A task offloading method based on NOMA heterogeneous cellular network

Abstract

The application claims a task offloading method based on a NOMA heterogeneous cellular network, and belongs to the technical field of communication. In a sparse network scenario where a macro base station is not equipped with a server, a micro base station is equipped with a server, and effective coverage areas do not overlap, uneven distribution of mobile terminals can cause load imbalance of the micro base station server, thereby causing offloading performance to decrease and user service experience to deteriorate. A task offloading method based on a NOMA heterogeneous cellular network is provided. The method adopts a hybrid offloading mode combining local micro base station offloading, macro-micro base station cooperative offloading, and micro base station inter-cooperative offloading. Through corresponding task scheduling, power control, and computing resource allocation strategies, the resource utilization rate of the micro base station server is improved, the task offloading delay is reduced, and thus the user service quality experience is effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of communication network technology. Specifically, it relates to a task offloading method based on NOMA heterogeneous cellular networks. Background Technology

[0002] With the rapid growth in the number of mobile communication users, traditional cellular network spectrum resources are insufficient to meet the access needs of a large number of terminal devices. By introducing heterogeneous cellular networks and non-orthogonal multiple-access (NOMA) technology, the spectrum resource utilization of wireless networks can be effectively improved. NOMA heterogeneous cellular networks consist of macro base stations and multiple micro base stations. Among them, micro base stations reuse sub-channel resources through NOMA technology to improve spectrum efficiency.

[0003] NOMA allows multiple users to share the same subchannel, achieving signal differentiation in the power domain by allocating different power levels to users with different link qualities on the same channel. Compared to traditional Orthogonal Multiple Access (OMA), NOMA offers higher spectrum resource efficiency. However, a large number of users sharing a subchannel can cause severe co-channel interference and significant transmission delays. Therefore, power interference management plays a crucial role in NOMA heterogeneous cellular network applications. Thus, in NOMA heterogeneous cellular networks, reasonable task offloading and resource allocation strategies are needed to reduce task transmission and computation delays during edge task offloading.

[0004] To address the task offloading problem in NOMA heterogeneous cellular networks, existing research primarily employs the approach of configuring servers separately on macro and micro base stations, utilizing a vertical collaboration model between macro and micro base stations to reduce transmission and computation latency during task offloading. A few studies utilize a network model where servers are configured only on micro base stations with densely overlapping coverage areas, optimizing offloading latency while reducing computational resource allocation. However, in sparse network scenarios where macro base stations lack servers and micro base station coverage areas do not overlap, the uneven distribution of mobile terminals can lead to load imbalance on micro base station servers, resulting in degraded offloading performance and a worsened user experience. Therefore, optimizing offloading task scheduling and computational resource allocation among micro base stations is crucial for reducing network resource costs and improving user experience.

[0005] This invention proposes a task offloading method based on NOMA heterogeneous cellular networks. This method employs a hybrid offloading mode combining local micro base station offloading, macro-micro base station cooperative offloading, and inter-micro base station cooperative offloading. Through corresponding task scheduling, power control, and computing resource allocation strategies, it improves the resource utilization of micro base station servers, reduces task offloading latency, and thus effectively improves the user's quality of service experience. Summary of the Invention

[0006] This invention aims to solve the problems of the prior art mentioned above. It proposes a task offloading method based on NOMA heterogeneous cellular networks. The technical solution of this invention is as follows:

[0007] A task offloading method based on NOMA heterogeneous cellular networks, for sparse network scenarios where macro base stations are not equipped with servers, micro base stations are equipped with servers, and their effective coverage areas do not overlap, includes the following steps:

[0008] 101. Let I be the set of all users and S be the set of micro base stations. Place the users within the effective coverage area of ​​each micro base station s in S into set I. s Initialize the transmit power of each user i in I.

