Task Scheduling for Multi-Cloud Computing Subject to Security and Reliability Constraints Qing-Hua Zhu, Senior Member, IEEE, Huan Tang, Jia-Jie Huang, and Yan Hou Abstract—The rise of multi-cloud systems has been spurred. For safety-critical missions, it is important to guarantee their security and reliability. To address trust constraints in a heterogeneous multi-cloud environment, this work proposes a novel scheduling method called matching and multi-round allocation (MMA) to optimize the makespan and total cost for all submitted tasks subject to security and reliability constraints. The method is divided into two phases for task scheduling. The first phase is to find the best matching candidate resources for the tasks to meet their preferential demands including performance, security, and reliability in a multi-cloud environment; the second one iteratively performs multiple rounds of re-allocating to optimize tasks execution time and cost by minimizing the variance of the estimated completion time. The proposed algorithm, the modified cuckoo search (MCS), hybrid chaotic particle search (HCPS), modified artificial bee colony (MABC), max-min, and min-min algorithms are implemented in CloudSim to create simulations. The simulations and experimental results show that our proposed method achieves shorter makespan, lower cost, higher resource utilization, and better trade-off between time and economic cost. It is more stable and efficient. Index Terms—Multi-cloud environment, multi-quality of service (QoS), reliability, security, task scheduling.   List of Symbols n The number of tasks m The number of clouds in a multi-cloud env- ironment h The number of virtual machines (VMs) in a multi-cloud environment N q {1, 2,…, q} N n T = {t i |i∈ } The set of tasks N m C = {c k |k∈ } The set of clouds N h V = {v j |j∈ } The set of VMs J The number of implementation parameters F The number of performance parameters L The number of cost parameters N J I k = {I k, j | j ∈ } The set of implementation security service level objective (SSLO) par- ameters N J I = {I j | j ∈ } The set of implementation SSLO parameter baseline values N F P k = {P k, f | f ∈ } The set of performance SSLO pa- rameters N F P = {P f | f ∈ } The set of performance SSLO pa- rameter baseline values N L A k = {A k, l | l ∈ } The set of cost SSLO parameters N L A = {A l | l ∈ } The set of cost SSLO parameter baseline values w j I The weights of SSLO parameters I j w f P The weights of SSLO parameters P f w l A The weights of SSLO parameters A l N m CSP ={CSP k , k ∈ } The set of cloud service providers (CSPs) b k The deviation baseline level of the CSP k sl k The security level of cloud c k t i The i-th task t i sd The security demand of t i t i ws The workload of t i t ∗ i = { t α i    α ∈ N 5 } The resource demand for CPU, RAM, bandwidth, disk and relia- bility of t i c k The k-th cloud |c k | The number of VMs in c k v j The j-th VM v k, j The j-th VM in cloud c k . v ∗ k, j v k, j The resource attribute of v k, j cp k,j The processing capacity of v k, j cc k,j The computing cost of v ∗ k, j = {v α k, j |α ∈ N 5 } v k, j The attribute of CPU, RAM, band- width, disk, and reliability of D The number of VM types in a multi- cloud environment N D Y = {Y d |d ∈ } The set of VM types N h N D δ(j) The VM type No. of v j , j ∈ , δ(j) ∈ MD(t i , v j ) The matching degree of t i on v j ECT The estimated completion time ECC The estimated completion cost avgECT The average estimated comple- tion time VA The variance of the estimated completion time Manuscript received September 18, 2019; accepted November 4, 2020. This work was supported in part by the National Natural Science Foundation of China (61673123, 61603100), and in part by the Natural Science Foundation of Guangdong Province, China (2020A151501482). Recommended by Associate Editor Peiyun Zhang. (Corresponding author: Qing-Hua Zhu.) Citation: Q.-H. Zhu, H. Tang, J.-J. Huang, and Y. Hou, “Task scheduling for multi-cloud computing subject to security and reliability constraints,” IEEE/CAA J. Autom. Sinica, vol. 8, no. 4, pp. 848–865, Apr. 2021. The authors are with the School of Computer Science and Technology, Guangdong University of Technology, Guangzhou 510006, China (e-mail: zhuqh@gdut.edu.cn; 2111705194@mail2.gdut.edu.cn; huangjj@mail2.gdut. edu.cn; houyan@gdut.edu.cn). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JAS.2021.1003934 848 IEEE/CAA JOURNAL OF AUTOMATICA SINICA, VOL. 8, NO. 4, APRIL 2021