An adaptive fixed-priority semi-perspective mixed-criticality task energy consumption optimization method

Abstract

This invention discloses an adaptive fixed-priority semi-perspective mixed-criticality task energy consumption optimization method, comprising the following steps: establishing a semi-perspective, non-precise mixed-criticality scheduling model; calculating the response time of mixed-criticality tasks in both low and high system modes using a response time analysis method; determining the priority of tasks based on the relationship between response time and their deadline using an optimal priority allocation method; and calculating the optimal energy consumption rate for the task set. By using the response time analysis method and the optimal priority allocation method to calculate the optimal energy consumption rate, the present invention not only effectively improves system utilization but also reduces system energy consumption.

Description

Technical Field

[0001] The present invention relates to real-time systems, industrial control systems and real-time scheduling of mixed-criticality systems, and in particular to an adaptive fixed-priority semi-perspective mixed-criticality task energy consumption optimization method. Background Art

[0002] Hybrid criticality systems integrate applications of varying criticality onto a shared platform to meet size, weight, and energy requirements. A common example of hybrid systems is aerospace control systems, which are governed by the international standard DO-178C. This standard specifies five levels of criticality (AE), with A being the highest and E being the lowest. Battery-powered drones are a typical example of hybrid criticality systems, where energy consumption is crucial due to limited battery capacity.

[0003] Existing research on mixed-criticality scheduling primarily focuses on system schedulability, with relatively little consideration of energy consumption. The few studies that do exist focus on non-perspective mixed-criticality models, which results in limited task set schedulability and low system resource utilization. Therefore, this paper proposes an adaptive fixed-priority semi-perspective mixed-criticality task energy optimization method. This method adopts a preemptive fixed-priority scheduling strategy and uses an optimal priority assignment method to determine the priority of tasks based on the relationship between response time and deadline. Finally, it determines the speed of energy optimization, effectively improving system utilization while reducing system energy consumption. Summary of the Invention

[0004] The purpose of the present invention is to overcome the deficiencies of the prior art and provide an adaptive fixed-priority semi-perspective hybrid critical task energy consumption optimization method.

[0005] In order to achieve the above object, the technical solution of the present invention is:

[0006] Establish a semi-perspective non-precise mixed-criticality scheduling model;

[0007] Calculate the response time of mixed critical tasks when the system is in low mode and high mode using response time analysis method;

[0008] Determine the priority of tasks using the optimal priority assignment method based on the relationship between response time and their deadline;

[0009] Calculate the optimal energy consumption rate for the task set;

[0010] The establishment of a semi-perspective imprecise mixed-key scheduling model comprises:

[0011] The main difference between the semi-transparent mixed-criticality task model and the traditional non-transparent mixed-criticality task model lies in the different system mode transition mechanism. In the semi-transparent mixed-criticality task model, when a high-criticality task arrives, it can be determined whether the system will switch from low mode to high mode. In the traditional non-transparent mixed-criticality task model, the system will not switch from low mode to high mode until the high-criticality task starts executing and its execution time exceeds the execution time of the low mode.

[0012] On a single processor, a set of semi-perspective inexact mixed critical period tasks Γ={τ1,τ2,…,τ n}, use the preemptive fixed priority strategy to schedule the task set. Mixed critical period task τ i (1≤i≤n, i is an integer) is composed of a quaternion {T i ,D i ,ξ i ,C i}, where T i Denotes a mixed-criticality period task τ i The cycle of D i Denotes a mixed-criticality period task τ i The relative deadline of i ξ i Denotes a mixed-criticality period task τ i The key level of ξ i ={LO,HI}, mixed critical period task τ i When the critical level is LO, it is a low-criticality task, and the mixed-criticality period task τ i When the critical level is HI, it is a high-criticality task; i Denotes a mixed-criticality period task τ i The worst-case execution time under different modes; C i (LO) and C i (HI) respectively represent the mixed critical period tasks τ i Execution time in low mode and high mode; The difference between the non-precise mixed criticality period task set and the traditional mixed criticality period task model is that when the system is in high mode, not only the high criticality level tasks improve their execution time, but also provide degraded services for low criticality level tasks. The so-called degraded service means that when the system is in high mode, the execution time of low criticality level tasks is lower than its execution time when the system is in low mode. The traditional mixed criticality period task model only executes high criticality level tasks and abandons all low criticality level tasks. The system is in low mode when any task τ i , its execution time does not exceed C i (LO); The system is in high mode when there are high-criticality tasks τ i, whose execution time exceeds C i (LO) but not exceeding C i (HI) and provide degraded service for low-criticality levels. If the mixed-criticality period task τ i For low-criticality tasks, C i (HI) <C i (LO); If the mixed critical period task τ i For high-criticality tasks, C i (HI)>=C i (LO). The so-called fixed priority strategy means that once the priority of a task is determined, its priority remains unchanged during the execution process.

