DF relay to RIS, and from RIS to BS respectively; let represents the transmission power of the DF relay on the nth subcarrier, let represents the subcarrier allocation variable;

[0101] Therefore, in the second period The achievable data rate of the received signal at the BS is expressed as:

[0102]

[0103] in, And v II =θ II 1 M×1 , A H Denotes the conjugate transpose of matrix A. The amount of task data offloaded from the kth user to the BS is

[0104] MEC performs computation offloading tasks in the remaining time, and the transmission energy loss in the second time period is:

[0105]

[0106] Therefore, the total unloading energy loss is expressed as:

[0107]

[0108] The calculation to calculate the energy loss is as follows:

[0109] The computing process is divided into three categories: the first category is local computing, the second category is DF relay computing, and the third category is MEC computing;

[0110] In the first type of local computation, the kth user performs tasks in the entire time period T Use CPU frequency; let c t represents the number of MEC CPU cycles in the first type of local computation to calculate the task bits for each user, so there is where f t,k and f t,max They represent the CPU frequency and maximum CPU frequency required by the local server to calculate the data of the kth user, respectively. t represents the effective capacitance factor of the local server; the total local computing energy loss of all users is when hour, Taking the minimum value, the local calculated energy loss is:

[0111]

[0112] In the second type of DF relay calculation, after decoding the offload task received from the kth user, the DF relay performs the task amount in the remaining time Assume that DF relay can execute the offloading tasks of different users in parallel, let c r represents the number of CPU cycles of MEC in the second type of DF relay calculation to calculate the task bits of each user, a r represents the effective capacitance factor of the DF relay computing server; let f r,k and fr,max They represent the CPU frequency and the maximum CPU frequency required by CN to calculate the data of the kth user. Since the DF relay performs the offloading task for all users, we get ∑ k f r，k ≤f r，max , Therefore, the energy loss of DF relay is calculated as:

[0113]

[0114] In the third type of MEC calculation, in the third time period, the MEC server performs the calculation offloading task; it is assumed that MEC can execute the offloading tasks of different users in parallel, and let c m represents the number of MEC CPU cycles in the third type of MEC calculation to calculate the task bits of each user; let f m,k and f m,max It represents the CPU frequency and maximum CPU frequency required by the MEC server to calculate the data of the kth user; thus, ∑ k f m，k ≤f m，max

[0115] Since the BS has a regulated power supply, the energy consumption of the base station is not considered.

[0116] S4. Construct an optimization problem to minimize the offloading energy loss and computing energy loss by optimizing subcarrier allocation, task allocation, time allocation, transmit power allocation and reconfigurable intelligent surface beamforming vector.

[0117] The present invention proposes an optimization problem under the network model setting, aiming to minimize the energy loss of offloading and calculation by optimizing subcarrier allocation, task allocation, time allocation, transmit power allocation and RIS beamforming vector.

[0118] The optimization problem is expressed as:

[0119]

[0120] in, represents the subcarrier allocation matrix, represents the task allocation matrix, represents the time allocation matrix, represents the transmit power allocation matrix, p max represents the maximum transmission power; |a| represents the absolute value of element a, v s [m] represents vector v s The mth element of I 、v IIThey represent the RIS beamforming vectors of the two stages respectively; C1 and C2 represent the input task constraints; C3 and C4 represent the transmission power constraints of the k-th user and DF relay respectively; C5, C6, and C7 represent the subcarrier allocation constraints at different time points; C8 and C9 represent the computing power constraints of the k-th user, CN, and MEC server; C10 and C11 represent the time allocation constraints; C12 represents the RIS beamforming matrix constraints.

[0121] It can be assumed that the location and channel estimation information of all devices can be perfectly obtained. At this point, an effective algorithm needs to be designed to solve the optimization problem.

[0122] S5. Decompose the optimization problem into three sub-problems and obtain the local optimal solution through alternating iteration method.

[0123] First, we need to remove the binary vector and reformulate the problem, decomposing the optimization problem into three sub-problems: Sub-problem 1, Sub-problem 2, and Sub-problem 3.

[0124] In subproblem 1, fix variables A, P, and v I and v II , jointly optimize D and T; in subproblem 2, fix the variables T, D, v I and v II , jointly optimize A and P; in subproblem 3, jointly optimize v I and v II , and keep other variables fixed; then, the local optimal solution is obtained by alternately iterating subproblems 1, 2, and 3.

