CN115665770A - IRS-BackCom energized 6G Internet of things multilayer computing method and system - Google Patents

Abstract

The invention discloses an IRS-BackCom enabled 6G Internet of things multilayer computing method and system, wherein a multilayer computing system model of the IRS-BackCom enabled 6G Internet of things is established; problem modeling is carried out, communication problems are simplified into a three-layer logic communication model, task scheduling is calculated, and an integral communication model of a task allocation scheme of local calculation and partial data unloading is obtained; the method comprises the steps of performing system problem expression and decomposition on a communication model to obtain a communication model, distributing local calculation and migration calculation problems to a user equipment layer, processing the communication model by an access point layer and a central processing unit layer, performing time scheduling on tasks, performing local calculation on problems meeting local calculation capability, performing partial data unloading calculation on problems exceeding the local calculation capability, performing decomposition calculation on the remaining problems, and realizing IRS-BackCom endowed 6G Internet of things multilayer calculation. The invention fully utilizes the optimal computing power of the user equipment, the access point and the central processing unit to carry out task computation, thereby greatly improving the efficiency of the whole communication system.

Description

IRS-BackCom energized 6G Internet of things multilayer computing method and system

Technical Field

The invention belongs to the technical field of wireless communication, intelligent reflector backscattering and multilayer calculation, and particularly relates to an IRS-BackCom enabling 6G Internet of things multilayer calculation method and system.

Background

With the development of 6G internet of things (IoT) ecosystem, more and more convenience facilities, such as implementation of applications of intelligent transportation, medical care, wearable devices, industrial automation, and the like, are provided. The need for wireless networks to support more and more user devices, while network capabilities are facing severe challenges.

The ever-present demand has led to various quality of service requirements. Such as high capacity, low latency, high reliability, low cost, etc. Such demands are difficult to meet with the limited energy budget and computing power in the internet of things devices. The computational data offloading technology and the Intelligent Reflecting Surface-backscattering technology (Intelligent Reflecting Surface-backsCom) provide solutions to this problem. Mobile Cloud Computing (MCC) and multi-access edge computing (MEC), two common computing data offloading techniques, are often used to mitigate conflicts between scarce-resource demands and scarce-resource internet-of-things devices.

The key idea of Mobile Cloud Computing (MCC) is to offload data bits from IOT devices to a compute-rich remote cloud data center for processing. This approach has some disadvantages such as high latency, large round trip consumption problems, etc. In contrast to MCC, MEC technology provides a more suitable approach to efficiently handle delay-sensitive service issues, such as real-time signal processing. This is mainly because in MEC technology the computation units are pushed to the edge of the network. In spite of the low latency advantage of edge deployment of computing units, the MEC technology has the problem of limited computing power and cannot meet the requirements of high-load task execution.

Multi-tier computing systems integrate MCCs and MECs, combining their respective advantages to improve the processing and feedback capabilities of data offload systems. Computing units are deployed in a hierarchical manner in a multi-tier computing system. When the load of requested computing tasks of the Internet of things equipment exceeds the data processing capacity of a lower layer, some computing tasks are migrated to a more powerful and higher layer. Multi-tier computing systems are superior to planar computing systems in terms of latency. Although multi-tier computing techniques may reduce the data processing load of the internet of things devices, the data offloading process greatly increases the transmit power overhead and thus exacerbates the energy consumption problem of the IOT devices. The emerging intelligent reflecting surface backscattering technology makes up for the deficiency.

The smart reflector backscatter technology incorporates backscatter into the smart reflector. The passive signal transmission can be realized under the condition of no active radio frequency link. The intelligent reflecting surface is a two-dimensional Electromagnetic (EM) hyperplane, and can efficiently adjust the reflection coefficient of each element unit to change the reflection characteristic of incident electromagnetic waves so as to achieve the effects of enhancing received signals and inhibiting interference noise. The power consumption of the IRS is much lower compared to an active antenna.

Disclosure of Invention

The technical problem to be solved by the present invention is to provide a method and a system for calculating a 6G internet of things in an IRS-BackCom enabling manner in order to solve the above-mentioned deficiencies in the prior art, and to solve the technical problems of low transmission efficiency, unstable transmission and high power consumption of data among a client, an access point and a central server.

The invention adopts the following technical scheme:

an IRS-BackCom energized 6G Internet of things multilayer computing method comprises the following steps:

s1, establishing an IRS-BackCom enabled 6G Internet of things multilayer computing system model;

s2, performing problem modeling on the IRS-Backcom enabled 6G Internet of things multilayer computing system model obtained in the step S1, simplifying the communication problem into a three-layer logic communication model, obtaining a user equipment layer, an access point layer and a central processor layer for computing task scheduling, and obtaining an overall communication model of a task allocation scheme of local computing and partial data unloading;

and S3, performing system problem expression and decomposition on the communication model obtained in the step S2 to obtain a three-layer communication model, distributing the problems of local calculation and migration calculation to a user equipment layer, processing the problems by an access point layer and a central processing unit layer, performing time scheduling on tasks, performing local calculation on the problems meeting the local calculation capability, performing partial data unloading calculation on the problems exceeding the local calculation capability, performing decomposition calculation on the remaining problems, and realizing the 6G Internet of things multilayer calculation with the IRS-BackCom function.

Specifically, in step S1, the multi-layer computing system model includes:

an energy station PB, K intelligent reflector IRS assisted user equipment UE on a T1 layer, M access points AP connected with an MEC server on a T2 layer, and a central server with rich resources on a T3 layer; the energy station PB, each uniform access point AP and the central server are respectively provided with N _p ,N _a And N _c Root antenna, N _c ≥M×N _a In a hierarchical network, each user equipment UE requests to perform a computation task, each task being independent bit by bit, being split into a number of bit subsets, all channels following quasi-static fading and the channel information being known.

Specifically, in step S2, a partial data offloading policy is adopted, all the first layer user equipment UE and the second layer access point AP perform computation offloading and local computation simultaneously, the computation tasks of the user equipment UE are divided, one part is computed locally at the user equipment UE, the other part is offloaded to the access point AP, and after task transfer from the user equipment UE to the access point AP is completed, the computation task bits received by each access point AP are divided into: processed locally at the access point AP and migrated to the central server for computation.

