Task Offloading and Resource Allocation in Mobile- Edge Computing System

 

ABSTRACT

 

Mobile-edge computing (MEC) system is a new par- adigm to provide cloud computing capacities at the edge of radio access network (RAN) which is close to mobile users. In this paper, we aim to promote QoS by offloading the computationally inten- sive tasks to the MEC server. There are many papers discuss this issue. Nevertheless, most of them just think over one-dimension re- source allocation, radio resources or computation resources, and make the MEC system less effective. Hence, we consider the allo- cation of both radio resources and computation resources of the MEC server to increase system effectiveness. Apart from this, we take the variety of tasks’ requirements into account. That is, we assume that different tasks may have different delay requirements. We formulate this problem as a cost minimization problem and design a heuristic algorithm to address it. Numerical results show that our algorithm can greatly promote QoS.

Keywords—mobile-edge computing, computation offloading

 

 

 

 

EXISTING SYSTEM:

 

As mobile devices become more and more popular, there are lots of mobile applications emerging and attracting great atten- tion. Most of these applications such as face recognition, inter- active gaming and augmented reality require intensive computa- tion capacity and high energy consumption. However, mobile devices are typically resources-constrained. Their computation resources and battery life are limited due to their small physical size. Mobile Cloud Computing (MCC) is the traditional archi- tecture to address these challenges. But offloading tasks to the remote cloud would always cause large delay. Consequently, Mobile-Edge Computing (MEC) is a new paradigm proposed to address this challenge as well as meet the delay requirements of applications. For instance, in, the authors assumed that a task can be further divided into some small subtasks. Thus, mobile devices can offload a portion of subtasks of a task and leave the others executing locally. They designed a mechanism to minimize the weighted sum of devices’ energy consumption as well as keep the execution delay lower than the delay require- ment. introduced 5G heterogeneous networks into the MEC system. In this network, there is a Macro Base Station (MBS) equipped with an MEC server and a Small Base Station (SBS). The service areas of these two stations are overlaid. Hence, mobile devices can offload tasks to not only the MBS but also the SBS. However, only the MBS is equipped with the MEC server, so the tasks offloaded to the SBS should be further transmitted to the MBS incurring backhaul transmission delay. The authors proposed an energy-efficiency mechanism with joint optimization of offloading decision and communication resource allocation. Nonetheless, radio resources between de- vices and the MEC server and computation resources of the MEC server are both limited. Therefore, allocation of these two types of resources must be included in our discussion. Introduced virtualization of the resources on BBU pool. Then, the authors divided the virtualized resources into two sets, baseband processing and Fog computing. Several experiments were performed on a General Purpose Processor (GPP) in [4]. The authors in considered a multiple F-RAN nodes scenario.

 

PROPOSED SYSTEM:

 

In this section, we will introduce our multiuse MEC system and formulate the problem. Further, we will propose our heuris- tic algorithm to address this problem. A. System Model We consider a multiuser MEC scenario shown in Fig. 1. In this scenario, there is a single wireless base station (BS) equipped with a MEC server with limited computation capacity 􀜨􀮼 and this BS serves N mobile devices. Each device 􀝊 with re- maining energy 􀜧􀯡􀯋 has one task 􀜮􀯡, which can be described in terms of input data size 􀜽􀯡, CPU cycles required 􀝀􀯡 and delay tolerance 􀜶􀯡􀯠􀯔 . Each task can be executed locally or offloaded to the MEC server. 1) Local Execution: We denote computation capacity of device n as 􀝂􀯖 (cycles/s) and 􀜶􀯡 􀯟 is the local task completion time which can be written as 􀜶􀯡 􀯟 􀵌 􀝀􀯡/􀝂􀯡

 

 

CONCLUSION:

 

In this paper, we formulate the multiuser offloading decision and two types of resource allocation problem as a cost minimization problem and proposed a heuristic algorithm considering both radio and computation resource allocation. In addition, we also discuss the variety of delay requirements of tasks in spite of only thinking over the average delay. By taking different delay requirements and two types of resource allocation into account, our algorithm can gain more effectiveness and be closer to reality. In the simulation, we evaluated the effectiveness of our pro- posed algorithm by comparing with the other two schemes, local execution only and remote execution only. Numerical results show that our proposed algorithm outperforms other two schemes and greatly promote QoS by taking two types of re- source allocation into account at the same time. Furthermore, we show that our algorithm leads to the lower value of the total cost than the other two schemes under any condition of the number of devices

 

 

 

REFERENCES:

 

[1] Y. Jararweh, A. Doulat, O. AlQudah, E. Ahmed, M. Al-Ayyoub and E. Benkhelifa, “The future of mobile cloud computing: Integrating cloudlets and Mobile Edge Computing”, 2016 23rd International Conference on Telecommunications (ICT), 2016.

 

[2] C. You, K. Huang, H. Chae and B. Kim, “Energy-Efficient Resource Allocation for Mobile-Edge Computation Offloading”, IEEE Transactions on Wireless Communications, vol. 16, no. 3, pp. 1397-1411, 2017.

 

[3] K. Zhang, Y. Mao, S. Leng, Q. Zhao, L. Li, X. Peng, L. Pan, S. Maharjan and Y. Zhang, “Energy-Efficient Offloading for Mobile Edge Computing in 5G Heterogeneous Networks”, IEEE Access, vol. 4, pp. 5896-5907, 2016.

 

[4] Y. Ku, D. Lin, C. Lee, P. Hsieh, H. Wei, C. Chou and A. Pang, “5G Radio Access Network Design with the Fog Paradigm: Confluence of Communications and Computing”, IEEE Communications Magazine, vol. 55, no. 4, pp. 46-52, 2017.

 

[5] Y.-Y. Shih, W.-H. Chung, A.-C. Pang, T.-C. Chiu, and H.-Y. Wei, “Enabling low-latency applications in fog-radio access networks”, IEEE Network, vol. 31, no. 1, pp. 52-58, 2017.

 

[6] M. Chen, B. Liang, “Joint offloading decision and resource allocation for multi-user multi-task mobile cloud”, 2016 IEEE International Conference on Communications (ICC), 2016. [7] X. Chen, L. Jiao, W. Li and X. Fu, “Efficient Multi-User Computation Offloading for Mobile-Edge Cloud Computing”, IEEE/ACM Transactions on Networking, vol. 24, no. 5, pp. 2795-2808, 2016.

 

[8] X. Lyu, H. Tian, C. Sengul and P. Zhang, “Multiuser Joint Task Offloading and Resource Optimization in Proximate Clouds”, IEEE Transactions on Vehicular Technology, vol. 66, no. 4, pp. 3435-3447, 2017.

 

[9] J. Liu, Y. Mao, J. Zhang and K. Letaief, “Delay-optimal computation task scheduling for mobile-edge computing systems”, 2016 IEEE International Symposium on Information Theory (ISIT), 2016