[0125] For subproblem 1, the task allocation matrix and the time allocation matrix are jointly optimized; therefore, subproblem 1 is reformulated as:

[0126]

[0127] The unloading energy loss is an affine function of the time distribution. About The convex function of About and Convex function, constraints C1, C2, C8, C10, C11 are affine, C9 is non-convex, which makes sub-problem 1 a non-convex function; use SCA to calculate the first-order Taylor expansion of the left-hand side of the two equations in C9, and get the upper term of these two terms respectively; for the first inequality of C9, the left side The previous term was:

[0128] C12:

[0129] in yes The value at the i-th generation, t I(i) t I The value in the i-th generation; for the left side of the second inequality of C9 For the last term, it is represented by:

[0130] C13:

[0131] in, yes The value at generation i, for The value at the i-th generation; thus, the approximate convex form of subproblem 1 is obtained:

[0132]

[0133] The above problem is solved directly through the CVX tool library.

[0134] Subproblem 2: Jointly optimize the subcarrier allocation matrix A and the transmit power allocation matrix P; the binary variable A makes the C4-C7 constraints non-convex. To avoid the binary attribute of C5, the binary variable is relaxed to:

[0135] C14:

[0136] Then, by introducing an auxiliary variable Decouple A and P in unloading energy loss; Subproblem 2 is reformulated as:

[0137]

[0138] In subproblem 2, the objective function is affine, the constraints C4, C6, C7, C15, C16 are affine, and C1 is convex; therefore, subproblem 2 is a standard convex function and can be solved directly using the CVX tool. Note that each subcarrier can only be assigned to one user, so the maximum subcarrier assignment variable is 1, and the others are 0.

[0139] Subproblem 3 performs RIS beamforming, and the optimization variable is v I and v II , and keep the variables A, D, T, and P unchanged. However, the beamforming vector does not affect the objective function of the optimization problem. In order to efficiently solve the optimal beamforming vector, it is necessary to obtain an effective objective function that includes the beamforming vector. Since RIS is a passive device and does not consume energy, it can reduce the transmission energy by increasing the transmission rate.

[0140] The present invention designs a beamforming vector optimization problem, the goal of which is to maximize the total data rate in each time period, which is expressed as:

[0141]

[0142] Subproblem 3 is non-convex. To solve this problem, define and make and rank() and tr() represent the rank and trace of the matrix respectively; thus we get:

[0143]

[0144] The equivalent form of subproblem 3 is expressed as:

[0145]

[0146] in Representation Matrix The existence of the non-convex rank 1 constraint makes the problem non-convex. The rank 1 constraint can be transformed into:

[0147]

[0148] At this time, the penalty function framework is used to solve this problem.

[0149] In RIS-assisted multi-layer edge computing, the pseudo code of the optimization algorithm for combining communication and computing variables proposed in the present invention is shown in Table 1:

[0150] Table 1

[0151]

[0152] The effects of the present invention are verified by means of specific examples below.

[0153] Consider a three-dimensional coordinate system, where BS, DF relay, and RIS are distributed at (0m, 0m, 10m), (25m, 0m, 10m), and (25m, 5m, 10m), respectively. There are 5 communication users randomly and evenly distributed in a circular area with a radius of 25m, and the center of the circle is located at (25m, 0m, 0m). This embodiment uses the Rice channel result. The remaining parameter settings are shown in Table 2:

[0154] Table 2

[0155]

[0156] To demonstrate the effectiveness of the proposed method, the performance of the following two benchmark solutions is compared:

[0157] 1) Traditional two-layer mobile edge computing solution (TL-MEC);

[0158] 2) Multi-layer computing scheme (MLC).

[0159] Figure 4 The total energy performance is shown as a function of the number of reflective smart surface (RIS) units, where the total task input data volume is 5Mbit. Obviously, the traditional two-tier mobile edge computing (TL-MEC) and multi-tier computing (MLC) schemes are not equipped with RIS, so their total energy consumption remains unchanged. As the number of RIS units increases, the total energy consumption of the method described in the present invention decreases, which highlights the role of RIS in energy efficiency and proves the effectiveness of the two-stage task offloading assisted by RIS.

[0160] Figure 5 The variation of total energy consumption under different task input data amounts is shown. As the amount of task input data increases, the total energy consumption of the method of the present invention and all benchmark schemes also increases. The reason is that the increase in the amount of task input data leads to an increase in the CPU computing amount of the user equipment (UEs) and the DF relay, which in turn leads to higher energy consumption. It is worth noting that MLC and the method of the present invention consume significantly less energy than TL-MEC, which proves the superiority of multi-layer computing. In addition, the method of the present invention has the lowest energy consumption under different input task bits, which verifies the effectiveness of the method of the present invention.