[0013] The response time analysis method is used to calculate the response time of the mixed critical task when the system is in low mode and high mode; specifically including:

[0014] When the system is in low mode, the task τ i Response time R i (LO) is calculated by the following formula:

[0015]

[0016] Where j is a positive integer representing the subscript of the task; hp(i) represents the priority of the task τ i High task set; C i (LO) and C j (LO) respectively represent the task τ i and τ j Execution time in low mode;

[0017] When the system is in high mode, the task τ i Response time R i (HI) is calculated by the following formula:

[0018]

[0019] Among them, t s is the moment when the system switches from low mode to high mode, R i (LO) is the task τ i Response time in low mode, W i (LO) is the task τ i The worst-case start execution time in low mode is given by the following formula:

[0020]

[0021] Among them, C j (LO) represents the task τ j Execution time in low mode;j Denotes a mixed-criticality period task τ j cycle. is to satisfy t s <R i (LO), the task τ i response time in high mode; The calculation formula is given by the following formula:

[0022]

[0023] Where j is a positive integer representing the subscript of the task; t s Is the moment when the system switches from low mode to high mode; C i (LO) and C j (LO) respectively represent the task τ i and τ j Execution time in low mode; C j (HI) represents the task τ j Execution time in high mode; hpL(i) and hpH(i) represent the priority ratio of task τ i A high set of low-criticality level tasks and a high-criticality level task set.

[0024] is to satisfy t s <W i (LO), the task τ i Response time in high mode, The calculation formula is given by the following formula:

[0025]

[0026] Where j is a positive integer representing the subscript of the task; t s Is the moment when the system switches from low mode to high mode; C i (HI) and C j (HI) respectively represent the task τ i and τ j Execution time in high mode; C j (LO) represents the task τ j Execution time in low mode; hpL(i) and hpH(i) represent the priority ratio of task τ i A high set of low-criticality level tasks and a high-criticality level task set.

[0027] The method of determining the priority of a task using an optimal priority allocation method based on the relationship between the response time and the deadline comprises:

[0028] The semi-perspective imprecise mixed critical period task set Γ={τ1,τ2,…,τ n}, the priority allocation rules are as follows:

[0029] 1) Sort the high-criticality tasks in descending order according to their duration;

[0030] 2) Sort the low-criticality tasks in descending order of their duration;

[0031] 3) First select the high-criticality task τ with the largest period i Assign the lowest priority; if R i (LO) <D i And R i (LO) <D i , assign the lowest priority to it and remove it from the high-criticality level queue; if not satisfied, go to step 4);

[0032] 4) Select the low-criticality task τ with the largest period i Assign the lowest priority; if R i (LO) <D i And R i (LO) <D i , assign the lowest priority to it and remove it from the low-criticality queue; if it is not satisfied, it means that the task set is not schedulable;

[0033] 5) Repeat steps 3) and 4); assign the next lowest priority; until all tasks are assigned a priority, or the task set is unschedulable;

[0034] The optimal energy consumption rate S according to the energy consumption of the computing task set is calculated by the following formula:

[0035]

[0036] sqFt i ′(LO)<D i (7)

[0037] sqFt i ′(HI)<D i (8)

[0038] s min ≤S≤1 (9)

[0039] Among them, P ind is the power consumption independent of processor speed, θ is the ratio of the maximum dynamic power consumption to the static power consumption of the processor; LCM is the super period of all tasks in the task set Γ, that is, the least common multiple of all task periods; s min is the normalized minimum speed provided by the processor; D i Denotes a mixed-criticality period task τ iThe relative deadline of R′ i (LO) and R′ i (HI) are tasks τ i Response time in low mode and high mode at speed S; R i ′(LO) and R i ′(HI) The formulas are as follows:

[0040]

[0041] Among them, S represents the unified execution speed of all tasks in LO mode; C i (LO) and C j (LO) respectively represent the task τ i and τ j Execution time in low mode.

