A method for non-uniform speed vehicle task offloading in a sparse rsu edge network

Abstract

The application claims a non-uniform speed vehicle task offloading method for coping with a sparse RSU edge network, belonging to the field of communication technology. In view of the problem that the high mobility and non-uniform speed driving of road vehicles may cause transmission link interruption and task offloading failure in the sparse RSU edge network scenario, a non-uniform speed vehicle task offloading method for coping with the sparse RSU edge network is proposed. The method solves the problem of short link duration between vehicles and RSUs and no RSU coverage by establishing a forwarding link between vehicles, adopts a hybrid offloading mode of V2I, V2V2I and local vehicle offloading, and improves the task offloading efficiency of the system. According to the real-time speed of the vehicle, the offloading period is dynamically adjusted, and according to the real-time position of the vehicle and the computing resource state of the system, the optimal offloading mode and computing resource allocation are selected, so as to maximize the system task completion rate and reduce the unit task offloading delay.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of communication, and particularly relates to a non-uniform vehicle task offloading method for coping with a sparse RSU edge network. BACKGROUND

[0002] With the rapid development of Internet of Vehicles technology, a large number of delay-sensitive and computation-intensive vehicle applications are emerging. Providing Vehicle to Infrastructure (V2I) based edge computing services for mobile vehicles on the road through Road Side Unit (RSU) has become a mainstream solution for current vehicle network business applications.

[0003] However, the existing research on vehicle task offloading in edge networks mostly considers the ideal scenario of vehicles driving on roads fully covered by RSUs, and less research is conducted on the problem that the high mobility and non-uniform speed characteristics of vehicles may further cause frequent interruption of transmission links or loss of tasks, thereby causing task offloading failure or deterioration of service quality, which may be caused by sparse deployment of RSUs or network function loss of partial road sections due to RSU failure.

[0004] In view of the above problems, the application provides a non-uniform vehicle task offloading method for coping with a sparse RSU edge network. According to the speed variation characteristics and task delay constraints of the task vehicle, the task offloading period is dynamically planned, a hybrid offloading mode combining local (Local), Vehicle to Infrastructure (V2I) and Vehicle to Vehicle to Infrastructure (V2V2I) is adopted, and the task scheduling strategy and computing resource allocation strategy are dynamically adjusted, so as to maximize the task completion rate and reduce the system unit task offloading delay. SUMMARY

[0005] The application aims to solve the problems of the prior art. A non-uniform vehicle task offloading method for coping with a sparse RSU edge network is provided. The technical scheme of the application is as follows:

[0006] A non-uniform vehicle task offloading method for coping with a sparse RSU edge network, comprising the following steps:

[0007] 101、According to the offloading request of the task vehicle i, the task data amount s i , the task computing complexity c i , and the task delay threshold Initialize the remaining task amount of the task vehicle i Tolerated delay Offloading period number k = 0;

[0008] 102、Let k = k + 1, and according to the current speed v i, calculate the kth unloading cycle length t of the task vehicle i i,k , the maximum task amount s that can be handled in the Local mode i,k , and the unit task processing delay τ in the Local mode i,k , let the unloading mode of the task vehicle i in the kth cycle be model = Local, and the unloading task amount of the task vehicle i unit task processing delay

[0009] 103, if the task vehicle i is in the coverage range of any roadside unit RSU j, according to the link duration of the task vehicle i and the RSU j and the computing resources that the RSU j can allocate to the task vehicle i calculate the maximum task amount s that the RSU j can handle in the V2I unloading mode of the task vehicle i in the time i,k , jump to step 104, otherwise, jump to step 106;

[0010] 104, if s i,k > 0, calculate the unit task processing delay τ of the task vehicle i in the V2I unloading mode i,k , jump to step 105, otherwise, jump to step 106;

[0011] 105, if let model = V2I, jump to step 107, otherwise, jump to step 107;

[0012] 106, find the best relay vehicle i' of the task vehicle i in the V2V2I unloading mode, where i ≠ i', calculate the maximum task amount s and the unit task processing delay τ that can be handled in the V2V2I mode i,k , if s i,k > 0 and let model = V2V2I, jump to step 107, otherwise, jump to step 107;

[0013] 107, perform task unloading according to the current unloading mode model, let

[0014] 108, the kth cycle task unloading is completed, let if and jump to step 102, otherwise, jump to step 109;

[0015] 109, the step ends.