[0161] Accordingly, an embodiment of the present invention also provides a multi-layer computing task offloading device assisted by a reconfigurable intelligent surface, such as Figure 6 As shown, the device comprises:

[0162] A model building module 201 is used to build a multi-layer mobile edge computing network model assisted by a reconfigurable intelligent surface;

[0163] The strategy design module 202 is used to design a two-stage offloading strategy, including: the first stage is offloading from the user equipment to the decoding forward repeater, and the second stage is offloading from the decoding forward repeater to the base station;

[0164] A calculation module 203, used for calculating unloading energy loss and calculating energy loss;

[0165] An optimization problem construction module 204 is used to construct an optimization problem to minimize offloading energy loss and computing energy loss by optimizing subcarrier allocation, task volume allocation, time allocation, transmit power allocation and reconfigurable intelligent surface beamforming vectors;

[0166] The solution module 205 is used to decompose the optimization problem into three sub-problems and obtain the local optimal solution through an alternating iterative method.

[0167] For ease of explanation, Figure 6Only the main components of the device are shown. The device of this embodiment can be used to perform Figure 1 The technical solution of the method embodiment shown has similar implementation principles and technical effects, which will not be repeated here.

[0168] The RIS-assisted multi-layer computing task offloading method and device proposed in the present invention provides an algorithm for minimizing the total energy consumption of the system, in which subcarrier allocation, power allocation, time allocation, task bit allocation and RIS beamforming vector are jointly optimized. Simulation results show that compared with traditional TL-MEC and MLC systems, the present invention has more advantages and can significantly reduce system energy consumption.

[0169] In an exemplary embodiment, the present invention further provides an electronic device, the electronic device comprising:

[0170] processor;

[0171] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are loaded and executed by the processor, the steps of the multi-layer computing task offloading method as described above are implemented.

[0172] Figure 7 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Figure 7 As shown, the electronic device 300 may include a processor 3001 and a memory 3002. Optionally, the electronic device 300 may further include a transceiver 3003. The processor 3001, the memory 3002, and the transceiver 3003 may be connected, for example, via a communication bus. The memory 3002 stores computer-readable instructions, which, when executed by the processor 3001, implement the steps of the above-mentioned multi-layer computing task offloading method.

[0173] In a specific implementation, as an embodiment, the processor 3001 may include one or more CPUs, such as Figure 7 CPU0 and CPU1 are shown in FIG.

[0174] In a specific implementation, as an embodiment, the electronic device 300 may also include multiple processors, such as Figure 7 3001 and processor 3004 are shown in FIG. Each of these processors may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). A processor herein may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0175] The memory 3002 is used to store the software program for executing the solution of the present invention, and the execution is controlled by the processor 3001. The specific implementation method can refer to the above method embodiment, which will not be repeated here.

[0176] The transceiver 3003 is used to communicate with a network device or a terminal device.

[0177] Optionally, the transceiver 3003 may include a receiver and a transmitter, wherein the receiver is used to implement a receiving function, and the transmitter is used to implement a sending function.

[0178] Optionally, the transceiver 3003 may be integrated with the processor 3001 , or may exist independently and be coupled to the processor 3001 via an interface circuit of the electronic device 300 , which is not specifically limited in the embodiment of the present invention.

[0179] It should be noted that Figure 7 The structure of the electronic device 300 shown in the figure does not constitute a limitation on the electronic device. The actual electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. In addition, the technical effects of the electronic device 300 can refer to the technical effects of the above-mentioned method embodiment, which will not be repeated here.