[0042]

[0043] Among them, t s It is the moment when the system switches from low mode to high mode, is to satisfy t s <R i ′(LO), task τ i The response time in high mode at speed s is given by the following formula

[0044]

[0045] Among them, C i (LO) and C j (LO) respectively represent the task τ i and τ j The execution time in low mode; S represents the uniform execution speed of all tasks in LO mode; j is a positive integer representing the subscript of the task; C j (HI) represents the task τ j Execution time in high mode; hpL(i) and hpH(i) represent the priority ratio of task τ i A high set of low-criticality level tasks and a high-criticality level task set.

[0046] is to satisfy t s <W i ′(LO), task τ i Response time in high mode at speed s, W i ′(LO) is the task τ i The worst-case start execution time at speed s in low mode is given by the following formula:

[0047]

[0048] Among them, C j (LO) represents the task τ j Execution time in low mode; T j Denotes a mixed-criticality period task τ j cycle; S represents the unified execution speed of all tasks in LO mode.

[0049] The calculation formula is given by the following formula:

[0050]

[0051]

[0052] Where j is a positive integer representing the subscript of the task; t s Is the moment when the system switches from low mode to high mode; C i (HI) and C j (HI) respectively represent the task τ i and τ j Execution time in high mode; C j (LO) represents the task τ j Execution time in low mode; hpL(i) and hpH(i) represent the priority ratio of task τ i A high set of low-criticality level tasks and a high-criticality level task set.

[0053] From the above description of the present invention, it can be seen that compared with the prior art, the present invention has the following beneficial effects:

[0054] (1) The method of the present invention saves 40.28% of energy consumption compared with other traditional algorithms;

[0055] (2) When the system is in high mode, the method of the present invention can provide degraded services for non-critical tasks, thereby improving the utilization of system resources;

[0056] (3) The method of the present invention can ensure the real-time performance and reliability of the system;

[0057] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. However, the adaptive fixed-priority semi-perspective mixed-criticality task energy consumption optimization method of the present invention is not limited to the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 Schematic diagram of a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0059] The following will describe and discuss the technical solutions in the embodiments of the present invention in detail in conjunction with the accompanying drawings of the present invention. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0060] See also Figure 1 As shown, the present invention provides an adaptive fixed-priority semi-perspective hybrid critical task energy consumption optimization method, comprising the following steps:

[0061] Step 1: Establish an inexact mixed-criticality scheduling model.

[0062] The main difference between the semi-transparent mixed-criticality task model and the traditional non-transparent mixed-criticality task model lies in the different system mode transition mechanism. In the semi-transparent mixed-criticality task model, when a high-criticality task arrives, it can be determined whether the system will switch from low mode to high mode. In the traditional non-transparent mixed-criticality task model, the system will not switch from low mode to high mode until the high-criticality task starts executing and its execution time exceeds the execution time of the low mode.

[0063] A set of semi-perspective inexact mixed critical period tasks Γ = {τ1, τ2, ..., τ ni}, use the preemptive fixed priority strategy to schedule the task set. Mixed critical period task τ i (1≤i≤n, i is an integer) is composed of a quaternion {T i , D i ,ξ i , C i}, where T i Denotes a mixed-criticality period task τ i The cycle of D i Denotes a mixed-criticality period task τ i The relative deadline of i ξ i Denotes a mixed-criticality period task τ i The key level of ξ i ={LO,HI}, mixed critical period task τ i When the critical level is LO, it is a low-criticality task, and the mixed-criticality period task τ i When the critical level is HI, it is a high-criticality task; i Denotes a mixed-criticality period task τ i The worst-case execution time under different modes; C i (LO) and C i (HI) respectively represent the mixed critical period tasks τ iExecution time in low mode and high mode; The difference between the non-precise mixed criticality period task set and the traditional mixed criticality period task model is that when the system is in high mode, not only the high criticality level tasks improve their execution time, but also provide degraded services for low criticality level tasks. The so-called degraded service means that when the system is in high mode, the execution time of low criticality level tasks is lower than its execution time when the system is in low mode. The traditional mixed criticality period task model only executes high criticality level tasks and abandons all low criticality level tasks. The system is in low mode when any task τ i , its execution time does not exceed C i (LO); The system is in high mode when there are high-criticality tasks τ i , whose execution time exceeds C i (LO) but not exceeding C i (HI) and provide degraded service for low-criticality levels. If the mixed-criticality period task τ i For low-criticality tasks, C i (HI) <C i (LO); If the mixed critical period task τ i For high-criticality tasks, C i (HI)>=C i (LO). The so-called fixed priority strategy means that once the priority of a task is determined, its priority remains unchanged during the execution process.