[0016] Further, the step 102, the kth period length t i,k The calculation method is shown in formula (1):

[0017]

[0018] In formula (1), p and s represent constants for adjusting v i The relationship between t i,k , v i represents the speed of task vehicle i, v min represents the minimum speed of the vehicle, represents the remaining tolerance time delay of the task of task vehicle i.

[0019] Further, the step 102, 103, 106, the maximum task amount s i,k The calculation method is shown in formula (2):

[0020]

[0021] In formula (2), represents the floor function, f i represents the local computing resource of task vehicle i, and q represents the size of a unit task, represents the remaining task amount of task vehicle i, represents the link duration between task vehicle i and RSU j, represents the computing resource that RSU j can allocate to task vehicle i, represents the link duration between vehicle i' and its associated RSU j, represents the data transmission rate between vehicle i and RSU j, and the calculation method is shown in formula (3):

[0022]

[0023] In formula (3), B V2I represents the bandwidth of V2I link, p i represents the transmission power of task vehicle i, represents the channel gain between task vehicle i and RSU j, and N0 represents the noise power, represents the path loss between task vehicle i and RSU j.

[0024] In formula (2), represents the link duration between vehicle i and vehicle i', and the calculation method is shown in formula (4):

[0025]

[0026] wherein the calculation methods of t1, t2, t3, t4 are shown in formula (5) - formula (8) respectively:

[0027]

[0028] In formula (5) - formula (8), a max and a min respectively represent the maximum and minimum acceleration of the vehicle, R V2V represents the maximum communication distance between vehicles, d i,i′ represents the distance between task vehicle i and relay vehicle i', μ i,i′ represents the position relationship between task vehicle i and relay vehicle i', the calculation method is shown in formula (9):

[0029]

[0030] In formula (9), x i represents the position of task vehicle i, x i′ represents the position of relay vehicle i';

[0031] In formula (2), V represents the V2V2I link transmission rate, the calculation method is shown in formula (10):

[0032]

[0033] In formula (10), V represents the V2V link transmission rate, the calculation method is shown in formula (11):

[0034]

[0035] In formula (11), B V2V represents the V2V link bandwidth, represents the channel gain between task vehicle i and relay vehicle i', d represents the path loss between task vehicle i and relay vehicle i'.

[0036] Further, the unit task processing delay τ i,k in the local offloading Local mode, V2I offloading mode, V2V2I offloading mode in steps 102, 103, 106

[0037]

[0038] Further, the link duration between task vehicle i and RSU j in step 103 The calculation method is shown in equation (13):

[0039]

[0040] In equation (13), x i represents the position of the task vehicle i, x j represents the position of the RSU j, a max represents the maximum acceleration of the vehicle, and R represents the coverage radius of the RSU.

[0041] Further, the RSU j can assign the calculation resource The calculation method is shown in equation (14):

[0042]

[0043] In equation (14), x represents the currently available calculation resource of the RSU j, represents the total calculation resource in the RSU j, γ represents a constant number other than 0, and n represents the number of task vehicles in the coverage range of the RSU j.

[0044] Further, in the step 106, the best relay vehicle i' in the V2V2I offloading mode is found for the task vehicle i, and the maximum task amount s i,k and the unit task processing delay τ i,k that can be processed in the V2V2I mode are calculated, and the specific steps include the following steps.

[0045] 1) The maximum task amount s i,k = 0 that can be processed by the task vehicle i in the V2V2I mode, the unit task processing delay τ i,k = ∞, and the candidate relay vehicle set I' are initialized.

[0046] 2) The idle vehicles within one hop range from the task vehicle i and in the coverage range of any RSU j are added to the set I';

[0047] 3) If any vehicle i' is taken out from I', the maximum task amount s that can be processed by the V2V2I link within t time is calculated according to the link duration between the task vehicle i and the vehicle i' the link duration between the vehicle i' and the associated RSU j of the vehicle i' and the calculation resource of the RSU j that can be assigned to the task vehicle i. and the unit task processing delay τ jump to step 4), otherwise, jump to step 5).