[0009] 102. For each user i in set I, if Uninstall mode m i =1, otherwise, let the uninstallation mode m i =2, put user i into temporary set I c middle;

[0010] 103. For each set I s For each user i, the number of task units n to be uninstalled based on the user's request. i The transmission rate from the user to the micro base station s m i Calculate the computing resource requirements. and I s Users in The values ​​are sorted in ascending order, according to... and the maximum available computing resources f of the micro base station s s Find the computing resources f that micro base station s can allocate to user i. i s ;

[0011] 104. For each set I s Sub-algorithm 1 is called respectively to optimize I. s Transmit power for each user and update and f i s ,if Let m i =3, change I s Users are added to set I in sequence. c If the condition is met, proceed to step 105; otherwise, proceed to step 105.

[0012] 105. According to f i s Calculate the number of task units that user i can offload to micro base station s. According to f i s , Update f s ,make if Skip to step 106; otherwise, skip to step 109.

[0013] 106. Take f from set S s Micro base stations s with a value greater than 0 are moved into temporary set S. c ,if Set I c All users in the middle according to n i If sorted in ascending order, proceed to step 107; otherwise, proceed to step 110.

[0014] 107. Set S c Chinese f s The largest micro base station s, as the current cooperating micro base station, will set I c The first user i in the set of users I of the cooperative micro base station s s And according to m i , and f s Calculate f i s and Let f s =f s -f i s , if Skip to step 107; otherwise, skip to step 108.

[0015] 108. Set S c Chinese f s Micro base stations s with a value of 0 are from set S c Move into S, if Proceed to step 109; otherwise, all users in set I shall perform task uninstallation and proceed to step 110.

[0016] 109. For each user set I s ,make Update the computing resources f allocated to each user i in the micro base station s. i s All users in set I execute the task to uninstall;

[0017] 110. End.

[0018] Furthermore, in step 103, the resource requirements are calculated. The calculation method is shown in formula (1):

[0019]

[0020] in, This represents the number of task units scheduled from user i to micro base station s, where Δd represents the minimum task unit, and c i Indicates computational complexity. r represents the maximum tolerable delay for user i. i s Let r represent the rate at which user i transmits tasks to micro base station s. The calculation method is shown in formula (2). i o Let represent the rate at which user i transmits tasks to macro base station o. The calculation method is shown in formula (3). The rate at which macro base station o transmits tasks to micro base station s′ is represented by the formula (4). The rate at which micro base station s transmits tasks to macro base station o is represented by the formula (5).

[0021] The rate at which user i transmits tasks to micro base station s The calculation method is shown in formula (2):

[0022]

[0023] Where B represents the communication bandwidth between user i and micro base station s, p i This represents the transmit power of user i. σ represents the channel gain between user i and micro base station s. 2 Let i' represent the power of additive white Gaussian noise, and let i' represent the user set I. s Any user other than user i, If it is a binary variable, Then user i′ is the interfering user of user i, let Otherwise, let This represents the sum of interference power from other users to user i.

[0024] The rate at which user i transmits tasks to macro base station o is r i o The calculation method is shown in formula (3):

[0025]

[0026] Where b represents the orthogonal sub-channel bandwidth allocated by the macro base station. This represents the channel gain between user i and macro base station o.

[0027] Rate of transmission of tasks from macro base station o to micro base station s′ The calculation method is shown in formula (4):

[0028]

[0029] Where, p o This indicates the transmit power of macro base station o. This represents the channel gain between the cooperating micro base station s′ and the macro base station o;

[0030] The rate at which micro base station S transmits tasks to macro base station O The calculation method is shown in formula (5):

[0031]

[0032] Where, p s This represents the transmit power of the micro base station s. This represents the channel gain between the micro base station s and the macro base station o.

[0033] Furthermore, the number of task units scheduled by user i to micro base station s The calculation method is shown in formulas (6)-(8). If the unloading mode m i =1, the number of task units scheduled from user i to micro base station s The calculation method is shown in formula (6):

[0034]

[0035] Among them, f i s This represents the computing resources that micro base station s can allocate to user i.