Further, the first layer calculates the local energy consumption E calculated at the k user equipment UE locally _loc,k Comprises the following steps:

wherein, t _loc,k Representing the execution time, t, of the local calculation at the kth user equipment UE _off Indicating that the energy station PB and all the intelligent reflecting surfaces IRS are in an open state during unloading time, mu indicating that the power consumption of one element unit of the intelligent reflecting surfaces IRS is positively correlated with the phase resolution of the intelligent reflecting surfaces IRS, and L being the number of elements in the IRS;

energy consumption E of the mth access point AP _AP,m Comprises the following steps:

wherein, t _AP,m And t _mig,m Respectively representing the computation time and data migration time, p, at the mth AP _AP,m Represents the information transmission power of the mth AP, epsilon _AP,m The power consumption coefficient of the processor chip at m APs is closely related to the chip shelf,

the CPU frequency of m APs.

Furthermore, the first layer of user equipment UE adopts IRS communication, when the energy station PB utilizes the directional antenna to radiate electromagnetic waves, energy-carrying radio-frequency signals reaching the intelligent reflecting surface IRS of each user equipment UE are used for backscattering communication, when the IRS executes backscattering communication, incident signals serving as signal carriers are remodulated, and backscattering vectors theta are calculated according to the modulated incident signals _k Modulated and converted into a signal x _kr Passive beamforming vector theta of _kr 。

Further, the remodulation is specifically:

where s denotes the original data signal,

x _kr representing the modulated data signal for the kth access point AP of the r-th UE,

w _k beamforming vectors on the kth antenna group of PB.

Specifically, in step S3, the data is unloaded to the access point AP in the first phase, and the data unloaded from the access point AP to the central server is executed in the second phase; the optimization problem is expressed by jointly optimizing the active beam forming at the energy station PB, the passive beam forming at the user equipment UE, the active beam forming at the access point AP, the bandwidth and power allocation among all the user equipments UE, and the computation and bit expression of the local computation time maximization system, as follows:

C8:t _off ≤T ₁

C10:t _AP,m R _AP,m +t _mig,m R _mig,m ≥S _m

C12:T ₁ +T ₂ ＝T

wherein S is _m Represents the calculation sum bit, T, received by the mth AP ₁ And T ₂ Respectively representing the duration of the first and second phases,

are respectively represented by w _k ,θ _kr ,p _k ,t _loc,k And v _m C1 and C2 represent the active and passive beamforming constraints of PB and all UEs, respectively, C3, C4, C5 represent the power allocation or bandwidth allocation constraint between UEs, P represents the total power of PB, C5 aims to ensure fairness between UEs by setting minimum and bit maximum bandwidth size bounds, C6 is the energy constraint of each UE,

represents the energy threshold of all UEs, C7 and C8 are time constraints, C9 represents the beamforming constraint at the AP; c10 is that the number of data bits received by each AP is lower than the data processing capacity of local calculation and unloading; c11 is a time constraint;

calculating and maximizing the bit in the first stage to obtain a problem 1, and calculating and obtaining a problem 2 by a problem transformation and alternative parameter optimization method in the minimization of delay in the second stage; solving the problems 1 and 2 to obtain calculation and bit and delay minimization, decomposing the calculation problem of the voice video and the live video, and distributing time and calculation tasks to obtain time scheduling of multilayer calculation, wherein the first layer obtains a distribution scheme of local calculation and bit and partial data unloading from the first layer to the second layer and the third layer;

problem 1 is as follows:

s.t.C1-C8

problem 2 is as follows:

s.t.C9-C11

further, question 1 equivalent is expressed as:

wherein, B _k Is the bandwidth of the channel, alpha _m,k For Lagrange dual transformation of the auxiliary variable, Ω _m,k Introducing variables, β, to facilitate matrix computation _m,k For positively determining an auxiliary variable, T _m,k For the joint channel gains of the k-th to m-th,

in order to introduce variables that facilitate the computation of the matrix,

is the set of APs, Θ _kr In order to realize the purpose of the method,

is a set of UEs that are to be served,

is the set of IRS element units of the UE, and l is the l-th element unit.

Further, problem 2 is equivalently expressed as:

wherein, (.) ^H Representing the conjugate transpose of a matrix or vector, X _mig,m And Y _mig,m,r Is a variable, R _lb Is composed of

Lower boundary of (a) _mig,m To positively determine the auxiliary variable, V _m And

is a variable of the number of the main chain,

is the set of APs, B is the total bandwidth, S _m For sum of data bits, p _AP,m Is the transmit power of the mth AP.

In a second aspect, an embodiment of the present invention provides an IRS-BackCom enabled 6G internet of things multilayer computing system, including:

the system module is used for establishing a multi-layer computing system model of the IRS-BackCom enabled 6G Internet of things;

the problem module is used for carrying out problem modeling on the IRS-Backcom enabled 6G Internet of things multilayer computing system model obtained by the system module, simplifying the communication problem into a three-layer logic communication model, obtaining a user equipment layer, an access point layer and a central processor layer for computing task scheduling, and obtaining an overall communication model of a task allocation scheme of local computing and partial data unloading;

and the calculation module is used for performing system problem expression and decomposition on the communication model obtained by the problem module to obtain a three-layer communication model, distributing the local calculation and migration calculation problems to the user equipment layer, processing the problems by the access point layer and the central processing unit layer, performing time scheduling on tasks, performing local calculation on the problems meeting the local calculation capability, performing partial data unloading calculation on the problems exceeding the local calculation capability, performing decomposition calculation on the rest problems, and realizing IRS-BackCom enabled 6G Internet of things multilayer calculation.

Compared with the prior art, the invention has at least the following beneficial effects:

an IRS-BackCom enabled 6G Internet of things multilayer computing method comprises the steps of establishing an IRS-BackCom enabled 6G Internet of things multilayer computing system model; performing problem modeling on the obtained IRS-Backcom enabled 6G Internet of things multilayer computing system model to simplify the original complex communication problem into a three-layer logic communication model, and obtaining communication models of time scheduling, local computing and partial data unloading of a user equipment layer, an access point layer and a central processor layer; the original complex task allocation problem is converted into two calculation problems of maximum calculation and bit and time delay minimization; and carrying out system problem expression and decomposition to obtain an optimized calculation result.

Furthermore, the originally complex local computation and migration computation problem can be divided into three layers of processing. The method comprises the steps of distributing and processing tasks by a user equipment layer, an access point layer and a central processor layer, reasonably scheduling the tasks in time, carrying out local calculation on simple problems, carrying out partial data unloading calculation on the problems exceeding the local calculation capacity, and decomposing and efficiently calculating complex calculation problems. Therefore, the overall computing efficiency of the system can be improved, the burden of computing tasks of all layers is reduced, and the overall computing efficiency is improved.