[0048] 4) If and Let Let i' be the relay vehicle of task vehicle i, jump to step 3), otherwise, jump to step 3);

[0049] 5) Output the maximum task amount s that the task vehicle i can handle in the V2V2I mode i,k and the unit task processing delay τ i,k ;

[0050] 6) The step ends.

[0051] The advantages and beneficial effects of the present application are as follows:

[0052] The application discloses a non-uniform vehicle task offloading method for a sparse RSU edge network. Existing researches on vehicle task offloading problems in edge networks mostly consider ideal scenarios in which vehicles drive on roads fully covered by RSUs, and less researches are conducted on problems that sparse deployment of RSUs or RSU failures may cause network function loss on partial road sections, and high mobility and non-uniform speed characteristics of vehicles may further cause frequent interruption of transmission links or loss of tasks, thereby causing task offloading failure or deterioration of service quality. In view of the above problems, the application provides a non-uniform vehicle task offloading method for a sparse RSU edge network. According to speed changes of task vehicles and task delay constraints, a task offloading period is dynamically planned, a mixed offloading mode of Local, V2I and V2V2I is adopted according to link duration between vehicles and RSUs and between vehicles and vehicles in different periods, a task scheduling strategy and a computing resource allocation strategy are dynamically adjusted, system resource utilization is optimized, task completion rate is maximized, and system unit task offloading delay is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 is a flow chart of the non-uniform vehicle task offloading method for a sparse RSU edge network provided by the application. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the application will be described clearly and in detail with reference to the accompanying drawings in the embodiments of the application. The described embodiments are only some of the embodiments of the application.

[0055] The technical solution of the application to solve the above technical problems is:

[0056] The concepts and models involved in the content of the application are as follows:

[0057] The present application assumes that a sparse RSU edge network is composed of RSUs and vehicles, the RSUs are sparsely deployed on one side of the road, the RSU set is represented as J (j∈J), an edge server is configured on each RSU, and the edge server can provide task offloading services for vehicle terminals. Vehicles on the road travel at a non-uniform speed within a certain speed range, the vehicle set is represented as I (i∈I), each vehicle can obtain real-time information such as the current position and speed through an on-board unit (OBU), a task vehicle can offload tasks to an RSU through a V2I or V2V2I mode, or the tasks can be calculated locally. Considering the uncertainty of the non-uniform vehicle travel state, a dynamic periodic scheduling mode is adopted for task offloading. The offloading cycle set of vehicle i is represented as K i (k∈K i ), since the result size is much smaller than the input data size, the result return delay is ignored.

[0058] The task request of vehicle i is represented as where s i represents the task data size, c i represents the task computing complexity, represents the maximum tolerable delay of the task. The task A i is split into subtasks with the same computing complexity in units of tasks, and is calculated in cycles. Let represent the task offloading size and the unit task processing delay of vehicle i in the kth cycle, respectively. Let I' (i'∈I') represent the relay vehicle set of task vehicle i.

[0059] The technical solutions of the present application are as follows:

[0060] 1. The length t i,k of the kth cycle of task vehicle i is calculated as shown in formula (1):

[0061]

[0062] In formula (1), ρ and σ are constants, which are used to adjust the relationship between v i and t i,k , v i represents the speed of task vehicle i, v min represents the minimum speed of the vehicle, represents the remaining tolerable delay of task vehicle i.

