A star-ris assisted hybrid noma computing offloading method

Abstract

The application relates to the technical field of Internet of Things, and discloses a STAR-RIS assisted hybrid NOMA computing offloading method, which comprises the following steps: S1, before the start of the offloading process, a main base station collects and caches hyper-surface and auxiliary base station basic information in a network; S2, when a user needs to offload a task, the user first sends an offloading request to the main base station; S3, the main base station receives the user offloading requirement; S4, after the optimization of the offloading strategy is completed, the main base station can notify the user, the auxiliary base station and RIS to prepare for task offloading; and S5, after the hyper-surface and the auxiliary base station are configured, the main base station sends the strategy to the user. Compared with the traditional only-reflection RIS technology, the STAR-RIS technology can transmit the offloading task to both sides of the hyper-surface through reflection and refraction signals, the offloading range is expanded, and for the Internet of Things equipment, computing nodes can be selected for computing offloading in a wider communication range.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of Internet of Things, and particularly relates to a STAR-RIS assisted hybrid NOMA computing offloading method. BACKGROUND

[0002] In order to support new technologies such as extended reality, intelligent devices, holographic radio, and satellite networks, the future 6G network architecture proposes the requirement of ultra-low latency communication under the 6G network. However, the 6G network access device is expected to reach the trillion level, the communication data volume is large, and the computing capability and energy configuration of most devices cannot support large data volume and low latency processing. Existing research proposes edge computing offloading to solve the 6G ultra-low latency problem, that is, by offloading the data generated by the device to the edge node with rich computing resources for processing, the low-latency task data computing is realized. However, in the case of a large number of Internet of Things devices accessing, a large amount of task data processing causes a serious burden on the computing node, and further causes network congestion and increased latency.

[0003] The offloading methods used in the prior art mainly include traditional wireless transmission offloading, RIS-assisted enhanced wireless offloading, and NOMA-assisted wireless offloading, and the technical solutions and technical defects of these methods are as follows:

[0004] Traditional wireless transmission offloading: traditional wireless transmission usually transmits the offloading link through a wireless line-of-sight link, and the processing result is also transmitted back to the user end through a wireless link. The limitation is that the Internet of Things users can only select one offloading computing node to transmit the task at the same time, the offloading flexibility is low, and the offloading transmission performance is severely restricted by the actual channel condition. Especially when the channel is blocked by an obstacle, or there are many offloading transmissions in the network, the severe path loss and signal interference significantly reduce the transmission rate and increase the time consumption.

[0005] RIS-assisted enhanced wireless offloading: conventional RIS-assisted enhanced wireless offloading, that is, an additional transmission path is constructed by means of RIS assistance to improve the offloading rate and offloading range. The limitation is that the conventional RIS can only transmit the offloading task to one side of the reflecting surface, and cannot realize multi-directional offloading, and the utilization rate of the computing resources in the network is limited. Moreover, the RIS-constructed auxiliary channel complicates the network radio environment, and the communication interference is intensified, which reduces the transmission rate in the case of multiple users and affects the network offloading processing performance.

[0006] NOMA assisted wireless offloading: NOMA assisted wireless communication, mainly through NOMA technology multiplexing time-frequency resources, realizing simultaneous multi-directional transmission, for single Internet of Things users, can realize simultaneous offloading to multiple computing nodes, and improve task processing performance by using multiple node resources. However, due to the limitations of the decoding order and other transmission rules of NOMA, compared with the conventional TDMA transmission mode, NOMA is not always the best strategy under different offloading scenarios and offloading conditions. And NOMA assisted offloading is still restricted by time channel conditions.

[0007] In summary, the present patent proposes a STAR-RIS assisted hybrid NOMA offloading scheme, which can realize simultaneous multi-directional transmission, expand the offloading range, and meet the low-latency computing offloading needs under various offloading conditions. SUMMARY

[0008] (I) Technical problems to be solved

[0009] In view of the shortcomings of the prior art, the present application provides a STAR-RIS assisted hybrid NOMA computing offloading method, which has the advantages of realizing flexible allocation of computing offloading resources by applying RIS technology and multiple access communication technology, further reducing the computing offloading delay, and meeting the low-latency communication needs of massive device access, solving the problems raised in the above background technology.