Further, a partial data offloading strategy is adopted, all the user equipment UE and the access point AP perform computation offloading and local computation simultaneously, computation tasks of the user equipment UE are divided, one part is computed locally in the user equipment UE, the other part is offloaded to the access point AP, and after task transfer from the user equipment UE to the access point AP is completed, computation task bits received by each access point AP are divided into: the system is processed locally at the AP and is migrated to the central server for calculation, so that the local calculation capability and the calculation capability of the previous layer are reasonably used, and when the throughput of calculation tasks is larger, the system has higher calculation efficiency and calculation capability and can adapt to the calculation tasks with high concurrency and high throughput. And computing tasks and computing resources are fully utilized through partial data unloading and local computing.

Further, according to the establishment of the model, the local energy consumption E at the kth user equipment UE is respectively calculated _loc,k And energy consumption E of the mth access point AP _AP,m And then, the maximum power consumption limit of each part is calculated, so that reasonable time scheduling and task allocation of local calculation and task unloading are obtained.

Further, when the energy station PB radiates electromagnetic waves using the directional antenna, energy-carrying radio frequency signals reaching the intelligent reflection surface IRS of each user equipment UE are used for backscatter communication, and when the IRS performs backscatter communication, incident signals as signal carriers are remodulated, and a backscatter vector θ _k Modulated and converted into a signal x _kr Passive beamforming vector theta of _kr The intelligent reflector IRS can perform energy harvesting capability and retransmit information by modulation, thus achieving low-power communication capability.

Furthermore, the IRS intelligent reflecting surface remodulates and transmits the communication signal, so that the power consumption of each information transmission can be reduced, the communication efficiency of the system can be improved,

further, the first phase is unloaded into the access point AP, and the data unloaded from the access point AP to the central server is executed in the second phase; the optimization problem is expressed by jointly optimizing active beam forming at an energy station PB, passive beam forming at user equipment UE, active beam forming at an access point AP, bandwidth and power distribution among all user equipment UE and calculation and bits of a local calculation time maximization system, a problem 1 is obtained by calculation and bit maximization in a first stage, and a problem 2 is obtained by calculation and substitution parameter optimization through a problem transformation method and a substitution parameter optimization method in a second stage.

Further, the calculation of the problem 1 and the maximum equivalent expression of the bits are subjected to secondary conversion by utilizing Lagrangian dual introduced variables. Therefore, the original complex problem is solved by using methods such as a parameter substitution method, partial derivation, gaussian randomization and the like, and a solution with lower complexity is obtained.

Further, the minimization of the delay of the problem 2 is expressed equivalently, secondary optimization is utilized, and parameter optimization is carried out by replacing an optimization method, so that an optimal solution is obtained.

It is to be understood that, for the beneficial effects of the second aspect to the third aspect, reference may be made to the relevant description in the first aspect, and details are not described herein again.

In conclusion, the invention can more efficiently and reasonably carry out time scheduling and partial task unloading on the original complex calculation task problem. The optimal computing power of the user equipment, the access point and the central processing unit is fully utilized to carry out task computation, and the efficiency of the whole communication system is greatly improved.

The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.

Drawings

FIG. 1 is a schematic diagram of an IRS-BackCom multi-layer computing system;

FIG. 2 is a schematic time schedule diagram of a two-stage process;

FIG. 3 is a schematic diagram of IRS-BackCom multi-layer calculation optimization, wherein (a) is the rate of offloading from the UE to the AP, and (b) is the rate of offloading from the AP to the central server;

FIG. 4 is a diagram illustrating the effect of IRS element number of IRS-Backcom multi-layer calculation on the user equipment calculation bit sum, where (a) is the offloading rate from UE to AP, and (b) is the calculation sum bit;

fig. 5 is a diagram illustrating the effect of the total transmit power variation of an energy station calculated by IRS-backhaul multi-layer on the offload rate of a user equipment and the access point rate system calculation sum bits, where (a) is the offload rate from a UE to an AP and (b) is the calculation sum bits;

fig. 6 is a schematic diagram illustrating the influence of the average distance from the user end to the access point in the IRS-backhaul multi-layer calculation on the offloading rate and the calculated bits and bits of the system, where (a) is the offloading rate from the UE to the AP, and (b) is the calculated bits and bits;

FIG. 7 is a diagram illustrating the effect of the number of IRS-Backcom multi-layer computing access points and the number of antennas on the system computing bit sum, where (a) is the number of access points and (b) is the number of access point antennas;

FIG. 8 is a schematic flow chart of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

In the description of the present invention, it should be understood that the terms "comprises" and/or "comprising" indicate the presence of the stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.

It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

It should be further understood that the term "and/or" as used in this specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and including such combinations, e.g., a and/or B, may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the character "/" herein generally indicates that the former and latter associated objects are in an "or" relationship.

It should be understood that although the terms first, second, third, etc. may be used to describe preset ranges, etc. in embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish preset ranges from one another. For example, the first preset range may also be referred to as a second preset range, and similarly, the second preset range may also be referred to as the first preset range, without departing from the scope of the embodiments of the present invention.

The word "if" as used herein may be interpreted as "at 8230; \8230;" or "when 8230; \8230;" or "in response to a determination" or "in response to a detection", depending on the context. Similarly, the phrase "if determined" or "if detected (a stated condition or event)" may be interpreted as "upon determining" or "in response to determining" or "upon detecting (a stated condition or event)" or "in response to detecting (a stated condition or event)", depending on the context.

Various structural schematics according to the disclosed embodiments of the invention are shown in the drawings. The figures are not drawn to scale, wherein certain details are exaggerated and possibly omitted for clarity of presentation. The shapes of various regions, layers and their relative sizes and positional relationships shown in the drawings are merely exemplary, and deviations may occur in practice due to manufacturing tolerances or technical limitations, and a person skilled in the art may additionally design regions/layers having different shapes, sizes, relative positions, according to actual needs.

The invention provides a 6G Internet of things multilayer computing method with IRS-BackCom enabling, in a hierarchical network, data of a computing task requested by each User Equipment (UE) is divided into three parts which are respectively calculated on the UE of a T1 layer, an access Point AP (Acces Point) of a T2 layer and a central server of a T3 layer; compared with the traditional active antenna transmission mode, the UE utilizes the passive Intelligent reflection Surface IRS (Intelligent reflection Surface) to unload data bits to the AP; is quite different from the traditional active antenna mode; based on the established network framework; the invention aims to maximize the computation of the system and the optimization problem of bits; the present invention maximizes the computation and bit of the system by jointly optimizing active beamforming at the energy station, passive beamforming at the UE, active beamforming at the AP, bandwidth and power allocation among all UEs, and local computation time within the considered time block.