[0036] If uninstallation mode m i =2, the number of task units scheduled from user i to cooperative micro base station s′. The calculation method is shown in formula (7):

[0037]

[0038] If uninstallation mode m i =3, the number of task units scheduled from user i to cooperative micro base station s′. The calculation method is shown in formula (8):

[0039]

[0040] Furthermore, in step 103, the micro base station s can allocate computing resources f to user i. i s The calculation method is shown in formula (9):

[0041]

[0042] Where, f s The currently available resources for micro base station s. Let be a binary variable. If micro base station s allocates computing resources to user i′, let Otherwise, let

[0043] Furthermore, in step 105, the micro base station s currently has available resources f. s The calculation method is shown in formula (10):

[0044]

[0045] Furthermore, sub-algorithm 1 in step 104 specifically includes:

[0046] The user communicates with the local micro base station using NOMA technology. A user power optimization model is constructed using the particle swarm optimization algorithm, specifically including:

[0047] Step 1): Let the number of population iterations be K, the number of particles in each iteration be U (i.e., the number of feasible solutions), the power optimization random factors γ1∈[0,1], γ2∈[0,1], the power influence factors θ1, θ2, and the maximum power adjustment factor δ. max Minimum power adjustment factor δ min ;

[0048] Step 2): Let set I s Power of each user i Power change Initialize U feasible solutions in Current optimal solution Global optimal solution Allocable computing resources for user i Among them, f i s Given the available computing resources for user i on micro base station s, the fitness function is... Let the iteration count variable k = 0, and the feasible solution count variable u = 0;

[0049] Step 3): Let k = k + 1. If k ≤ K, jump to step 4); otherwise, jump to step 11.

[0050] Step 4): Let u = u + 1. If u ≤ U, jump to step 5); otherwise, let u = 0 and jump to step 3.

[0051] Step 5): Update set I s The power change for each user i power Power adjustment factor δ(k);

[0052] Step 6): For set I s For each user i in the list, if the power of user i is... make If yes, proceed to step 7); otherwise, proceed to step 7.

[0053] Step 7): For set I s For each user i in the list, if the power of user i is... make If yes, proceed to step 8); otherwise, proceed to step 8.

[0054] Step 8): Let set I s Power of each user i According to m i , and renew Set I s User i in The values ​​are sorted in ascending order, according to... and f s Update f i s ,make renew

[0055] Step 9): If make If the previous step is not successful, proceed to step 10; otherwise, proceed to step 10.

[0056] Step 10): If make If the previous step is incorrect, proceed to step 4; otherwise, proceed to step 4.

[0057] Step 11): For set I s For each user i, let p i equal to φ g The i-th dimension component φ g,i According to m i , and r i s ,calculate Set I s User i in The values ​​are sorted in ascending order, according to... and f s Calculate f i s ;

[0058] Step 12): The algorithm ends.

[0059] Furthermore, in step 5), the set I is updated. s The power change for each user i power The method for calculating the power adjustment factor δ(k) is shown in formulas (11), (12), and (13):

[0060]

[0061]

[0062]

[0063] Where, δ max Indicates the maximum power regulation value, δ min Let K represent the minimum power adjustment value, K represent the number of population iterations, θ1 and θ2 represent power influence factors, γ1 and γ2 represent power optimization stochastic factors, γ1∈[0,1], γ2∈[0,1], and φ u,i Indicates the current optimal solution The optimal power of user i, φ g,i Represents the global optimal solution The optimal power for user i.

[0064] The advantages and beneficial effects of this invention are as follows:

[0065] This invention proposes a task offloading method based on NOMA heterogeneous cellular networks. Currently, traditional task offloading methods for heterogeneous cellular networks typically use OFDMA technology to allocate channel resources and do not consider cooperative offloading between micro base stations, resulting in low user task transmission rates and low resource utilization of micro base station servers. This invention utilizes the particle swarm optimization algorithm to construct a user power optimization model and employs a hybrid offloading mode combining local micro base station offloading, macro-micro base station cooperative offloading, and inter-micro base station cooperative offloading to complete task offloading. Through corresponding task scheduling, power control, and computational resource allocation strategies, it improves the resource utilization of micro base station servers, reduces task offloading latency, and thus effectively improves the user's quality of service experience. Attached Figure Description

[0066] Figure 1 This is a flowchart of a preferred embodiment of the present invention for a task offloading method in a NOMA heterogeneous cellular network. Detailed Implementation

[0067] The technical solutions of the present invention will be clearly and thoroughly described below with reference to the accompanying drawings of the embodiments. The described embodiments are merely some embodiments of the present invention.