[0010] (II) Technical solutions

[0011] The present application provides the following technical solutions: a STAR-RIS assisted hybrid NOMA computing offloading method, the offloading method comprising the following steps:

[0012] S1, before the start of the offloading process, the main base station collects and caches the super surface and auxiliary base station basic information in the network;

[0013] S2, when the user needs to offload tasks, first send an offloading request to the main base station;

[0014] S3, after the main base station receives the user offloading requirement, first judges whether the user task meets the offloading processing requirement, and judges whether to support processing the offloading task type of the user according to the service capability of the main base station itself and the service capability of the cached auxiliary base station;

[0015] S4, after the offloading strategy optimization is completed, the main base station can notify the User, the auxiliary base station, and the RIS to prepare for task offloading. The main base station needs to first notify the super surface and the auxiliary base station to prepare for offloading. After receiving the information, the auxiliary base station enters the offloading receiving state until receiving the offloading completion message, and receives the User offloading task during this period;

[0016] S5, after the super surface and the auxiliary base station configuration is completed, the main base station sends the strategy to the user, and makes itself enter the data receiving state, after the user receives the strategy, starts to unload the task according to the optimal unloading strategy.

[0017] Preferably, the main base station collects the super surface and the auxiliary base station basic information in the network, which can be synchronized when the super surface and the auxiliary base station access the network, and the basic information is used for channel estimation, trust degree calculation and other operations.

[0018] Preferably, the main base station receives the User unloading request: including channel state information, User unloading task quantity and size base station calculation capacity.

[0019] Preferably, the main base station optimizes the unloading strategy based on the scheme and transmission algorithm, which specifically includes time allocation, unloading order, STAR-RIS phase modulation, unloading task allocation, power allocation, decoding order and the like, and minimizes the total time consumption.

[0020] Preferably, based on the optimized unloading strategy, the main base station sends the phase configuration to the RIS, and sends the unloading request to the auxiliary base station, at this time the RIS adjusts the element phase, the auxiliary base station enters the task receiving state, and then sends the confirmation information to the main base station.

[0021] Preferably, based on the hybrid NOMA unloading strategy, when the unloading scheme changes, the NOMA transmission is converted into TDMA transmission, or the TDMA transmission is converted into NOMA transmission, the super surface needs to be reconfigured, at this time the User suspends the unloading, and sends the unloading scheme change information to the main base station.

[0022] Preferably, after the User unloading is completed, the unloading completion information is sent to the main base station, the main base station receives the confirmation information, sends the unloading completion information to the auxiliary base station, closes the task receiving, starts the task calculation process, and the auxiliary base station receives the unloading completion confirmation, also closes the task receiving, and enters the task calculation process.

[0023] Compared with the prior art, the application provides a STAR-RIS assisted hybrid NOMA computing unloading method, which has the following beneficial effects:

[0024] 1. The STAR-RIS assisted hybrid NOMA computing offloading method introduces STAR-RIS technology to assist the computing offloading of Internet of Things devices, and further designs a hybrid NOMA offloading scheme to reduce offloading delay. This scheme fully considers using multiple offloading to fully utilize network resources for offloading computing, and simultaneously offloads Internet of Things devices to multiple directions through STAR-RIS technology, and combines NOMA technology to realize time-frequency resource multiplexing in simultaneous multi-direction offloading, further improving offloading performance. Compared with the traditional reflection RIS assisted offloading or single NOMA / TDMA offloading scheme, this scheme has higher flexibility and wider offloading communication range.

[0025] 2. The STAR-RIS assisted hybrid NOMA computing offloading method, the simultaneous refraction / reflection characteristics of STAR-RIS can simultaneously assist and enhance multiple links in the simultaneous multi-direction offloading scenario of Internet of Things devices, and can dynamically adjust the transmission gain of the super surface phase change to multiple offloading paths according to specific conditions. Compared with the traditional RIS, it has higher flexibility.

[0026] 3. The STAR-RIS assisted hybrid NOMA computing offloading method, the traditional NOMA or TDMA scheme has advantages and disadvantages in different scenarios, and cannot be used as the optimal transmission strategy in all offloading scenarios. The hybrid NOMA scheme considers the NOMA scheme and the TDMA scheme, and dynamically switches between NOMA and TDMA in the offloading process according to the specific offloading scenario, to realize lower latency offloading. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 The STAR-RIS and hybrid NOMA assisted computing offloading scheme of the present application is shown in the figure;

[0028] Figure 2 The flow scheme of the present application is shown in the figure;

[0029] Figure 3 The User first offloads to the BS schematic diagram of the present application;

[0030] Figure 4 The User first offloads to the SBS schematic diagram of the present application. DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0032] Referring to Figures 1-4 A STAR-RIS assisted hybrid NOMA computing offloading method, the offloading method comprising the following steps:

[0033] S1, before the start of the offloading process, the main base station collects and caches the super surface and auxiliary base station basic information in the network, when the RIS and auxiliary base station access the network, the super surface and auxiliary base station need to be authenticated, wherein the RIS and auxiliary base station need to send their respective basic information to the main base station, including RIS information such as super surface type, location, and element number, and auxiliary base station information such as auxiliary base station type, location, and computing resources. The main base station will cache the collected data and perform authentication, trust calculation and other operations on the super surface and auxiliary base station, and the processing results will also be cached at the main base station end;

[0034] S2, when the user needs to offload tasks, first send an offload request to the main base station, wherein the user offload request includes offload task information, specifically including offload task data volume, offload task type, task processing delay requirement, and Internet of Things user self information, specifically including transmission power size, configured antenna number, user location information, etc.