The present invention decomposes the optimization problem into two sub-problems, namely the computation and bit maximization problem in time phase 1 and the minimization of time delay in time phase 2. And the solutions of the two subproblems are obtained by using an objective function conversion and alternative optimization method. The invention carries out a large amount of simulation, verifies the feasibility of the system and shows that the system can realize high-performance calculation and bit in the aspect of processing the calculation bit.

Referring to fig. 8, the invention relates to a 6G internet of things multilayer calculation method with IRS-BackCom enablement, which includes the following steps:

s1, establishing an intelligent reflector backscatter (IRS-BackCom) enabled multilayer computing system model of the 6G Internet of things;

referring to fig. 1, the IRS-BackCom-based 6G internet of things enabled multi-tier computing system model includes an energy station PB (Power Beacon), K intelligent reflector IRS-assisted user equipments UE on a T1 tier, M access points AP on a T2 tier connected to an MEC server, and a central server on a T3 tier having rich resources. PB, each unified AP and the central server are respectively provided with N _p ,N _a And N _c Root antenna, N _c ≥M×N _a 。

Order to

The set of IRS element units and APs, representing UE, kth UE respectively, in such a hierarchical network each UE is requesting to perform a computational task. Each task is assumed to be bit-by-bit independent and thus can be partitioned into multiple bit subsets. All control links for information exchange between all entities are substantially smooth. And all channels follow quasi-static fading and the channel information is perfectly known.

The system employs a partial data offload policy. For all UEs and APs, their computation and offloading units are separated in circuit structure, thus supporting both computation offloading and local computation. The UE itself has very weak computational power and its data processing speed is rather limited. To achieve efficient computation and low latency, the UE's computation task is divided into multiple parts, one part being computed locally at the UE, the rest being offloaded to the AP. Once the task transfer from the UE to the AP is completed, the computation task bits received by each AP are further divided into two parts, one part being processed locally at the AP and the other part being migrated to the central server for computation.

S2, problem modeling is carried out on the basis of the IRS-BackCom enabled 6G Internet of things multilayer computing system model obtained in the step S1;

based on the intelligent reflector backscatter enabled 6G internet of things multilayer computing system model established in the step S1, a partial data offloading strategy is adopted, and for all the UEs and the APs, the computing units and the offloading units of the UEs and the APs are separated in the circuit structure, so that the computation offloading and the local computation are simultaneously carried out.

Due to the low computational power, the data processing speed of each UE is rather limited. For efficient computation and low latency, the UE's computation task is divided into parts. Some are computed locally at the UE and the rest are offloaded to the AP. Once the task transfer from the UE to the AP is complete, the computation task bits received by each AP are further divided into two parts. One part is processed locally at the AP and the other part is migrated to a central server for computation.

S201, establishing a calculation and energy model;

the calculation time of the data and the data migration time are transmitted in different calculation layers to calculate the consumed time and energy.

When the UE requests to process a calculation task, only a small part of task bits are calculated locally due to low calculation capability; subsequently, after the task is offloaded from the UE to the AP, the data received by the AP must be divided into two parts. One part is computed locally while the other part is migrated to the central server. Let f be _loc,k And C _loc,k Respectively representing the frequency of the CPU and the number of cycles required to compute a unit bit at the kth UE. Epsilon _loc,k Indicating the capability of the processor chip at the kth UEThe dissipation factor is closely related to the chip shelf. Let f _AP,m ,C _AP,m ，ε _AP,m Indicating the corresponding content of the mth AP.

The calculation rates performed at the kth UE and the mth AP are respectively expressed as:

because the central server has a powerful computing power with a negligible execution delay, the local energy consumption at the kth UE consists of data computation and IRS operation, expressed as:

wherein, t _loc,k Denotes the execution time, t, for local computation at the kth UE _off Indicating the unload time (PB and all IRS are on during unload), μ indicates that the power consumption of one element unit of a single IRS is positively correlated with the phase resolution of the IRS.

The IRS consumes more energy as the number of IRS element units increases. The energy consumption at the mth AP is given by:

wherein, t _AP,m And t _mig,m Respectively representing the computation time and data migration time, p, at the mth AP _AP,m Indicating the information transmission power of the mth AP.

S202, establishing a communication model

The system generally has a much smaller scale of calculation results than the calculation task, so downlink communication will not be considered; and the return delay can also beTo be reasonably ignored. A time block T is considered for uplink task offloading during which all channel gains remain unchanged. When task offloading is performed, the time block T is divided into two phases T ₁ And T ₂ Two time phases, as shown in fig. 2. In the first stage T ₁ Offloading data tasks from the UE to the AP; in a second stage T ₂ Part of task bits received by the AP are migrated to the central server; note that both phases share the same spectral resource B.

The electromagnetic waves emitted by the deployed energy stations PB at the first stage are used as the backscatter carrier signal at the IRS at each UE. Each UE occupies a different spectrum resource to transmit some of its data bits to the AP; in particular, the total spectrum resource total bandwidth B is divided into K resource blocks

Wherein B is _k Is allocated to the kth UE. The antennas at PB are divided into k groups.

Wherein, N _p,k Indicates the number of antennas of the k-th group,

the kth group of antennas points to the kth UE and shares the same frequency spectrum resource B with the kth UE _k 。

Order to

And

channel gain matrices are indicated from the kth group antenna at PB to the kth UE, from the kth group antenna of PB to the mth AP, and from the kth UE to the mth AP, respectively.

The signal transmission from the PB to the central server is ignored. Definition of theta _k And theta _k Respectively, the kth backscatter matrix and vector of the IRS, theta _k ＝diag{θ _k }。

When the PB utilizes directional antennas to transmit radiated electromagnetic waves, the energy-carrying radio frequency signals of the IRS arriving at each UE are used as backscatter communications.

When an IRS performs backscatter communications, the process of incident signal as a signal carrier being remodulated is described mathematically as:

where s denotes the original data signal,

x _kr representing the kth AP modulated data signal for the r-th UE,

symbol w _k The beamforming vector on the kth antenna of finger PB. Backscatter vector θ _k Modulated and converted to a signal x _kr Passive beamforming vector theta of _kr And is and

[·] _l,l the ith diagonal element of the matrix is represented.

And reasonably distributing the computing tasks among the UE, the AP and the central server by adopting a partial data unloading strategy. For efficient computation and low latency, the UE's computation task is divided into multiple parts. One part is computed locally at the UE and the rest is offloaded to the AP. Once the task transfer from the UE to the AP is complete, the computation task bits received by each AP are further divided into two parts. One part is processed locally at the AP and the other part is migrated to a central server for computation.

