CN-122028113-A - Mobile edge computing uplink communication and computing joint optimization method and related equipment

CN122028113ACN 122028113 ACN122028113 ACN 122028113ACN-122028113-A

Abstract

The embodiment of the application provides a mobile edge computing uplink communication and computing joint optimization method and related equipment, and belongs to the technical field of wireless communication and edge computing. The method comprises the steps of constructing an uplink mobile edge computing system model of a base station and a multi-terminal provided with a steerable antenna array, deriving a maximum computing time delay expression of the system, establishing a mathematical model of joint optimization with the aim of minimizing the time delay, and solving a non-convex problem iteratively into a computing side sub-problem and a communication side sub-problem by adopting an alternative optimizing framework, wherein the computing side sub-problem is solved by a closed solution and convex optimizing technology, and the communication side sub-problem is further subjected to half-positive relaxation and binary search to optimize beam forming and is subjected to fractional programming and continuous convex approximation to optimize antenna pointing. According to the application, through collaborative optimization of antenna space pointing and communication computing resources, the wireless channel is actively improved, the maximum computing time delay of the system is effectively reduced, and the service capability of the system to low-time-delay business is improved.

Inventors

ZHENG BEIXIONG
WANG QIYAO
LUO DANDAN
LIN SHAOE
TANG JIE

Assignees

华南理工大学

Dates

Publication Date: 20260512
Application Date: 20260212

Claims (10)

1. A method for jointly optimizing mobile edge computing uplink communication and computing, the method comprising the steps of: The method comprises the following steps of S1, establishing a system model for calculating an uplink communication system based on the mobile edge of a steerable antenna, wherein a base station is provided with an antenna array formed by a plurality of steerable antennas, and each terminal device is provided with a single omni-directional antenna; S2, deducing the calculation time delay of each terminal device based on the system model, and establishing a closed expression of the maximum calculation time delay of the system; S3, constructing a joint optimization problem P1 aiming at minimizing the maximum calculation time delay of the system based on a closed expression of the maximum calculation time delay of the system, wherein the optimization variables comprise the unloading data size of each terminal device, the calculation resources distributed to each terminal device by an edge calculation server, a base station receiving beam forming vector and a directional matrix of a steerable antenna, and the constraint conditions comprise the constraint of the total amount of the edge calculation resources, the constraint of the norm of the receiving beam forming vector, the constraint of the main shaft deflection angle of the steerable antenna beam and the constraint of the unloading data size range; S4, decomposing the joint optimization problem P1 into two sub-problems by adopting an alternative optimization strategy to carry out iterative solution, wherein the first sub-problem P2 is used for joint optimization of unloading data size and edge computing resource allocation under the condition of fixing a received beam forming vector and a steerable antenna pointing matrix; s5, solving the first sub-problem P2 to obtain the optimal unloading data size and the computing resource allocation scheme under the current communication parameters; s6, solving the second sub-problem P3 to obtain updated received beam forming vectors and steerable antenna pointing matrixes under the current calculation parameters; And S7, alternately and repeatedly executing the step S5 and the step S6 until a preset convergence condition is met, and outputting the finally optimized unloading data size, the calculation resource allocation, the received beam forming vector and the steerable antenna pointing matrix.
2. The method according to claim 1, wherein the system model in step S1 is specifically: the system comprises a base station and K terminal devices, wherein the base station is provided with a uniform plane array formed by M steerable antennas, the plane array is arranged in an x-y plane of a three-dimensional Cartesian coordinate system, the reference orientation of each antenna is parallel to the positive direction of an x axis, and the beam main axis direction of each steerable antenna is formed by zenith angles And azimuth angle Definition of the zenith angle Is constrained to a range of values Wherein For a preset maximum allowed zenith angle, and for representing beam pointing, a unit-mode pointing vector is defined for each steerable antenna.
3. The method according to claim 1, wherein in step S2, the calculation delay of the terminal device k is calculated Expressed as the maximum of the local computation delay and the edge computation delay, i.e Wherein For the local computation of the time delay, The maximum calculation time delay of the system is calculated for the total unloading time delay including the data transmission time delay and the edge processing time delay 。
4. The method according to claim 1, wherein the mathematical expression of the joint optimization problem P1 constructed in step S3 is: Wherein, the Representing the maximum computational delay of the system, A task offload data amount vector representing each terminal device, Representing the computing resource vectors allocated by the edge server for each terminal, Representing the base station receiving beamforming matrix, Representing the pointing matrix of the steerable antenna, Representation for receiving the th The beamforming vectors of the individual terminal signals, Indicating the index of the terminal device, Represent the first The steerable antennas being relative to a reference direction Is provided with a pointing deflection angle of (a), Represent the first The directional unit vectors of the individual steerable antennas, A steerable antenna index is represented and, Represent the first The amount of task data offloaded by the individual terminal devices to the edge server, Represent the first The total amount of data of the tasks of the individual terminal devices, Representing the total number of terminal devices in the system, Representation allocation to the first The edge computing resources of the individual terminal devices, Representing the maximum computational resource constraints that the edge server can provide.
5. The method according to claim 4, wherein the solving of the first sub-problem P2 in step S5 comprises the following sub-steps: s5.1 at fixed computing resource Allocation Determining an optimal offload data size by closed-form solution ; S5.2, the optimal unloading data size Substitution with respect to computing resource allocation Convex optimization problem P2.1; S5.3, solving the convex optimization problem P2.1 by using a KKT condition and a binary search method to obtain an optimal computing resource allocation scheme 。
6. The method according to claim 5, characterized in that said solving the second sub-problem P3 in step S6 comprises the following sub-steps: S6.1 optimal offload data size based on step S5 And computing resource allocation Converting the second sub-problem P3 into a problem P3.1 which aims at minimizing the maximum edge calculation time delay and takes the constraint of signal to interference and noise ratio as a core; S6.2, introducing auxiliary variables to convert the problem P3.1 into an equivalent problem P3.2; S6.3, adopting an alternate optimization framework to process the problem P3.2, and decomposing the problem P3.3 into a beam forming optimization sub-problem P3.4 and an antenna pointing optimization sub-problem P3.4; s6.4 optimizing the sub-problem for beamforming P3.3 at a given antenna pointing matrix Under the condition of (1) using semi-positive relaxation technology to relax non-convex rank-one constraint, and solving to obtain high-quality received beam forming vector by combining binary search method with Gaussian randomization process ; S6.5, for the antenna pointing optimization sub-problem P3.4, under the condition of given received beam forming vector w, auxiliary variables are introduced by using a split-type programming technology, non-convex constraint is processed by using a continuous convex approximation method, and a local optimal steerable antenna pointing matrix is obtained by iterative solution 。
7. The method according to claim 6, wherein in step S6.4, for a given antenna pointing matrix The beamforming optimization sub-problem P3.3 is expressed as a threshold with respect to minimizing the common signal-to-interference-plus-noise ratio Is determined for a given set by dichotomy Whether a value exists a feasible semi-positive programming solution, and finally recovering a feasible rank-one beamforming vector from the SDR solution by using Gaussian randomization.
8. The method according to claim 6, wherein in step S6.5, for a given received beamforming vector By introducing auxiliary variables The partial SINR constraint in the antenna pointing optimization sub-problem P3.4 is converted into a quadratic form, and then, the quadratic form is fixed Iterative optimization orientation matrix And (3) approximating the non-convex problem to a convex problem by using a method of first-order Taylor expansion and construction of a secondary upper bound function in each iteration, and calling a convex optimization solver to solve until the algorithm converges.
9. The method according to claim 1, wherein the convergence condition in the step S7 is that the maximum computation delay of the system is the following two adjacent iterations The variation of (2) is smaller than a preset threshold value Or the iteration number reaches the preset maximum iteration number.
10. An electronic device comprising a memory storing a computer program and a processor implementing the method of any one of claims 1 to 9 when the computer program is executed by the processor.