[0068] The technical solution of the present invention to solve the above-mentioned technical problems is:

[0069] The concepts and models involved in this invention are as follows.

[0070] 1. Network Model

[0071] In sparse network scenarios where macro base stations are not configured with servers and micro base station coverage areas do not overlap, the macro base station acts as a central controller, collecting information from all micro base stations and users within its coverage area. It uses OFDMA technology to communicate with micro base stations and users, while micro base stations use NOMA technology to communicate with users within their effective coverage area. The macro base station and each micro base station operate on different frequency bands. To address the issue that uneven distribution of mobile terminals may lead to unbalanced micro base station server load, resulting in decreased offloading performance and degraded user service experience, a task offloading method based on NOMA heterogeneous cellular networks is proposed. This method employs a hybrid offloading mode combining local micro base station offloading, macro-micro base station collaborative offloading, and inter-micro base station collaborative offloading. Through appropriate task scheduling, power control, and computational resource allocation strategies, it improves the resource utilization of micro base station servers, reduces task offloading latency, and thus effectively improves the user's quality of service experience.

[0072] 2. Other symbols involved in this invention are explained below.

[0073] Δd: Task unit size

[0074] N i The number of task units generated by user i

[0075] n i The number of task units to be scheduled for user i

[0076] User i's task delay constraint

[0077] r i s : Transmission rate between user i and micro base station s

[0078] B: Communication bandwidth between user i and micro base station s

[0079] Number of task units scheduled by user i to micro base station s

[0080] f i s The computing resources that micro base station s can allocate to user i

[0081] σ 2 Additive white Gaussian noise power

[0082] i: User

[0083] s: micro base station

[0084] o: Macro base station

[0085] I: The set of users in the system

[0086] I s : User set of micro base station s

[0087] S: Micro base station set

[0088] p i Transmit power of user i

[0089] Determine whether user i′ under micro base station s is an interfering user.

[0090] Channel gain between user i and micro base station s

[0091] The rate at which micro base station S transmits tasks to macro base station O

[0092] Rate of transmission of tasks from macro base station o to cooperative micro base station s′

[0093] The binary variable used to determine whether user i has unloaded the task to the collaborative micro base station s′

[0094] b: Orthogonal subchannel bandwidth allocated to macro base station o

[0095] p s The transmit power of micro base station s

[0096] Channel gain between micro base station s and macro base station o

[0097] p o Transmission power of macro base station o

[0098] The rate at which user i transmits tasks to macro base station o

[0099] Channel gain between user i and macro base station o

[0100] binary variable of the connection state between user i and micro base station s

[0101] m i User i's task uninstallation mode

[0102] A binary variable indicating whether micro base station s allocates computing resources to user i.

[0103] f s Available computing resources of micro base station s

[0104] δ(k): Power adjustment factor in the k-th iteration

[0105] The power change of user i corresponding to the u-th feasible solution in the k-th iteration The power of user i corresponding to the u-th feasible solution in the k-th iteration

[0106] The u-th feasible solution in the k-th iteration

[0107] θ1, θ2: Power influence factors

[0108] δ max Maximum power adjustment value

[0109] δ min Minimum power adjustment value

[0110] γ1, γ2: Random factors controlling the power

[0111] φ u,i The optimal power of user i corresponding to the current optimal solution of the u-th feasible solution.

[0112] φ g,i The optimal power for user i corresponding to the global optimal solution.

[0113] φ u The current optimal solution for the u-th feasible solution.