And S3, performing problem modeling based on the step S2 in the multilayer computing system model of the IRS-Backcom enabled 6G Internet of things established in the step S1 to perform system problem expression and decomposition.

The calculation and bit maximization problem of the system and the execution time of the local calculation and the time allocation between the two phases are difficult to solve directly by methods that jointly optimize the active beamforming at PB, the passive beamforming at UE, the active beamforming at AP, the bandwidth and power allocation between all UEs and the local calculation time, respectively, thus proposing that the time block is divided into two consecutive phases that are interrelated.

Some of the computed bits are offloaded to the AP in the first stage. The data offloaded from the AP to the central server is performed in the second phase. The problem can be divided into two problems corresponding to the two phases, namely phase 1 calculation and bit maximization and phase 2 delay minimization, and the problem is solved through a method of problem transformation and substitution parameter optimization.

S301, problem expression is carried out

Considering the computation task and data migration through two processes, when the computation task migration from the AP to the central server is completed, the computation result can be obtained and returned immediately due to the strong computation power of the central server. Performing optimization problem calculation by jointly optimizing active beam forming at PB, passive beam forming at UE, active beam forming at AP, bandwidth and power distribution among all UEs, local computation time maximization calculation and bits for processing as many computation bits as possible in a time block;

in the first phase, signals of different UEs are received at the mth AP, only the transmission from the kth UE is considered, the received signal at the mth AP:

wherein the content of the first and second substances,

is a complex Gaussian random vector with a power spectral density of N _m And is and

since the antenna at PB is pointed towards each UE, the AP experiences little interference from the PB. Occupying resource block B of bandwidth according to received signal _k OnSignal-to-noise ratio (SINR) at mth AP:

the sum rate from the kth UE to all APs is therefore:

in the second phase, all APs occupy the same spectrum resource B, and transmit the partially received task bits to the central server over the wireless backhaul link. Order to

Representing the channel gain matrix from the AP to the central server. The signals received at the central server are:

wherein n is (0, σ) ² I) Is a complex Gaussian random vector with a power spectral density of N and σ ² = BN. The transmission rate from the mth AP to the central server according to the received signal is:

R _mig,m ＝B log ₂ (1+γ _mig,m )

wherein, the first and the second end of the pipe are connected with each other,

due to the powerful computing power of the central server, when the migration of the computing tasks from the APs to the central server is completed, the computing results can be obtained and returned immediately. In order to process as many computation bits as possible in the considered time block, the invention maximizes the computation and bits of the system by jointly optimizing the active beamforming at PB, the passive beamforming at UE, the active beamforming at AP, the bandwidth and power allocation among all UEs and the local computation time.

The optimization problem is expressed as:

C8:t _off ≤T ₁ ,

C10:t _AP,m R _AP,m +t _mig,m R _mig,m ≥S _m ,

C12:T ₁ +T ₂ ＝T,

wherein the content of the first and second substances,

representing the computation sum bit received by the mth AP. T is ₁ And T ₂ Respectively, the duration of the first and second phases.

Are respectively represented by w _k ,θ _kr ,p _k ,t _loc,k And v _m A set of (a); c3, C4, C5 represent power allocation or bandwidth allocation constraints among the UEs, wherein P represents the total power of the PB, and C5 aims to ensure fairness among the UEs by setting upper and lower bounds of minimum and maximum bandwidths; c6 is the energy constraint for each UE,

represents the energy thresholds of all UEs; c7 and C8 are time constraints; c9 represents the beamforming constraint at the AP; c10 holds because the number of data bits received by each AP does not exceed its locally calculated and offloaded data processing capabilities; c11 is a time constraint.

Furthermore, the energy budget of the AP is not an important limiting factor and is therefore not considered as a constraint.

S302, performing problem decomposition after problem expression based on the step S301;

on the basis of establishing a multilayer calculation model, the original complex and difficult-to-solve local calculation task and partial data unloading task allocation scheme of a three-layer communication model is converted into two sub-problems which are easy to solve; namely computation and bit maximization and delay minimization problems; this ensures a sufficient utilization of the overall communication capacity of the system in order to cope with high throughput; improving the communication efficiency of the system during the peak communication time period

The non-convex optimization problem (P0) is difficult to solve directly due to the presence of multiple coupled variables. These coupling variables are the active and passive beamforming vectors, power and bandwidth, local computation execution time, and time allocation between the two phases. In the system under consideration, the time block is divided into two consecutive phases associated with each other. Some of the computation bits are offloaded to the AP in the first stage with corresponding constraints C1-C8. In the second phase, the data that is offloaded from the AP to the central server is executed, with the corresponding constraints C9-C11. Corresponding to these two phases, the problem (p 0) is divided into two problems as follows:

i.e., phase 1 computation and bit maximization and phase 2 delay minimization.

Two optimization problems, problem 1 and problem 2, for a given time allocation between the two phases, can be derived by bisection once they are solved ₁ And T ₂ With an optimal time allocation in between. A solution to problem P1 may be obtained by solving problems (P1) and (P2) and using a dichotomy.

Problem 1: phase 1 calculation and bit maximization:

s.t.C1-C8

problem 2: phase 2 latency minimization:

s.t.C9-C11.

for given time distribution, respectively giving the solving processes of the two optimization problems; thereafter, an optimal time allocation between T1 and T2 may be obtained by bisection. In other words, by solving the problems P1 and P2, a solution to the original problem P0 can be obtained by using the dichotomy.

As seen from the problem (P1), when given

The problem (P1) reduces to a linear programming problem and the objective function therein is limited only by the constraints C6-C8. This allows for a linear programming problem to be easily solvedThe solution method is omitted. On the other hand, in the case of a liquid,

depend on

They are only concerned with the constraints C1-C5. And therefore focus on the following issues.

s.t.C1-C5

In problem (P3), active beamforming, passive beamforming, power and bandwidth are coupled. Furthermore, the objective function is a sum of logarithmic functions, and this problem (P3) remains difficult to directly solve in view of these factors. In order for the problem (P3) to work, an efficient solution is proposed, the invention first converts the sum of the logarithmic functions into a more manageable form and then proposes an alternating optimization method to optimize the variables.