[0063] 2. The maximum task size s i,k that can be processed by task vehicle i in the local offloading Local mode, the V2I offloading mode, and the V2V2I offloading mode is calculated as shown in formula (2):

[0064]

[0065] In formula (2), denotes the floor function, f i denotes the local computing resource of task vehicle i, and θ denotes the unit task size, denotes the remaining task amount of task vehicle i, denotes the link duration between task vehicle i and RSU j, denotes the computing resource that RSU j can allocate for task vehicle i, denotes the link duration between vehicle i' and its associated RSU j, denotes the data transmission rate between vehicle i and RSU j, and the calculation method is shown in formula (3):

[0066]

[0067] In formula (3), B V2I denotes the bandwidth of the V2I link, p i denotes the transmission power of task vehicle i, denotes the channel gain between task vehicle i and RSU j, and N0 denotes the noise power, denotes the path loss between task vehicle i and RSU j;

[0068] In formula (2), denotes the link duration between vehicle i and vehicle i', and the calculation method is shown in formula (4):

[0069]

[0070] wherein the calculation methods of t1, t2, t3, and t4 are shown in formula (5) to formula (8) respectively:

[0071]

[0072] In formula (5) to formula (8), a max and a min denote the maximum and minimum acceleration of the vehicle respectively, R V2V denotes the maximum communication distance between vehicles, d i,i′ denotes the distance between task vehicle i and relay vehicle i', and μ i,i′ denotes the positional relationship between task vehicle i and relay vehicle i', and the calculation method is shown in formula (9):

[0073]

[0074] In formula (9), x i denotes the position of task vehicle i, and x i′denotes the position of the relay vehicle i';

[0075] In formula (2), denotes the V2V2I link transmission rate, and the calculation method is shown in formula (10):

[0076]

[0077] In formula (10), denotes the V2V link transmission rate, and the calculation method is shown in formula (11):

[0078]

[0079] In formula (11), B V2V denotes the V2V link bandwidth, denotes the channel gain between the task vehicle i and the relay vehicle i', denotes the path loss between the task vehicle i and the relay vehicle i';

[0080] 3, the unit task processing delay τ in the local offloading Local mode, V2I offloading mode, V2V2I offloading mode i,k The calculation method is shown in formula (12):

[0081]

[0082] 4, the link duration of the task vehicle i and the RSU j The calculation method is shown in formula (13):

[0083]

[0084] In formula (13), x i denotes the position of the task vehicle i, x j denotes the position of the RSU j, a max denotes the maximum acceleration of the vehicle, and R denotes the coverage radius of the RSU;

[0085] 5, the computing resources that the RSU j can allocate to the task vehicle i The calculation method is shown in formula (14):

[0086]

[0087] In formula (14), denotes the currently available computing resources of the RSU j, denotes the total computing resources of the RSU j, γ denotes a constant not equal to 0, and n denotes the number of task vehicles in the coverage range of the RSU j; 6, sub algorithm 1: finding the best relay vehicle i' for the task vehicle i in the V2V2I offloading mode

[0088] Step 1: initialize the maximum amount of tasks s that the task vehicle i can handle in the V2V2I offloading mode i,k = 0, unit task processing delay τ i,k = ∞, candidate relay vehicle set

[0089] Step 2: add idle vehicles within one-hop range of the task vehicle i and within the coverage range of any RSU j to the set I';

[0090] Step 3: if take any vehicle i' from I', according to the link duration between the task vehicle i and the vehicle i' the link duration between the vehicle i' and its associated RSU j and the computing resources that the RSU j can allocate for the task vehicle i calculate the maximum amount of tasks s that the V2V2I link can handle in time and the unit task processing delay τ jump to Step 4), otherwise, jump to Step 5);

[0091] Step 4: if and let let i' be the relay vehicle of the task vehicle i, jump to Step 3), otherwise, jump to Step 3);

[0092] Step 5: output the maximum amount of tasks s that the task vehicle i can handle in the V2V2I offloading mode i,k and the unit task processing delay τ i,k ;

[0093] Step 6: the step ends.