[0064] Step 2: Use the response time analysis method to calculate the response time of mixed-criticality tasks when the system is in low mode and high mode.

[0065] When the system is in low mode, the task τ i Response time R i (LO) is calculated by the following formula:

[0066]

[0067] Where j is a positive integer representing the subscript of the task; hp(i) represents the priority of the task τ i High task set; C i (LO) and C j (LO) respectively represent the task τ i and τ j Execution time in low mode;

[0068] When the system is in high mode, the task τ i Response time R i (HI) is calculated by the following formula:

[0069]

[0070] Among them, t s is the moment when the system switches from low mode to high mode, R i (LO) is the task τ i Response time in low mode, W i (LO) is the task τ i The worst-case start execution time in low mode is given by the following formula:

[0071]

[0072] Among them, C j (LO) represents the task τ j Execution time in low mode; j Denotes a mixed-criticality period task τ j cycle. is to satisfy t s <R i (LO), the task τ i response time in high mode; The calculation formula is given by the following formula:

[0073]

[0074]

[0075] Where j is a positive integer representing the subscript of the task; t s Is the moment when the system switches from low mode to high mode; C i (LO) and C j (LO) respectively represent the task τ i and τ j Execution time in low mode; C j (HI) represents the task τ j Execution time in high mode; hpL(i) and hpH(i) represent the priority ratio of task τ i A high set of low-criticality level tasks and a high-criticality level task set.

[0076] is to satisfy t s <W i (LO), the task τ i Response time in high mode, The calculation formula is given by the following formula:

[0077]

[0078] Where j is a positive integer representing the subscript of the task; t s Is the moment when the system switches from low mode to high mode; C i(HI) and C j (HI) respectively represent the task τ i and τ j Execution time in high mode; C j (LO) represents the task τ j Execution time in low mode; hpL(i) and hpH(i) represent the priority ratio of task τ i A high set of low-criticality level tasks and a high-criticality level task set.

[0079] Step 3: Determine the priority of the tasks using the optimal priority assignment method based on the relationship between the response time and its deadline.

[0080] Given a set of semi-perspective inexact mixed critical period tasks Γ={τ1,τ2,…,τ n}, the priority allocation rules are as follows:

[0081] 1) Sort the high-criticality tasks in descending order according to their duration;

[0082] 2) Sort the low-criticality tasks in descending order of their duration;

[0083] 3) First select the high-criticality task τ with the largest period i Assign the lowest priority; if R i (LO) <D i And R i (LO) <D i , assign the lowest priority to it and remove it from the high-criticality level queue; if not satisfied, go to step 4);

[0084] 4) Select the low-criticality task τ with the largest period i Assign the lowest priority; if R i (LO) <D i And R i (LO) <D i , assign the lowest priority to it and remove it from the low-criticality queue; if it is not satisfied, it means that the task set is not schedulable;

[0085] 5) Repeat steps 3) and 4); assign the next lowest priority; until all tasks are assigned a priority, or the task set is unschedulable;

[0086] Step 4: Calculate the optimal energy consumption rate of the task set.