[0114] φ g Global optimal solution

[0115] The technical solution of the present invention is described below.

[0116] 1. Location relationship between users and micro base stations

[0117] Let the distance between user i and micro base station s be . As shown in formula (1):

[0118]

[0119] 2. Task unloading delay

[0120] If user i is within the coverage area of ​​micro base station s and the offloading mode is user-micro base station offloading, let the delay from task offloading to micro base station s be . As shown in formula (2):

[0121]

[0122] Where, r i sf represents the rate at which user i transmits tasks to micro base station s. i s This represents the computing resources that micro base station s can allocate to user i. r represents the number of task units that user i is scheduled to micro base station s. i s As shown in formula (3):

[0123]

[0124] Where B represents the communication bandwidth between user i and micro base station s, p i This represents the transmit power of user i. σ represents the channel gain between user i and micro base station s. 2 Let i' represent the power of additive white Gaussian noise, and let i' represent the user set I. s Any user other than user i, If it is a binary variable, Then user i′ is the interfering user of user i, let Otherwise, let This represents the sum of interference power from other users to user i.

[0125] If user i is within the effective coverage area of ​​micro base station s and the offloading mode is user-micro base station-macro base station-cooperative micro base station offloading, let its task offloading delay be... As shown in formula (4):

[0126]

[0127] in, This represents the rate at which micro base station s transmits tasks to macro base station o. f represents the rate at which macro base station o transmits tasks to cooperative micro base station s′. i s′ This represents the computing resources allocated by the collaborative micro base station server s′ to user i. This represents a binary variable. If user i unloads the task to the cooperative micro base station s′, let... Otherwise, let and in, As shown in formulas (5) and (6) respectively:

[0128]

[0129]

[0130] Where b represents the orthogonal subchannel bandwidth allocated by the macro base station, p s This represents the transmit power of the micro base station s. p represents the channel gain between micro base station s and macro base station o. o This indicates the transmit power of macro base station o. This represents the channel gain between the cooperating micro base station s′ and the macro base station o.

[0131] If user i is not within the effective coverage area of ​​the micro base station and the offloading mode is user-macro base station-cooperative micro base station offloading, let its task offloading delay be... As shown in formula (7):

[0132]

[0133] Where, r i o The rate at which user i transmits tasks to macro base station o is represented by formula (8):

[0134]

[0135] in, This represents the channel gain between user i and macro base station o.

[0136] set up Let t be a binary variable representing the connection state between user i and micro base station s, and let t be the task offloading delay for user i. i As shown in formula (9):

[0137]

[0138] Among them, if make Otherwise, let Since the effective coverage areas of micro base stations do not overlap, user i can only be within the effective coverage area of ​​one micro base station, i.e.

[0139] Let the normalized latency of user i's task be t. i As shown in formula (10):

[0140]

[0141] Let ω represent the task completion status of user i. i As shown in formula (11):

[0142]

[0143] Wherein, if the task generated by user i is completed within the maximum tolerable delay, ω i =0, otherwise, ω i =1.

[0144] 3. Computing resource requirements

[0145] Let user i's computing resource requirements be... As shown in formula (12):

[0146]

[0147] 4. The computing resources f that micro base station s can allocate to user i i s The calculation method is shown in formula (13):

[0148]

[0149] Among them, f s The currently available resources for micro base station s. Let be a binary variable. If micro base station s allocates computing resources to user i′, let Otherwise, let

[0150] 5. Currently available resources f of micro base station s s The calculation method is shown in formula (14):

[0151]

[0152] 6. Number of task units that user i schedules tasks to micro base station s The calculation method is shown in formulas (15)-(17):

[0153] If uninstallation mode m i =1, the number of task units scheduled from user i to micro base station s The calculation method is shown in formula (15):

[0154]

[0155] Among them, f i s This represents the computing resources that micro base station s can allocate to user i.