[0180] In an exemplary embodiment, the present invention further provides a computer-readable storage medium, wherein at least one instruction is stored in the computer-readable storage medium, and the at least one instruction is loaded and executed by a processor to implement the steps of the multi-layer computing task offloading method as described above. For example, the computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0181] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0182] References in the specification to "one embodiment", "an embodiment", "an exemplary embodiment", "some embodiments", etc. indicate that the described embodiments may include a particular feature, structure, or characteristic, but not every embodiment may include the particular feature, structure, or characteristic. In addition, when a particular feature, structure, or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the relevant art to implement such feature, structure, or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0183] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0184] In the present invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0185] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0186] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0187] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0188] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0189] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0190] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0191] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A multi-layer computing task offloading method assisted by a reconfigurable intelligent surface, characterized in that: The following steps are involved: S1. Construct a multi-layer mobile edge computing network model assisted by reconfigurable smart surface; S2. Design a two-stage offloading strategy, including: the first stage is offloading from the user equipment to the decoding forward repeater, and the second stage is offloading from the decoding forward repeater to the base station; S3, calculating unloading energy loss and calculating energy loss; S4. Construct an optimization problem to minimize the offloading energy loss and computing energy loss by optimizing subcarrier allocation, task allocation, time allocation, transmit power allocation and reconfigurable smart surface beamforming vector; S5. Decompose the optimization problem into three sub-problems and obtain the local optimal solution through alternating iteration method; The step S1 specifically includes: A multi-layer mobile edge computing (MEC) network model assisted by a reconfigurable intelligent surface (RIS) is constructed, which includes K user equipments (UEs), a decoding forward repeater (DF relay), and a base station (BS). The DF relay, BS, and each UE are equipped with a single antenna. A RIS consists of M antenna arrays to assist in computing. The channel bandwidth B is divided into N subcarriers, and the bandwidth of each subcarrier is W = B / N. The step S2 specifically includes: Assume that the DF relay and the computing node CN are located at the same location, and the BS and MEC server are also located at the same location; the data volume of the kth user is D k , the amount of data is successfully transmitted within the time length T; the time length T is divided into three time periods, where the first time period is denoted as t I , the amount of data uninstalled by the kth user To the DF relay, the DF relay is in the remaining Tt I The amount of data executed in time is task; in the second time period DF relay unloaded data volume to BS; in the third time period In the MEC server, the computing offloading task is performed; the offloading strategy includes two stages: the first stage is offloading to the DF relay, and the second stage is offloading to the BS; In step S3, the calculation of the unloading energy loss is as follows: When unloading to the DF relay in the first stage, the amount of data unloaded by the kth user is To DF relay, the transmission power is On the nth subcarrier, the channel gain from the kth user to the DF relay is defined as h k,n ; The channel gain from the kth user to the RIS is g k,n ; The channel gain from RIS to DF relay is h n ; Therefore, the achievable data rate from the kth user to the DF relay is described as: in, is the power of Gaussian white noise, diag(h n ) is h n The diagonal matrix of represents the RIS beamforming matrix; v I =θ I 1 M×1 represents the RIS beamforming vector; definition Is the nth subcarrier allocated to the kth user? When the nth subcarrier is allocated to the kth user, otherwise, Therefore, the amount of tasks that the kth user offloads to the DF relay is The unloading energy loss in the first time period is: When unloading to the BS in the second phase, in the second time period When the received signal is successfully decoded, the DF relay unloads the task to BS; at this time, on the nth subcarrier, assuming θ II and f n denote the channel gains from BS to DF relay, from DF relay to RIS, and from RIS to BS respectively; let represents the transmission power of the DF relay on the nth subcarrier, let represents the subcarrier allocation variable; Therefore, in the second period The achievable data rate of the received signal at the BS is expressed as: in, And v II =θ II 1 M×1 , A H Denotes the conjugate transpose of matrix A. The amount of task data offloaded from the kth user to the BS is MEC performs computation offloading tasks in the remaining time, and the transmission energy loss in the second time period is: Therefore, the total unloading energy loss is expressed as: In step S3, the energy loss is calculated as follows: The computing process is divided into three categories: the first category is local computing, the second category is DF relay computing, and the third category is MEC computing; In the first type of local computation, the kth user performs tasks in the entire time period T Let ct denote the number of CPU cycles of MEC in the first type of local computation to calculate the task bits for each user, so we have where f t,k and f t,max They represent the CPU frequency and maximum CPU frequency required by the local server to calculate the data of the kth user, respectively. t represents the effective capacitance factor of the local server; the total local computing energy loss of all users is when hour, Taking the minimum value, the local calculated energy loss is: In the second type of