Description

Mobile edge computing uplink communication and computing joint optimization method and related equipment Technical Field The application relates to the technical field of wireless communication and edge calculation, in particular to a mobile edge calculation uplink communication and calculation combined optimization method and related equipment. Background With the development of low-delay services such as the Internet of things, augmented reality, automatic driving and the like, a mobile edge computing technology has been developed. The MEC can effectively reduce the end-to-end time delay and the terminal energy consumption by offloading the computing task from the terminal equipment to the server at the network edge for processing. However, the performance of existing MEC systems is often limited by unstable wireless channels and limited communication resources. The communication process of task unloading is tightly coupled with the computing process of the edge server, so that the overall optimal performance is difficult to realize by simple communication resource optimization or computing resource scheduling, and high system time delay determined by bottleneck users is easy to generate. On the other hand, steerable antenna technology has recently received attention in the field of wireless communications because of its ability to dynamically adjust the radiation main lobe direction, provide directional gain, and suppress interference. However, the existing researches focus on improving the safety, perception or simple channel capacity of a physical layer by utilizing RA, and no mature scheme is introduced into an MEC framework, and deep joint optimization is performed with calculation task unloading and resource allocation so as to systematically solve the time delay bottleneck problem in MEC. Therefore, how to design an effective joint optimization method, fully utilize the spatial degree of freedom introduced by the steerable antenna, actively improve the multi-user wireless channel condition, and cooperatively optimize the communication and computing resources, thereby remarkably reducing the maximum computing time delay of the MEC system, and becoming a technical problem to be solved urgently. Disclosure of Invention The main purpose of the embodiment of the application is to provide a mobile edge calculation uplink communication and calculation combined optimization method, electronic equipment, a storage medium and a program product based on a steerable antenna, wherein a unified optimization framework containing the directional freedom degree of the steerable antenna is constructed, and an efficient solving algorithm is designed, so that the cooperative configuration of communication and calculation resources is realized, the maximum calculation time delay of a system is effectively reduced, and the service capability of low-time-delay business is improved. In order to achieve the above objective, an aspect of an embodiment of the present application provides a method for jointly optimizing uplink communication and computation of mobile edge computation, where the method includes: The method comprises the following steps of S1, establishing a system model for calculating an uplink communication system based on the mobile edge of a steerable antenna, wherein a base station is provided with an antenna array formed by a plurality of steerable antennas, and each terminal device is provided with a single omni-directional antenna; S2, deducing the calculation time delay of each terminal device based on the system model, and establishing a closed expression of the maximum calculation time delay of the system; S3, constructing a joint optimization problem P1 aiming at minimizing the maximum calculation time delay of the system based on a closed expression of the maximum calculation time delay of the system, wherein the optimization variables comprise the unloading data size of each terminal device, the calculation resources distributed to each terminal device by an edge calculation server, a base station receiving beam forming vector and a directional matrix of a steerable antenna, and the constraint conditions comprise the constraint of the total amount of the edge calculation resources, the constraint of the norm of the receiving beam forming vector, the constraint of the main shaft deflection angle of the steerable antenna beam and the constraint of the unloading data size range; S4, decomposing the joint optimization problem P1 into two sub-problems by adopting an alternative optimization strategy to carry out iterative solution, wherein the first sub-problem P2 is used for joint optimization of unloading data size and edge computing resource allocation under the condition of fixing a received beam forming vector and a steerable antenna pointing matrix; s5, solving the first sub-problem P2 to obtain the optimal unloading data size and the computing resource allocation scheme under the current communi