S3021, performing objective function conversion on the first-stage calculation and bit maximization problem by using a secondary conversion method based on the step S302;

the lagrange dual transform is used to transform the objective function in order to convert the logarithmic function into a more manageable form. Introducing an auxiliary variable alpha _m,k And an auxiliary variable vector beta _m,k Again using a quadratic transformation, the sum of the logarithmic functions is rewritten as:

wherein

Wherein

A _m,k ＝T _m,k θ _km

The problem (p 3) is newly expressed based on the new objective function as:

s.t.C1-C5

s3022, alternately optimizing problem parameters by using methods such as derivative quadratic programming (QCQP) semi-definite relaxation (SDR) and the like for the first-stage calculation and bit maximization problem based on the step S302;

optimization of variable alpha by looping _m,k ,β _m,k ,

Solving a problem (P4), comprising the steps of:

step 1: optimizing alpha _m,k And beta _m,k ；

Given a

In case of (2), respectively for alpha _m,k And beta _m,k Derivation to obtain optimum

And

order to

Then derive the optimum

Respectively as follows:

and 2, step: optimization

When given α _m,k ,β _m,k ,

Then, the problem (P4) is restated as:

s.t.C4,C5

obviously, (P5) is a quadratic programming (QCQP) problem and is therefore easy to solve.

And step 3: optimization

And

given alpha _m,k ,β _m,k ,

Thereafter, the objective function (P4) of the problem can be simplified as:

Y _k x _k respectively as follows:

then, the problem (P4) is restated as:

s.t.C1,C3

it is lifted to a higher dimension using semi-definite relaxation (SDR), let

And

this problem is then equivalently rewritten as:

by removing the rank 1 constraint C16, (P6) is relaxed to a semi-definite programming (SDP) problem, which can be easily solved with existing CVX solving software. Then, a solution of rank 1 can be restored by Singular Value Decomposition (SVD) or gaussian randomization.

And 4, step 4: optimization of

Given alpha _m,k ,β _m,k ,

The objective function of the problem (P4) is simplified to:

the problem (P4) is then restated as:

s.t.C2

deducing:

wherein the content of the first and second substances,

then the problem (P4) is equivalently restated as:

neglecting the constraint C20 of rank 1, the problem is

The top is convex and easy to solve. According to its optimal solution

The rank-1 solution may be recovered using Singular Value Decomposition (SVD) or gaussian randomization. Based on which θ can be obtained _kr 。

Problem 2: phase 2 latency minimization:

s.t.C9-C11

s3023, converting the phase two delay minimization problem based on the step S302;

the problem (P2) is equivalently rewritten as:

s.t.C9,C10

from thisIn the optimization problem, t can be easily derived _AP,m ＝t _mig,m Is a necessary condition for an optimal solution. To prove this, let t _AP,m ≤t _mig,m (ii) a According to the constraint C10 has

When t is _AP,m When the value is larger, t _mig,m And correspondingly reduced. Thus, at t _AP,m ＝t _mig,m When the utility model is used, the water is discharged,

is minimized.

The optimization problem (P2) is further equivalently expressed as:

for the sake of simplicity the problem is expressed as:

consider that

The problem is equivalently converted into:

the left side of constraint C23 according to quadratic transformation is represented as:

wherein the content of the first and second substances,

the problem (P2) then translates into:

and S3024, performing parameter alternation optimization on the second-stage time delay minimization problem by using derivation and semi-positive definite relaxation based on the step S302.

By optimizing the variable alpha _mig,m β _mig,m And

solving the problem (P8), this process is divided into the following steps:

step 1: optimizing alpha _mig,m β _mig,m ；

Given the

Respectively to alpha _mig,m β _mig,m Derivation to find optimality

The optimal is then derived:

step 2: optimization of

Given alpha _mig,m β _mig,m The constraint C24 is equivalent to:

order to

Constraint C24 is converted to:

then the problem (P8) is simplified to:

s.t.C9,C25

by performing semi-positive relaxation (SDR), this problem is equivalently translated into:

wherein, the first and the second end of the pipe are connected with each other,

ignoring the constraint C30 of rank 1, makes (P9) a convex problem. According to its optimal solution

A solution with the rank of 1 can be restored by using an SVD or Gaussian randomization method; on the basis of which obtain

In another embodiment of the invention, the system can be used for realizing the IRS-BackCom enabled 6G Internet of things multilayer computing method, and specifically, the IRS-BackCom enabled 6G Internet of things multilayer computing system comprises a system module, a problem module and a computing module.

The system module establishes an IRS-BackCom enabled 6G Internet of things multilayer computing system model;

the problem module is used for carrying out problem modeling on the IRS-Backcom enabled 6G Internet of things multilayer computing system model obtained by the system module, simplifying the communication problem into a three-layer logic communication model, obtaining a user equipment layer, an access point layer and a central processor layer for computing task scheduling, and obtaining an overall communication model of a task allocation scheme of local computing and partial data unloading;

and the calculation module is used for performing system problem expression and decomposition on the communication model obtained by the problem module to obtain a three-layer communication model, distributing the local calculation and migration calculation problems to the user equipment layer, processing the problems by the access point layer and the central processing unit layer, performing time scheduling on tasks, performing local calculation on the problems meeting the local calculation capability, performing partial data unloading calculation on the problems exceeding the local calculation capability, performing decomposition calculation on the remaining problems, and realizing the 6G Internet of things multilayer calculation endowed with the IRS-BackCom.

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The system of the invention considers the problems of calculation in a time block, maximization of bits and minimization of delay. To solve this problem the patent jointly optimizes active beamforming at PB, passive beamforming at UE, active beamforming at AP, bandwidth and power allocation among all UEs, and time of local computation.

In the multi-tier computing system under consideration, the offload rate from the UE to the AP and the computation and bit of the system may be expressed as the communication capability of the proposed IRS backscatter strategy and the task processing capability of the system, respectively. Therefore, the present invention studies how these two performance indicators depend on several important parameters, including the number of IRS element units, the total transmission power of PB, the average distance from UE to AP, the number of APs, and the number of AP antennas. In addition to the proposed optimization schemes, several simplified optimization schemes are presented for comparison. The achievable communication performance of the backscattering multilayer computational optimization is evaluated by numerical simulation below.

Combining: the figure shows the optimization scheme of the multilayer computing system of IRS backscatter support considered, where active beamforming at PB, passive beamforming on UE, active beamforming on AP, bandwidth and power allocation among all UEs, and locally computed time are jointly optimized.

Initiative: this figure represents a simplified optimization scheme of the multi-tier computing system for IRS backscatter support under consideration, where passive beamforming is randomly generated at the UE and other variables include active beamforming at the PB, active beamforming at the AP, bandwidth and power allocation among all UEs, and local computation time are jointly optimized.

And (3) passive: in multi-tier computing systems with IRS backscatter support in mind, the PB may not be a specially deployed BS. Instead, the existing surrounding signal stations can also be used as PB. The active beamforming at PB in this case is randomly generated. This figure represents the case where all variables are jointly optimized except for active beamforming at PB.