[0035] S3, the main base station receives the user offload request, first judges whether the user task meets the offload processing requirement, according to the main base station's own service ability and the cached auxiliary base station service ability, judges whether it supports processing the offload task type of the user, if there is no resource in the network to process the type of user task, the user is told to handle the task by himself, otherwise continue to process the offload task; If the user task meets the offload computing requirement, appropriate computing resources can be allocated to the user task to process the task. The main base station decides to use the auxiliary base station. For the auxiliary base station existing in the network, the main base station selects the optimal auxiliary base station based on the location information, computing ability, trust degree and other comprehensive considerations of the auxiliary base station. The trust degree is calculated by the historical task processing probability of the auxiliary base station, which is the probability of completing a certain type of task within a predetermined time by the auxiliary base station. The main base station calculates the direct trust value of the auxiliary base station through historical records. At the same time, the main base station can also collect the trust value of the current auxiliary base station from other base stations in the network, that is, the indirect trust value. The sum of the direct and indirect trust values is calculated to obtain the trust degree of the main base station to the auxiliary base station. The trust degree of the auxiliary base station can be used for auxiliary base station selection strategy and setting of the proportion of offloading tasks to the auxiliary base station;

[0036] S4, After the optimization of the offloading strategy is completed, the main base station can notify the user and the auxiliary base station that the RIS is ready to perform task offloading. The main base station needs to first notify the metasurface and the auxiliary base station to be ready for offloading. After receiving the information, the auxiliary base station enters an offloading receiving state until it receives an offloading completion message, during which it maintains the reception of the user offloading task; the RIS needs to adjust the phase according to the optimized offloading. The RIS can be dynamically controlled by the main base station, that is, the RIS is configured with a processing unit that can receive phase configuration information sent by the main base station, dynamically adjust the phase of the metasurface elements, and send a configuration completion message to the main base station after the adjustment is completed.

[0037] S5, After the metasurface and the auxiliary base station are configured, the main base station sends the strategy to the user and enters a data receiving state. After receiving the strategy, the user starts to offload tasks according to the optimal offloading strategy. If offloading strategy switching occurs during the offloading process, the user first sends switching information to the main base station. After receiving the switching information, the main base station sends phase configuration information to the metasurface according to the offloading strategy, changes the metasurface configuration, and then notifies the user to continue the offloading operation after the configuration is completed.

[0038] If the processing unit of the metasurface configuration has a cache function, the main base station can send the phase configuration information of each offloading stage to the RIS and store it locally after optimizing the offloading strategy. When the user switches the offloading scheme, the RIS can automatically reconfigure the phase and notify the user to continue offloading after the reconfiguration is completed.

[0039] The specific offloading scheme process is as follows:

[0040] (1) The main base station collects the basic information of the metasurface and the auxiliary base station in the network. This operation can be performed synchronously when the metasurface and the auxiliary base station are connected to the network. Based on the basic information, channel estimation, trust degree calculation, and other operations are performed

[0041] (2) The main base station receives the user offloading request: including channel state information, user offloading task quantity, and large / small base station computing capability.

[0042] (3) The main base station optimizes the offloading strategy based on the proposed scheme and transmission algorithm. Specifically, it includes time allocation, offloading sequence, STAR-RIS phase modulation, offloading task allocation, power allocation, decoding order, and minimization of total time consumption.

[0043] (4) Based on the optimized offloading strategy, the main base station sends phase configuration to the RIS and sends the offloading request to the auxiliary base station. At this time, the RIS adjusts the element phase, and the auxiliary base station enters a task receiving state. Then, the confirmation information is sent to the main base station.

[0044] (5) The main base station sends the offloading strategy related information to the User, and after receiving the user's confirmation, the main base station enters the receiving task state and sends the start offloading information to the User.

[0045] (6) After receiving the main base station information, the User offloads the task according to the algorithm optimization result, and the metasurface assists to offload the task to the main base station and the auxiliary base station.

[0046] (7) Based on the mixed NOMA offloading strategy, when the offloading scheme changes (i.e., the NOMA transmission is converted to TDMA transmission, or the TDMA transmission is converted to NOMA transmission), the metasurface needs to be reconfigured. At this time, the User suspends the offloading and sends the offloading scheme change information to the main base station.

[0047] (8) The main base station receives the offloading scheme change information and adjusts the RIS phase based on the optimized offloading strategy. After the phase configuration is completed, the User is sent the continue offloading information. After receiving the information, the User continues the task offloading.

