Deep Reinforcement Learning Based Task Offloading Algorithm for Mobile-edge Computing Systems

Mobile-edge computing(MEC) is deemed to a promising paradigm. By deploying high-performance servers on the mobile access network side, MEC can provide auxiliary computing power for mobile devices, greatly reducing the computing pressure of mobile devices and improving the quality of the computing experience. In this paper, we consider the offloading problem of tasks in single-user MEC system. In order to minimize the mean energy consumption of mobile devices and the mean slowdown of tasks in the queue, we propose a deep reinforcement learning(DRL) based task offloading algorithm, and a new reward function is designed, which can guide the algorithm to optimize the trade-off between mean energy consumption and mean slowdown. The simulation results show that the deep reinforcement learning based algorithm outperforms the baseline algorithms.

[1]  Yoshua Bengio,et al.  Deep Sparse Rectifier Neural Networks , 2011, AISTATS.

[2]  Qimei Cui,et al.  An energy-optimal offloading algorithm of mobile computing based on HetNets , 2015, 2015 International Conference on Connected Vehicles and Expo (ICCVE).

[3]  Mohamed Kamoun,et al.  Energy-optimal resource scheduling and computation offloading in small cell networks , 2015, 2015 22nd International Conference on Telecommunications (ICT).

[4]  Khaled Ben Letaief,et al.  Delay-optimal computation task scheduling for mobile-edge computing systems , 2016, 2016 IEEE International Symposium on Information Theory (ISIT).

[5]  Yishay Mansour,et al.  Policy Gradient Methods for Reinforcement Learning with Function Approximation , 1999, NIPS.

[6]  Antonio Pascual-Iserte,et al.  Optimization of Radio and Computational Resources for Energy Efficiency in Latency-Constrained Application Offloading , 2014, IEEE Transactions on Vehicular Technology.

[7]  Hui Tian,et al.  Fine-granularity based application offloading policy in cloud-enhanced small cell networks , 2016, 2016 IEEE International Conference on Communications Workshops (ICC).

[8]  Dusit Niyato,et al.  A Dynamic Offloading Algorithm for Mobile Computing , 2012, IEEE Transactions on Wireless Communications.

[9]  Richard S. Sutton,et al.  Reinforcement Learning: An Introduction , 1998, IEEE Trans. Neural Networks.

[10]  Mohamed Kamoun,et al.  Joint resource allocation and offloading strategies in cloud enabled cellular networks , 2015, 2015 IEEE International Conference on Communications (ICC).

[11]  Srikanth Kandula,et al.  Resource Management with Deep Reinforcement Learning , 2016, HotNets.

[12]  Khaled Ben Letaief,et al.  Joint Task Offloading Scheduling and Transmit Power Allocation for Mobile-Edge Computing Systems , 2017, 2017 IEEE Wireless Communications and Networking Conference (WCNC).

[13]  Kevin L. Priddy,et al.  Artificial neural networks - an introduction , 2005, Tutorial text series.

[14]  Massoud Pedram,et al.  A semi-Markovian decision process based control method for offloading tasks from mobile devices to the cloud , 2013, 2013 IEEE Global Communications Conference (GLOBECOM).

[15]  Min Sheng,et al.  Mobile-Edge Computing: Partial Computation Offloading Using Dynamic Voltage Scaling , 2016, IEEE Transactions on Communications.