Random time: this diagram represents another simplified optimization scheme of the multi-tier computing system considered for IRS backscatter support, where the time allocation between the two phases is not optimized. Instead, the time is divided randomly, which is the only difference from the joint scheme.

Algorithm 1 summarizes the overall solution of the problem (P0) where t represents the tth iteration and ε represents a smaller positive accuracy limit or iteration accuracy. Algorithm 1 is convergent, and as can be clearly seen from algorithm 1, there are three repeat-end cycles. In the first repeat-end loop, the computational complexity depends mainly on the solution process of (P7). Because A is _m,k ,B _m,k ,α _m,k β _m,k The solution can be quickly realized without a complex optimization process. Furthermore the number of element units of the IRS far exceeds the number of antennas at the AP. By adopting the Interior Point Method (IPM), the complexity of the problem (P7) is given:

wherein, the first and the second end of the pipe are connected with each other,

the computational complexity is dominated in the nested repeat-end loop (P9). With IPM, the complexity of the problem (P9) is:

wherein the content of the first and second substances,

after problem solving (p 7) and (p 9), the solution is typically a matrix of rank 1, so the complexity is small and the recovery of a vector of rank 1 can be neglected. The overall complexity (p 0) of the problem is therefore approximately:

C _P0 ＝t _loop1 C _P7 +t _loop2 n ₃ C _P9

wherein, t _loop1 t _loop2 Representing the number of cycles of the first and third repeat-end cycles respectively,

representing the complexity of the bisection method in the second iteration.

In numerical simulations all channel data were randomly generated with a leis distribution of factor k, for all channels the path loss is expressed as PL = PL ₀ -25lg(d/d ₀ )dB，PL ₀ Indicates the reference distance d ₀ Path loss, d represents the transmission distance; the number of antenna groups in all groups on PB is the same, and the antenna gain of each group of antennas is represented by η. Only a small portion of the electromagnetic power emitted from the PB reaches each AP. Furthermore, the number of element units is the same for all IRS. Some important simulation parameter settings are listed in table 1. Distances from PB to UE, from UE to AP, from PB to AP and from AP to the central server are respectively in an interval [ d ] by intervals _pu0 -10m,d _pu0 +10m]、[d _ua0 -20m,d _ua0 +20m]、[d _pa0 -20m,d _pa0 +20m]And [ d _ac0 -20m,d _ac0 +20m]And (4) randomly and uniformly generating.

TABLE 1 simulation parameter values

Referring to fig. 3, the convergence behavior of all optimization schemes in random observation is shown. It is clear from fig. 3 (a) and 3 (b) that all optimization schemes converge very fast. In fig. 3 (a), the "joint" optimization scheme has the highest offloading rate from the UE to the AP. As can be seen from fig. 3 (b), the proposed optimization scheme does not achieve the best performance when data is offloaded from the AP to the central server.

Referring to fig. 4, the offload rate from the UE to the AP and the system's calculation and bit relationships to the number of elements of the IRS at the UE are described. From fig. 4 (a) and 4 (b), it can be seen that as the number of IRS elements increases, the offload rate and computation and bits of the system also increase. This result indicates that an increase in the number of IRS elements contributes to an increase in system performance. Furthermore the performance of the combined scheme is superior to the active and passive schemes. In fig. 4 (b), the system computation sum bit obtained by the random time scheme is smaller than the joint state, which shows that the time allocation between t1 and t2 plays a crucial role in improving the performance of the whole system.

Referring to fig. 5 and 6, it is shown how the total transmission power of PB and the average distance from UE to AP affect the offload rate from UE to AP and the computation and bit, respectively. It can be seen from the figure that an increase in the total transmit power at PB and a decrease in the average distance from the UE to the AP help to improve the rate and system computation and bits of offloading from the UE to the AP.

Referring to fig. 7, it is shown how the number of APs and their number of antennas affect the computation and the bits. It is readily apparent from fig. 7 (a) that the offload rate from UE to AP and the computation and bit of the system increase as more APs are deployed.

In summary, the 6G internet of things multilayer computing method and system based on IRS-BackCom enablement according to the present invention divide the computing process into the client, the access point, and the central server, and optimize each process to improve the computing efficiency and achieve the purpose of low power consumption; the maximization problem of the calculation sum in the time block is solved; active beamforming of the energy station, passive beamforming at the user end, active beamforming at the access point, bandwidth and power allocation between all user equipments, and time of local computation are optimized. Simulations have shown that increasing the number of IRS elements, the total transmit power, the number of access points and the number of antennas of an access point can facilitate the computation and the bits. Conversely, performance decreases as the average distance of the user device from the access point increases; the result shows that the IRS-BackCom multilayer computing system is efficient and feasible, can greatly reduce the communication pressure of communication equipment and provides a reliable, efficient and low-power-consumption communication network.

Claims (10)

1. The IRS-BackCom energized 6G Internet of things multilayer computing method is characterized by comprising the following steps:

s1, establishing an IRS-BackCom enabled 6G Internet of things multilayer computing system model;

s2, performing problem modeling on the IRS-Backcom enabled 6G Internet of things multilayer computing system model obtained in the step S1, simplifying the communication problem into a three-layer logic communication model, obtaining a user equipment layer, an access point layer and a central processor layer for computing task scheduling, and obtaining an overall communication model of a task allocation scheme of local computing and partial data unloading;

and S3, performing system problem expression and decomposition on the communication model obtained in the step S2 to obtain a three-layer communication model, distributing the problems of local calculation and migration calculation to a user equipment layer, processing the problems by an access point layer and a central processing unit layer, performing time scheduling on tasks, performing local calculation on the problems meeting the local calculation capability, performing partial data unloading calculation on the problems exceeding the local calculation capability, performing decomposition calculation on the rest problems, and realizing IRS-BackCom enabled 6G Internet of things multilayer calculation.

2. The IRS-BackCom enabled 6G Internet of things multilayer computing method according to claim 1, wherein in step S1, the multilayer computing system model comprises:

k intelligent reflector IRS assisted user equipment UE on one energy station PB and T1 layer, M access points AP connected with one MEC server on T2 layer and M access points AP on T3 layerA central server with rich resources; the energy station PB, each uniform access point AP and the central server are respectively provided with N _p ,N _a And N _c Root antenna, N _c ≥M×N _a In a hierarchical network, each user equipment UE requests to perform a calculation task, each task being bit-independent, being split into a number of bit subsets, all channels following quasi-static fading and the channel information being known.