[0048] (9) After the User offloads, the User sends the offloading completion information to the main base station. After receiving the confirmation information, the main base station sends the offloading completion information to the auxiliary base station, and at the same time, the task receiving is closed and the task calculation process is started. After receiving the offloading completion confirmation, the auxiliary base station also closes the task receiving and enters the task calculation process.

[0049] For different offloading sequences, there are generally two offloading schemes π∈{π1,π2}, i.e., the User offloads to the BS first and then to the SBS, like Figure 3 , or the User offloads to the SBS first and then to the BS, like Figure 4 . Among them represents the time of User transmission to BS and SBS alone, and in this stage, the STAR-RIS is optimized to assist in enhancing one of the channels alone; represents the time of User offloading to two base stations simultaneously based on the NOMA mode, and in this stage, the STAR-RIS assists in reflecting and refracting two channels simultaneously; represents the time of BS and SBS computing the offloading task. Since the main time consumption is mainly in the task offloading stage and the task processing stage, the target function ignores the related information acquisition delay, the offloading strategy information and RIS control information transmission delay, and the task processing result feedback delay.

[0050] The problem model is shown below, which specifically considers the optimization of power allocation, reflection surface optimization, offloading strategy selection, time allocation, decoding order, and task division.

[0051]

[0052]

[0053]

[0054]

[0055]

[0056]

[0057] N s = pN,N b = (1 - p)N,

[0058]

[0059] 0≤|u b,m | 2 ,|u s,m | 2 ,|u b,m | 2 +|u s,m | 2 ≤1,

[0060] p b +p s ≤p,p b ≥0,p s ≥0,

[0061]

[0062]

[0063]

[0064] For specific problem solving can be divided into the following steps:

[0065] Fixed selection of a unloading order π and decoding order o

[0066] Time allocation And task division ratio p classification discussion, get different conditions under the time allocation and the optimal solution of the task ratio

[0067] Based on SCA iterative optimization power allocation p k And the phase of the reflecting surface Get the optimal solution

[0068] Comparison of different classification under the optimization of unloading delay, select the minimum delay as the optimal unloading scheme.

[0069] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A STAR-RIS assisted hybrid NOMA computing offloading method, characterized in that: The offloading method comprises the following steps: S1, before the offloading process starts, the main base station collects and caches the metasurface and auxiliary base station basic information in the network; S2, when the user needs to offload the task, first send an offloading request to the main base station; S3, the main base station receives the user offloading requirement, first judges whether the user task meets the offloading processing requirement, judges whether to support processing the offloading task type of the user according to the service capability of the main base station itself and the cached service capability of the auxiliary base station; S4, after the offloading strategy optimization is completed, the main base station can notify the User, the auxiliary base station, and the RIS to prepare for task offloading. The main base station needs to first notify the metasurface and the auxiliary base station to make offloading preparation. After receiving the information, the auxiliary base station enters the offloading receiving state until receiving the offloading completion message. During this period, the User offloading task is received; S5, after the metasurface and the auxiliary base station are configured, the main base station sends the strategy to the user and enters the data receiving state. After receiving the strategy, the user starts to offload the task according to the optimal offloading strategy; The main base station collects the metasurface and auxiliary base station basic information in the network. The operation can be performed synchronously when the metasurface and auxiliary base station are connected to the network, and channel estimation, trust degree calculation and other operations are performed based on the basic information; The main base station receives the User offloading request: including channel state information, User offloading task quantity and size base station calculation capability; The main base station optimizes the offloading strategy based on the proposed scheme and transmission algorithm, which specifically includes time allocation, offloading order, STAR-RIS phase modulation, offloading task allocation, power allocation, decoding order, etc., to minimize the total time consumption.

2. The STAR-RIS assisted hybrid NOMA computation offloading method according to claim 1, wherein: Based on the optimized offloading strategy, the main base station sends the phase configuration to the RIS and sends the offloading request to the auxiliary base station. At this time, the RIS adjusts the element phase, the auxiliary base station enters the task receiving state, and then sends the confirmation information to the main base station.

3. The STAR-RIS assisted hybrid NOMA computation offloading method according to claim 1, wherein: Based on the hybrid NOMA offloading strategy, when the offloading scheme changes, the NOMA transmission is converted into TDMA transmission, or the TDMA transmission is converted into NOMA transmission, the metasurface needs to be reconfigured. At this time, the User suspends the offloading and sends the offloading scheme change information to the main base station.

4. The STAR-RIS assisted hybrid NOMA computation offloading method of claim 1, wherein: After the User offloading is completed, the offloading completion information is sent to the main base station. After receiving the confirmation information, the main base station sends the offloading completion information to the auxiliary base station, closes the task receiving, starts the task calculation process, and the auxiliary base station receives the offloading completion confirmation, also closes the task receiving and enters the task calculation process.

Images