[0156] If uninstallation mode m i =2, the number of task units scheduled from user i to cooperative micro base station s′. The calculation method is shown in formula (16):

[0157]

[0158] If uninstallation mode m i =3, the number of task units scheduled from user i to cooperative micro base station s′. The calculation method is shown in formula (17):

[0159]

[0160] 7. The number of task units n to be scheduled for user i i The calculation method is shown in formula (18):

[0161]

[0162] 8. The power change of user i corresponding to the feasible solution u in the k-th iteration. Power corresponding to user i The power adjustment factor δ(k) in the kth iteration is calculated as shown in formulas (19)-(21):

[0163]

[0164]

[0165]

[0166] 9. Sub-algorithm 1: Power control algorithm based on particle swarm optimization

[0167] The user communicates with the local micro base station using NOMA technology. A user power optimization model is constructed using the idea of ​​particle swarm optimization algorithm to solve the communication interference that may exist between users under the same micro base station and improve the communication rate between the user and the local micro base station.

[0168] Step 1): Let the number of population iterations be K, the number of particles in each iteration be U (i.e., the number of feasible solutions), the power optimization random factors γ1∈[0,1], γ2∈[0,1], the power influence factors θ1, θ2, and the maximum power adjustment factor δ. max Minimum power adjustment factor δ min ;

[0169] Step 2): Let set I s Power of each user i Power change Initialize U feasible solutions in Current optimal solution Global optimal solution Allocable computing resources for user i Among them, f i s Given the available computing resources for user i on micro base station s, the fitness function is... Let the iteration count variable k = 0, and the feasible solution count variable u = 0;

[0170] Step 3): Let k = k + 1. If k ≤ K, jump to step 4); otherwise, jump to step 11.

[0171] Step 4): Let u = u + 1. If u ≤ U, jump to step 5); otherwise, let u = 0 and jump to step 3.

[0172] Step 5): Update set I according to formulas (19)-(21). s The power change for each user i power Power adjustment factor δ(k);

[0173] Step 6): For set I s For each user i in the list, if the power of user i is... make If yes, proceed to step 7); otherwise, proceed to step 7.

[0174] Step 7): For set I s For each user i in the list, if the power of user i is... make If yes, proceed to step 8); otherwise, proceed to step 8.

[0175] Step 8): Let set I s Power of each user i Update according to formula (12) Set I s User i in The values ​​are sorted in ascending order, and f is updated according to formula (13). i s ,make renew

[0176] Step 9): If make If the previous step is not successful, proceed to step 10; otherwise, proceed to step 10.

[0177] Step 10): If make If the previous step is incorrect, proceed to step 4; otherwise, proceed to step 4.

[0178] Step 11): For set I s For each user i, let p i equal to φ g The i-th dimension component φ g,i Calculate according to formula (12) Set I s User i in The values ​​are sorted in ascending order, and f is calculated according to formula (13). i s ;

[0179] Step 12): The algorithm ends.

[0180] A task offloading method based on NOMA heterogeneous cellular networks, the specific implementation method of which includes the following steps:

[0181] Step 1: Let the set of all users be I, and the set of micro base stations be S. Add the users within the effective coverage area of ​​each micro base station s in S to the set I. s Initialize the transmit power of each user i in I.

[0182] Step 2: For each user i in set I, if Uninstall mode m i =1, otherwise, let the uninstallation mode m i =2, put user i into temporary set I c middle;

[0183] Step 3: For each set I s For each user i in the process, based on the uninstallation mode m i And calculate using formula (12) Determine the computing resource requirements and I s Users in The values ​​are sorted in ascending order. According to formula (13), the computing resources f that micro base station s can allocate to user i can be calculated. i s ;

[0184] Step 4: For each set I s Sub-algorithm 1 is called respectively to optimize I. s Transmit power for each user and update and f i s ,if Let m i =3, change I s Users are added to set I in sequence. c If the condition is met, proceed to step 5; otherwise, proceed to step 5.

[0185] Step 5: Based on f i s Calculate the number of task units that user i can offload to micro base station s. According to uninstallation mode m i and update f using formulas (12)-(18) s and n i ,if Skip to step 6; otherwise, skip to step 9.