3. The IRS-Backcom enabling 6G Internet of things multilayer computing method of claim 1, wherein in step S2, a partial data offloading strategy is adopted, all the first layer user equipment UE and the second layer access point AP perform computation offloading and local computation simultaneously, the computation tasks of the user equipment UE are divided, one part is computed locally in the user equipment UE, the other part is offloaded to the access point AP, and after the task transfer from the user equipment UE to the access point AP is completed, the computation task bits received by each access point AP are divided into: processed locally at the access point AP and migrated to the central server for computation.

4. The IRS-BackCom enabled 6G IOT multi-layer computing method of claim 3, wherein the first layer of local computing computes local energy consumption E at kth UE _loc,k Comprises the following steps:

wherein, t _loc,k Representing the execution time, t, of the local calculation at the kth user equipment, UE _off Indicating that the energy station PB and all the intelligent reflecting surfaces IRS are in an open state during unloading time, mu indicating that the power consumption of one element unit of the intelligent reflecting surfaces IRS is positively correlated with the phase resolution of the intelligent reflecting surfaces IRS, and L being the number of elements in the IRS;

energy consumption E of the mth access point AP _AP,m Comprises the following steps:

wherein, t _AP,m And t _mig,m Respectively representing the computation time and data migration time, p, at the mth AP _AP,m Represents the information transmission power of the mth AP, epsilon _AP,m The power consumption coefficient of the processor chip at the m APs is closely related to the chip shelf,

the CPU frequency of m APs.

5. The IRS-BackCom-enabled 6G Internet of things multilayer calculation method according to claim 3, wherein IRS communication is adopted by first-layer User Equipments (UEs), when an energy station (PB) radiates electromagnetic waves by using a directional antenna, energy-carrying radio frequency signals reaching an Intelligent Reflection Surface (IRS) of each UE are used for backscatter communication, when the IRS performs backscatter communication, incident signals serving as signal carriers are re-modulated, and a backscatter vector theta is generated _k Modulated and converted into a signal x _kr Passive beamforming vector theta of _kr 。

6. The IRS-BackCom enabled 6G Internet of things multilayer computing method of claim 5, wherein the remodulation specifically comprises:

where s denotes the original data signal,

x _kr representing the modulated data signal for the kth access point AP of the r-th UE,

w _k beamforming vectors on the kth antenna group of PB.

7. The IRS-BackCom enabled 6G IOT multi-layer computing method according to claim 1, wherein in step S3, the data is offloaded to the access point AP in a first phase, and the data offloaded from the access point AP to the central server is performed in a second phase; the optimization problem is expressed by jointly optimizing active beam forming at the energy station PB, passive beam forming at the user equipment UE, active beam forming at the access point AP, bandwidth and power allocation among all the user equipments UE, and computation and bit of the local computation time maximization system, which is specifically as follows:

C8:t _off ≤T ₁

C10:t _AP,m R _AP,m +t _mig,m R _mig,m ≥S _m

C12:T ₁ +T ₂ ＝T

wherein S is _m Represents the calculation sum bit, T, received by the mth AP ₁ And T ₂ Respectively representing the duration of the first and second phases,

are respectively represented by w _k ,θ _kr ,p _k ,t _loc,k And v _m C1 and C2 represent the active and passive beamforming constraints of PB and all UEs, respectively, C3, C4, C5 represent the power allocation or bandwidth allocation constraints between UEs, P represents the total power of PB, C5 aims to ensure fairness between UEs by setting minimum and bit maximum bandwidth size bounds, C6 is the energy constraint of each UE,

represents the energy threshold of all UEs, C7 and C8 are time constraints, C9 represents the beamforming constraint at the AP; c10 is that the number of data bits received by each AP is lower than the data processing capacity of local calculation and unloading; c11 is a time constraint;

calculating and maximizing the bit in the first stage to obtain a problem 1, and calculating and obtaining a problem 2 by a problem transformation and alternative parameter optimization method in the minimization of delay in the second stage; solving the problems 1 and 2 to obtain calculation, bit and delay minimization, decomposing the calculation problems of the voice video and the live video, and distributing tasks of time and calculation to obtain time scheduling of multilayer calculation, wherein the first layer obtains a distribution scheme of local calculation and bit and partial data unloading from the first layer to the second layer and the third layer;

problem 1 is as follows:

s.t.C1-C8

problem 2 is as follows:

s.t.C9-C11

8. the IRS-BackCom enabled 6G Internet of things multilayer computing method of claim 7, wherein the problem 1 equivalent is expressed as:

wherein, B _k Is the bandwidth of the channel, alpha _m,k Is a Lagrangian dualConversion of the auxiliary variable, Ω _m,k Introducing a variable, beta, to facilitate matrix computation _m,k To positively define an auxiliary variable, T _m,k For the joint channel gains of the k-th to m-th,

in order to introduce variables that facilitate the computation of the matrix,

is a set of APs, Θ _kr In order to realize the purpose,

is a set of UEs that are to be served,

is the set of IRS element units of the UE, and l is the l-th element unit.

9. The IRS-BackCom enabled 6G Internet of things multilayer computing method of claim 7, wherein the problem 2 equivalent expression is as follows:

wherein, (.) ^H Representing a conjugate transpose of a matrix or vector, X _mig,m And Y _mig,m,r Is a variable, R _lb Is composed of

Lower bound of (a) _mig,m To positively determine the auxiliary variable, V _m And

is a variable, and is a function of,

is the set of APs, B is the total bandwidth, S _m For sum of data bits, p _AP,m Is the transmit power of the mth AP.

10. An IRS-BackCom enabled 6G Internet of things multilayer computing system, comprising:

the system module is used for establishing a multi-layer computing system model of the IRS-BackCom enabled 6G Internet of things;

the problem module is used for carrying out problem modeling on the IRS-BackCom enabled 6G Internet of things multilayer computing system model obtained by the system module, simplifying the communication problem into a three-layer logic communication model, obtaining a user equipment layer, an access point layer and a central processor layer for computing task scheduling, and obtaining an overall communication model of a task allocation scheme of local computing and partial data unloading;

and the calculation module is used for performing system problem expression and decomposition on the communication model obtained by the problem module to obtain a three-layer communication model, distributing the local calculation and migration calculation problems to the user equipment layer, processing the problems by the access point layer and the central processing unit layer, performing time scheduling on tasks, performing local calculation on the problems meeting the local calculation capability, performing partial data unloading calculation on the problems exceeding the local calculation capability, performing decomposition calculation on the remaining problems, and realizing the 6G Internet of things multilayer calculation endowed with the IRS-BackCom.

Images