[0094] A non-uniform vehicle task offloading method for sparse RSU edge networks, the specific implementation method comprising the following steps:

[0095] Step 1: according to the offloading request of the task vehicle i, obtain the task data amount s i , the task computing complexity c i , and the task delay threshold initialize the remaining task amount of the task vehicle i the remaining tolerance delay the offloading cycle number k = 0;

[0096] Step 2: let k = k + 1, according to the current speed v i of the task vehicle i, calculate the length t i,kthe maximum amount of tasks s that can be processed by the task vehicle i in the Local mode i,k and the unit task processing latency τ in the Local mode i,k , let the task vehicle i be in the kth period of the unloading mode model = Local, the unloading task amount of the task vehicle i the unit task processing latency

[0097] Step 3: If the task vehicle i is in the coverage range of any roadside unit RSU j, according to the link duration of the task vehicle i and the RSU j and the computing resources that the RSU j can allocate to the task vehicle i calculate the maximum amount of tasks s that can be processed by the task vehicle i in the V2I unloading mode at the RSU j in the time period i,k , jump to step 104, otherwise, jump to step 106;

[0098] Step 4: If s i,k > 0, calculate the unit task processing latency τ of the task vehicle i in the V2I unloading mode i,k , jump to step 105, otherwise, jump to step 106;

[0099] Step 5: If let model = V2I, jump to step 107, otherwise, jump to step 107;

[0100] Step 6: Call sub-algorithm 1 to find the best relay vehicle i' for the task vehicle i in the V2V2I unloading mode, where i≠ i', calculate the maximum amount of tasks s that can be processed in the V2V2I unloading mode i,k and the unit task processing latency τ i,k If s i,k > 0, and let model = V2V2I, jump to step 107, otherwise, jump to step 107;

[0101] Step 7: Perform task unloading according to the current unloading mode model, let

[0102] Step 8: The kth period of task unloading is completed, let If and jump to step 102, otherwise, jump to step 109;

[0103] Step 9: The step is ended.

[0104] The systems, apparatuses, modules or units disclosed in the above examples can be implemented by computer chips or entities, or by products with certain functions.

[0105] It should be further noted that the terms "comprise", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that processes, methods, articles, or apparatuses that comprise a list of elements do not include only those elements but can also include other elements not expressly listed or inherent to such processes, methods, articles, or apparatuses. Without further limitation, an element defined by the phrase "comprising a" does not exclude the existence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0106] The above examples are to be understood only as illustrative of the application and not a limitation of the scope of protection for the present application. After reading the specification, the skilled person can make various changes or modifications to the present application, and these equivalent changes and modifications also fall within the scope defined by the claims of the present application.

Claims

1. A method for non-uniform vehicle task offloading for sparse RSU edge network, characterized in that, The method comprises the following steps: 101、According to the unloading request of the task vehicle i, obtain the task data volume s i , task computing complexity c i , task time delay threshold Initialize the remaining task volume of the task vehicle i Remaining tolerance time delay Unloading cycle number k = 0; 102. Let k = k + 1, and calculate the current speed v of the task vehicle i. i Calculate the duration t of the kth unloading cycle for task vehicle i. i,k The maximum number of tasks that can be handled in Local mode (s) i,k And the unit task processing latency τ in Local mode i,k Let model = Local for the unloading mode of task vehicle i in the kth period, and let the unloading task volume of task vehicle i be... Unit task processing latency 103、If the task vehicle i is in the coverage range of any roadside unit RSUj, according to the link duration of the task vehicle i and the RSUj And the computing resources that the RSUj can allocate to the task vehicle i Calculate The maximum amount of tasks that the RSUj can handle in the V2I unloading mode within the time s i,k , jump to step 104, otherwise, jump to step 106; 104、If s i,k > 0, calculate the unit task processing delay τ i,k of the task vehicle i in the V2I unloading mode, jump to step 105, otherwise, jump to step 106; 105、if Let model = V2I, Go to step 107, otherwise, go to step 107; 106. Find the best relay vehicle i' for the task vehicle i in the V2V2I offloading mode, where i≠i', calculate the maximum amount of tasks s that can be processed in the V2V2I offloading mode i,k and the unit task processing delay τ i,k If s i,k >0 and Let model=V2V2I, Jump to step 107, otherwise, jump to step 107; 107. Perform task offloading according to the current offloading mode model, and 108、the kth period task offloading is completed, let If And Jump to step 102, otherwise, jump to step 109; 109、End of step.

2. The method of claim 1, wherein, The step 102 calculates the kth cycle duration t of the task vehicle i i,k The calculation method is shown in formula (1): In Equation (1), p and s represent constants for adjusting v i between t i,k and t i represents the speed of the task vehicle i, v min represents the minimum speed of the vehicle, represents the remaining tolerance time delay of the task of the task vehicle i.