[0087] The optimal energy consumption rate S of the task set is calculated by the following formula:

[0088]

[0089] sqFt i ′(LO) <D i (7)

[0090] sqFt i ′(HI) <D i (8)

[0091] s min ≤S≤1 (9)

[0092] Among them, P ind is the power consumption that is independent of processor speed, θ is the ratio of the maximum dynamic power consumption to the static power consumption of the processor; LCM is the super period of all tasks in the task set Γ, that is, the least common multiple of all task periods; S min is the normalized minimum speed provided by the processor; D i Denotes a mixed-criticality period task τ i The relative deadline of R i ′(LO) and R i ′(HI) are the tasks τ i Response time in low mode and high mode at speed S; R i ′(LO) and R i ′(HI) The formulas are as follows:

[0093]

[0094] Among them, S represents the unified execution speed of all tasks in LO mode; C i (LO) and C j (LO) respectively represent the task τ i and τ j Execution time in low mode.

[0095]

[0096] Among them, t s It is the moment when the system switches from low mode to high mode, is to satisfy t s <R i ′(LO), task τ i The response time at high speed S is given by the following formula:

[0097]

[0098] Among them, C i (LO) and C j (LO) respectively represent the task τ i and τ jThe execution time in low mode; S represents the uniform execution speed of all tasks in LO mode; j is a positive integer representing the subscript of the task; C j (HI) represents the task τ j Execution time in high mode; hpL(i) and hpH(i) represent the priority ratio of task τ i A high set of low-criticality level tasks and a high-criticality level task set.

[0099] is to satisfy t s <W i ′(LO), task τ i Response time in high mode at speed S, W i ′(LO) is the task τ i The worst-case start execution time at speed S in low mode is given by the following formula:

[0100]

[0101] Among them, C j (LO) represents the task τ j Execution time in low mode; T j Denotes a mixed-criticality period task τ j cycle; S represents the unified execution speed of all tasks in LO mode.

[0102] The calculation formula is given by the following formula:

[0103]

[0104] Where j is a positive integer representing the subscript of the task; t s Is the moment when the system switches from low mode to high mode; C i (HI) and C j (HI) respectively represent the task τ i and τ j Execution time in high mode; C j (LO) represents the task τ j Execution time in low mode; hpL(i) and hpH(i) represent the priority ratio of task τ i A high set of low-criticality level tasks and a high-criticality level task set.

[0105] In this embodiment, the mixed periodic task set Γ = {τ1, τ2, τ3} includes three periodic tasks; the periodic task τ1 has a period T1 of 5, a relative deadline D1 of 5, and a criticality level ξ1 of LO, that is, it is a low-criticality task, and its low-mode worst-case execution time C1(LO) is 1; its high-mode worst-case execution time C1(HI) is 0.5; the periodic task τ2 has a period T2 of 20, a relative deadline D2 of 20, and a criticality level ξ2 of HI, that is, it is a high-criticality task, and its low-mode worst-case execution time C2(LO) is 5; its high-mode worst-case execution time C2(HI) is 6; The period T3 of the scheduled task τ3 is equal to 100, its relative deadline D3 is 100, and the critical level ξ3 is HI, that is, it is a high-criticality task, and its high-mode worst-case execution time C3(LO) is 20; its high-mode worst-case execution time C3(HI) is 30; according to the algorithm rules of this paper, they are assigned priorities, and it is concluded that task τ1 has the highest priority, task τ2 has the second highest priority, and task τ3 has the lowest priority; by judging that this mode is schedulable, it is finally calculated that the energy consumption optimization speed S=0.65, the energy consumption of the method of the present invention is 50.46, and the energy consumption of other methods is 84.50, saving 40.28% of energy consumption.

[0106] The above is only a preferred embodiment of the present invention. However, the present invention is not limited to the above embodiment. Any equivalent changes and modifications made according to the present invention, as long as the resulting functions and effects do not exceed the scope of this solution, shall fall within the scope of protection of the present invention.

