CUDA-accelerated task scheduling in vehicular clouds with opportunistically available V2I

In this paper, we consider the use of CUDA-based Graphics Processing Units (GPUs) as high performance parallel computing for the purpose of accelerating the application task scheduling in Vehicular Cloud Computing (VCC) systems. We leverage the Single Instruction Multiple Data (SIMD) mode in General-Purpose Graphic Processing Units (GPGPUs) to solve the value iteration algorithm of the defined Markov Decision Process (MDP) of task scheduling on real VCC. We consider opportunistically available Vehicle to Infrastructure (V2I) communication in Dedicated Short Range Communication (DSRC) used in Vehicular ad hoc networks (VANETs) for the vehicular clouds.

[1]  Xuemin Shen,et al.  An SMDP-Based Resource Allocation in Vehicular Cloud Computing Systems , 2015, IEEE Transactions on Industrial Electronics.

[2]  Sergio Ruiz,et al.  A Parallel Solver for Markov Decision Process in Crowd Simulations , 2015, 2015 Fourteenth Mexican International Conference on Artificial Intelligence (MICAI).

[3]  Mario Gerla,et al.  Vehicular Cloud Computing , 2012, 2012 The 11th Annual Mediterranean Ad Hoc Networking Workshop (Med-Hoc-Net).

[4]  Samy El-Tawab,et al.  Towards Fault-Tolerant Job Assignment in Vehicular Cloud , 2015, 2015 IEEE International Conference on Services Computing.

[5]  Mohsen Guizani,et al.  RSU cloud and its resource management in support of enhanced vehicular applications , 2014, 2014 IEEE Globecom Workshops (GC Wkshps).

[6]  Ashok Kumar Turuk,et al.  Application of greedy algorithms to Virtual Machine Distribution across Data Centers , 2014, 2014 Annual IEEE India Conference (INDICON).

[7]  Rong Yu,et al.  Toward cloud-based vehicular networks with efficient resource management , 2013, IEEE Network.

[8]  Bryan C. Ward.  Relaxing Resource-Sharing Constraints for Improved Hardware Management and Schedulability , 2015, 2015 IEEE Real-Time Systems Symposium.

[9]  Ronald Duarte,et al.  On the performance and energy-efficiency of multi-core SIMD CPUs and CUDA-enabled GPUs , 2013, 2013 IEEE International Symposium on Workload Characterization (IISWC).

[10]  Mohsen Guizani,et al.  Advanced Activity-Aware Multi-Channel Operations1609.4 in VANETs for Vehicular Clouds , 2016, 2016 IEEE Global Communications Conference (GLOBECOM).

[11]  John Nickolls.  GPU parallel computing architecture and CUDA programming model , 2007, 2007 IEEE Hot Chips 19 Symposium (HCS).

[12]  Rupak Biswas,et al.  The impact of hyper-threading on processor resource utilization in production applications , 2011, 2011 18th International Conference on High Performance Computing.

[13]  Silvia Giordano,et al.  The Next Paradigm Shift: From Vehicular Networks to Vehicular Clouds , 2013 .

[14]  Nicholas Wilt,et al.  The CUDA Handbook: A Comprehensive Guide to GPU Programming , 2013 .

[15]  Roberto Rojas-Cessa,et al.  Greedy scheduling of tasks with time constraints for energy-efficient cloud-computing data centers , 2015, Journal of Cloud Computing.