3. The method of claim 2, wherein, The maximum task amount s that the task vehicle i can handle in the local unloading Local mode, V2I unloading mode, and V2V2I unloading mode in the steps 102, 103, and 106 i,k The calculation method is shown in formula (2): In formula (2), denotes the floor function, f i denotes the local computing resource of task vehicle i, and θ denotes the unit task size, denotes the remaining task amount of task vehicle i, denotes the link duration between task vehicle i and RSUj, denotes the computing resource that RSUj can allocate to task vehicle i, denotes the link duration between vehicle i' and its associated RSUj, denotes the data transmission rate between vehicle i and RSUj, and the calculation method is shown in formula (3): In Equation (3), B V2I denotes the bandwidth of the V2I link, p i denotes the transmit power of the task vehicle i, denotes the channel gain between the task vehicle i and the RSUj, N0denotes the noise power, denotes the path loss between the task vehicle i and the RSUj; In equation (2), denotes the link duration between vehicle i and vehicle i', which is calculated as shown in equation (4): Wherein, the calculation method of t1, t2, t3, t4 is shown in formula (5) - formula (8) respectively: In the formulas (5) - (8), a max and a min represent the maximum and minimum acceleration of the vehicle, R V2V represents the maximum communication distance between vehicles, d i,i′ represents the distance between the task vehicle i and the relay vehicle i', and μ i,i′ represents the positional relationship between the task vehicle i and the relay vehicle i', and the calculation method is shown in the formula (9): In Equation (9), x i represents the position of the task vehicle i, x i′ represents the position of the relay vehicle i'. In equation (2), denotes the V2V2I link transmission rate, which is calculated as shown in equation (10): In formula (10), denotes the V2V link transmission rate, and the calculation method is shown in formula (11): In Equation (11), B V2V denotes the V2V link bandwidth, denotes the channel gain between task vehicle i and relay vehicle i', denotes the path loss between task vehicle i and relay vehicle i'.

4. The method of claim 1, wherein, The unit task processing delay τ in the local offloading Local mode, V2I offloading mode, and V2V2I offloading mode in steps 102, 103, and 106 i,k The calculation method is shown in formula (12):

5. The method of claim 1, wherein, The link duration of the task vehicle i with the RSUj in the step 103 The calculation method is shown as formula (13): In Equation (13), x i represents the position of the task vehicle i, x j represents the position of the RSU j, a max represents the maximum acceleration of the vehicle, and R represents the coverage radius of the RSU.

6. The method of claim 1, wherein, The step 103, the RSUj can allocate the computing resource for the task vehicle i The calculation method is shown as formula (14): In formula (14), denotes the total available computing resources of RSUj, denotes the total computing resources within RSUj, γ denotes a constant not equal to 0, and n denotes the number of task vehicles within the coverage of RSUj.

7. The method of claim 1, wherein, The step 106 finds the best relay vehicle i' for the task vehicle i in the V2V2I offloading mode and calculates the maximum amount of tasks s that can be processed in the V2V2I offloading mode i,k and the unit task processing delay τ i,k , and specifically includes the following steps: 1) initialization of the maximum amount of tasks s that the task vehicle i can handle in the V2V2I offloading mode i,k = 0, unit task processing delay τ i,k = ∞, candidate relay vehicle set 2) Add the idle vehicles within one-hop range of the task vehicle i and within the coverage range of any RSU j to the set I'; 3) If For any vehicle i' from I', according to the link duration between the task vehicle i and the vehicle i' The link duration between the vehicle i' and its associated RSUj And the computing resources that RSUj can allocate for the task vehicle i Compute the maximum amount of tasks that the V2V2I link can handle in The time And the unit task processing delay Go to step 4), otherwise, go to step 5); 4) if and Let Let i' be the relay vehicle of task vehicle i, go to step 3), otherwise, go to step 3); 5) the maximum amount of tasks s that the task vehicle i can handle in the V2V2I offloading mode i,k and the unit task processing delay τ i,k ; 6) End of step.

Images