Claims

1. An adaptive fixed-priority semi-perspective hybrid critical task energy consumption optimization method, characterized in that: include: Establish a semi-perspective non-precise mixed-criticality scheduling model; Calculate the response time of mixed critical tasks when the system is in low mode and high mode using response time analysis method; Determine the priority of tasks using the optimal priority assignment method based on the relationship between response time and their deadline; Calculate the optimal energy consumption rate for the task set; The establishment of the semi-perspective inexact mixed-key scheduling model includes: On a single processor, a set of semi-perspective inexact mixed critical period tasks Γ={τ1,τ2,…,τ n }, use the preemptive fixed priority strategy to schedule the task set; mixed critical period task τ i By the quaternion {T i ,D i ,ξ i ,C i }, where 1≤i≤n, i is an integer; T i Denotes a mixed-criticality period task τ i The cycle of D i Denotes a mixed-criticality period task τ i The relative deadline of i ξ i Denotes a mixed-criticality period task τ i The key level is ξ i ={LO,HI}, mixed critical period task τ i When the critical level is LO, it is a low-criticality task, and the mixed-criticality period task τ i When the critical level is HI, it is a high-criticality task; i Denotes a mixed-criticality period task τ i The worst-case execution time under different modes; C i (LO) and C i (HI) respectively represent the mixed critical period tasks τ i Execution time in low mode and high mode; when the system is in high mode, it not only improves the execution time of high-criticality tasks, but also provides degraded services for low-criticality tasks; the degraded service means that when the system is in high mode, the execution time of low-criticality tasks is lower than its execution time when the system is in low mode; the system is in low mode means that any task τ i , its execution time does not exceed C i (LO); The system is in high mode when there are high-criticality tasks τ i , whose execution time exceeds C i (LO) but not exceeding C i (HI) and provide degraded service for low-criticality level; if the mixed-criticality period task τ i For low-criticality tasks, C i (HI) <C i (LO); If the mixed critical period task τ i For high-criticality tasks, C i (HI)>=C i (LO); The fixed priority strategy means that once the priority of a task is determined, its priority remains unchanged during the execution process; The response time analysis method is used to calculate the response time of the mixed critical task when the system is in low mode and high mode, specifically including: When the system is in low mode, the task τ i Response time R i (LO) is calculated by the following formula: Where j is a positive integer representing the subscript of the task; hp(i) represents the priority of the task τ i High task set; C i (LO) and C j (LO) distribution represents the task τ i and τ j Execution time in low mode; T j Denotes a mixed-criticality period task τ j Cycle When the system is in high mode, the task τ i Response time R i (HI) is calculated by the following formula: Among them, t s is the moment when the system switches from low mode to high mode, R i (LO) is the task τ i Response time in low mode, W i (LO) is the task τ i Worst-case start execution time in low mode; is to satisfy t s <R i (LO), the task τ i response time in high mode; is to satisfy t s <W i (LO), the task τ i Response time in high mode.

2. The adaptive fixed-priority semi-perspective hybrid critical task energy consumption optimization method according to claim 1, characterized in that: The method of determining the priority of a task by using an optimal priority allocation method based on the relationship between the response time and the deadline includes: The semi-perspective imprecise mixed critical period task set Γ={τ1,τ2,…,τ n }, the priority allocation rules are as follows: 1) Sort the high-criticality tasks in descending order according to their duration; 2) Sort the low-criticality tasks in descending order of their duration; 3) First select the high-criticality task τ with the largest period i Assign the lowest priority; if R i (LO) <D i And R i (LO) <D i , assign the lowest priority to it and remove it from the high-criticality level queue; if not satisfied, go to step 4); 4) Select the low-criticality task τ with the largest period i Assign the lowest priority; if R i (LO) <D i And R i (LO) <D i , assign the lowest priority to it and remove it from the low-criticality queue; if it is not satisfied, it means that the task set is not schedulable; 5) Repeat steps 3) and 4) to assign the next lowest priority; until all tasks are assigned a priority, or the task set is unschedulable.

3. The adaptive fixed-priority semi-perspective hybrid critical task energy consumption optimization method according to claim 2, characterized in that: The value of the optimal energy consumption rate S of the computing task set is calculated by the following formula: s.t. R i ′(LO)<D i (4) s.t. R i ′(HI)<D i (5) s.t. S min ≤S≤1 (6) Among them, P ind is the power consumption that is independent of processor speed, θ is the ratio of the maximum dynamic power consumption to the static power consumption of the processor; LCM is the super period of all tasks in the task set Γ, that is, the least common multiple of all task periods; S min is the normalized minimum speed provided by the processor; R i ′(LO) and R′ i (HI) are tasks τ i Response time at speed S in low mode and high mode; D i Denotes a mixed-criticality period task τ i relative deadlines.

Images