[0186] Step 6: Add f from set S sMicro base stations s with a value greater than 0 are moved into temporary set S. c ,if Set I c All users in the middle according to n i If sorted in ascending order, proceed to step 7; otherwise, proceed to step 10.

[0187] Step 7: Set S c Chinese f s The largest micro base station s, as the current cooperating micro base station, will set I c The first user i in the set of users I of the cooperative micro base station s s According to the uninstallation mode m i And calculate f using formulas (12)-(17) i s and if Skip to step 7; otherwise, skip to step 8.

[0188] Step 8: Set S c Chinese f s Micro base stations s with a value of 0 are from set S c Move into S, if Proceed to step 9; otherwise, all users in set I will perform task uninstallation and proceed to step 10.

[0189] Step 9: For each user set I s ,make Update the computing resources f allocated to each user i in the micro base station s. i s All users in set I execute the task to uninstall;

[0190] Step 10: Algorithm ends.

[0191] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions.

[0192] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0193] The above embodiments should be understood as illustrative only and not as limiting the scope of protection of the present invention. After reading the description of the present invention, those skilled in the art can make various alterations or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A task offloading method based on NOMA heterogeneous cellular networks, used in sparse network scenarios where macro base stations are not equipped with servers, micro base stations are equipped with servers, and their effective coverage areas do not overlap, characterized in that, Includes the following steps:

101. Let the set of all users be... The micro base station set is ,Will Each micro base station Users within the effective coverage area are added to the set. ,initialization Each user Transmission power ; 102. For sets Each user in ,if Uninstall mode Otherwise, enable uninstallation mode. , will users Add to temporary set middle; 103. For each set Each user in The number of task units to be uninstalled based on user requests. User to micro base station transmission rate , Calculate the computing resource requirements. and will Users in The values ​​are sorted in ascending order, according to... and micro base stations Maximum available computing resources Find the micro base station Available for users Allocated computing resources ; 104. For each set Sub-algorithm 1 is called respectively to optimize. Transmit power for each user and update and ,if ,make ,Will Users are added to the set in turn. If the condition is met, proceed to step 105; otherwise, proceed to step 105.

105. According to Calculate users Can be uninstalled to micro base stations Number of task units ,according to , renew ,make ,if If yes, proceed to step 106; otherwise, proceed to step 109.

106. Set middle micro base stations Move to temporary set ,if , will set All users in China If sorted in ascending order, proceed to step 107; otherwise, proceed to step 110.

107. Set middle Largest micro base station As a current collaborative micro base station, it will integrate The first user in Joining cooperative micro base stations user set and according to , and calculate and ,make , ,if If yes, proceed to step 107; otherwise, proceed to step 108.

108. Set middle micro base stations From the set move in ,if If not, proceed to step 109; otherwise, set. All users in the process will perform the uninstallation task. Proceed to step 110.