DF relay calculation, after decoding the offload task received from the kth user, the DF relay performs the task amount in the remaining time Assume that DF relay can execute the offloading tasks of different users in parallel, let c r represents the number of CPU cycles of MEC in the second type of DF relay calculation to calculate the task bits of each user, α r represents the effective capacitance factor of the DF relay computing server; let f r,k and f r,max They represent the CPU frequency and the maximum CPU frequency required by CN to calculate the data of the kth user. Since the DF relay performs the offloading task for all users, we get Therefore, the energy loss of DF relay is calculated as: In the third type of MEC calculation, in the third time period, the MEC server performs the calculation offloading task; it is assumed that MEC can execute the offloading tasks of different users in parallel, and let c m represents the number of MEC CPU cycles in the third type of MEC calculation to calculate the task bits of each user; let f m,k and f m,max It represents the CPU frequency and maximum CPU frequency required by the MEC server to calculate the data of the kth user; thus, Σ k f m，k ≤f m，max and In step S4, the optimization problem is expressed as: in, represents the subcarrier allocation matrix, represents the task allocation matrix, represents the time allocation matrix, represents the transmit power allocation matrix, p max represents the maximum transmission power; |a| represents the absolute value of element a, v s [m] represents vector v s The mth element of I 、v II Represent the RIS beamforming vectors of the two stages respectively; C1 and C2 represent the input task constraints; C3 and C4 represent the transmission power constraints of the k-th user and DF relay respectively; C5, C6, and C7 represent the subcarrier allocation constraints at different time points; C8 and C9 represent the computing power constraints of the k-th user, CN, and MEC server; C10 and C11 represent the time allocation constraints; C12 represents the RIS beamforming matrix constraints; The step S5 specifically includes: Decompose the optimization problem into three sub-problems: sub-problem 1, sub-problem 2 and sub-problem 3; In subproblem 1, fix variables A, P, and v I and v II , jointly optimize D and T; in subproblem 2, fix the variables T, D, v I and v II , jointly optimize A and P; in subproblem 3, jointly optimize v I and v II , and keep other variables fixed; then, obtain the local optimal solution by alternately iterating subproblems 1, 2, and 3; For subproblem 1, the task allocation matrix and the time allocation matrix are jointly optimized; therefore, subproblem 1 is reformulated as: stC1,C2,C8-C11 The unloading energy loss is an affine function of the time distribution. About The convex function of About and Convex function, constraints C1, C2, C8, C10, C11 are affine, C9 is non-convex, which makes sub-problem 1 a non-convex function; use SCA to calculate the first-order Taylor expansion of the left-hand side of the two equations in C9, and get the upper term of these two terms respectively; for the first inequality of C9, the left side The previous term was: in yes The value at the i-th generation, t I(i) The value in the i-th generation; for the left side of the second inequality of C9 For the last term, it is represented by: in, yes The value at generation i, for The value at the i-th generation; thus, the approximate convex form of subproblem 1 is obtained: stC1,C2,C8,C10,C11,C12,C13 The above problems are solved directly through the CVX tool library; Subproblem 2: Jointly optimize the subcarrier allocation matrix A and the transmit power allocation matrix P; the binary variable A makes the C4-C7 constraints non-convex. To avoid the binary attribute of C5, the binary variable is relaxed to: Then, by introducing an auxiliary variable Decouple A and P in unloading energy loss; Subproblem 2 is reformulated as: stC1,C4,C6,C7,C14 In subproblem 2, the objective function is affine, constraints C4, C6, C7, C15, C16 are affine, and C1 is convex; therefore, subproblem 2 is a standard convex function and can be solved directly using the CVX tool; Subproblem 3 performs RIS beamforming, and the optimization variable is v I and v II , and keep the variables A, D, T, and P unchanged; design a beamforming vector optimization problem whose goal is to maximize the total data rate in each time period, expressed as: stC1,C12 Subproblem 3 is non-convex. To solve this problem, define and make and rank() and tr() represent the rank and trace of the matrix respectively; thus we get: The equivalent form of subproblem 3 is expressed as: in Representation Matrix The existence of the non-convex rank 1 constraint makes the problem non-convex. The rank 1 constraint can be transformed into: At this time, the penalty function framework is used to solve this problem.

2. A multi-layer computing task offloading device assisted by a reconfigurable intelligent surface, used to implement the method according to claim 1, characterized in that: The device comprises: A model building module for building a multi-layer mobile edge computing network model assisted by a reconfigurable smart surface; A strategy design module, used for designing a two-stage offloading strategy, including: offloading from the user equipment to the decoding forward repeater in the first stage, and offloading from the decoding forward repeater to the base station in the second stage; A calculation module, used for calculating unloading energy loss and calculating energy loss; Optimization problem building module, used to build optimization problems, minimizing offloading energy loss and computing energy loss by optimizing subcarrier allocation, task allocation, time allocation, transmit power allocation and reconfigurable smart surface beamforming vector; The solving module is used to decompose the optimization problem into three sub-problems and obtain the local optimal solution through the alternating iterative method.

3. An electronic device, characterized in that: The electronic device comprises: processor; A memory having computer-readable instructions stored thereon, wherein the computer-readable instructions, when loaded and executed by the processor, implement the method according to claim 1.

4. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program code, which can be called by a processor to execute the method according to claim 1.

Images