109. For each user set ,make Update each user In micro base stations Computing resources allocated in ,gather All users in the system should perform the uninstallation task; 110. End; In step 103, the resource requirements are calculated. The calculation method is shown in formula (1): (1) in, Indicates user Dispatch to micro base station Number of task units This represents the smallest unit of task. Indicates computational complexity. Indicates user Maximum tolerable latency, Indicates user Transmit task to micro base station The rate is calculated as shown in formula (2). Indicates user Transmit task to macro base station The rate is calculated as shown in formula (3). Indicates macro base station Transmit task to micro base station The rate is calculated as shown in formula (4). Indicates micro base station Transmit task to macro base station The rate is calculated as shown in formula (5); user Transmit task to micro base station rate The calculation method is shown in formula (2): (2) in, Indicates user With micro base stations Inter-communication bandwidth, Indicates user The transmission power, Indicates user With micro base stations Channel gain between This represents the power of additive white Gaussian noise. Represents a set of users Excluding users Any other user besides If it is a binary variable, Then the user For users Interference with users, making Otherwise, let , Indicates other users to user The sum of interference power; user Transmit task to macro base station rate The calculation method is shown in formula (3): (3) in, This represents the bandwidth of the orthogonal sub-channels allocated to the macro base station. Indicates user With macro base stations Channel gain between; macro base station Transmit task to micro base station rate The calculation method is shown in formula (4): (4) in, Indicates macro base station The transmission power, Indicates cooperative micro base station With macro base stations Channel gain between; micro base station Transmit task to macro base station rate The calculation method is shown in formula (5): (5) in, Indicates micro base station The transmission power, Indicates micro base station With macro base stations Channel gain between; The sub-algorithm 1 in step 104 specifically includes: The user communicates with the local micro base station using NOMA technology. A user power optimization model is constructed using the particle swarm optimization algorithm, specifically including: Step 1): Let the number of population iterations be... The number of particles in each iteration, i.e., the number of feasible solutions, is Power optimization random factor , Power Influence Factor , Maximum power adjustment factor Minimum power adjustment factor ; Step 2): Let set Each user power Power change ,initialization One feasible solution ,in Current optimal solution Global optimal solution ,user Allocable computing resources ,in, micro base stations Up users Available computing resources, fitness function Let the iteration count variable Feasible solution count variables ; Step 3): Let ,if If yes, proceed to step 4; otherwise, proceed to step 11. Step 4): Let ,if (Jump to step 5), otherwise, let (Jump to step 3); Step 5): Update the collection Each user power change ,power Power adjustment factor ; Step 6): For the set Each user in If the user power ,make If yes, proceed to step 7); otherwise, proceed to step 7. Step 7): For the set Each user in If the user power ,make If yes, proceed to step 8); otherwise, proceed to step 8. Step 8): Let set Each user power ,according to , and renew , will set users in according to The values ​​are sorted in ascending order, based on and renew ,make ,renew ; Step 9): If ,make If yes, proceed to step 10); otherwise, proceed to step 10. Step 10): If ,make If yes, proceed to step 4); otherwise, proceed to step 4. Step 11): For the set Each user ,make equal The Dimensional components ,according to , and ,calculate , will set users in according to The values ​​are sorted in ascending order, based on and calculate ; Step 12): The algorithm ends.

2. The task offloading method based on NOMA heterogeneous cellular networks according to claim 1, characterized in that, In step 105, the user Dispatch to micro base station Number of task units The calculation method is shown in formulas (6)-(8). If the unloading mode ,user Dispatch to micro base station Number of task units The calculation method is shown in formula (6): (6) in, Indicates micro base station Available for users Allocated computing resources; If uninstallation mode ,user Dispatch to cooperative micro base station Number of task units The calculation method is shown in formula (7): (7) If uninstallation mode ,user Dispatch to cooperative micro base station Number of task units The calculation method is shown in formula (8): (8)。 3. The task offloading method based on NOMA heterogeneous cellular networks according to claim 1, characterized in that, In step 103, the micro base station Available for users Allocated computing resources The calculation method is shown in formula (9): (9) in, micro base stations The currently available resources, For binary variables, if the micro base station For users Allocate computing resources, let Otherwise, let .

4. The task offloading method based on NOMA heterogeneous cellular networks according to claim 1, characterized in that, In step 105, the micro base station Current available resources The calculation method is shown in formula (10): (10) For binary variables, if the micro base station For users Allocate computing resources, let Otherwise, let .

5. A task offloading method based on a NOMA heterogeneous cellular network according to claim 1, characterized in that, In step 5), the set is updated. Each user power change ,power Power adjustment factor The methods are shown in formulas (11), (12), and (13): (11) (12) (13) in, This indicates the maximum power adjustment value. This indicates the minimum power adjustment value. Indicates the number of population iterations. , Indicates the power influence factor. , This represents the random factor for power optimization. , , Indicates the current optimal solution Chinese users The optimal power, Represents the global optimal solution Chinese users The optimal power.

Images