Energy-efficient and delay-aware offloading scheme using D2D-enabled mobile edge computing / Ramtin Ranji

Ramtin , Ranji (2019) Energy-efficient and delay-aware offloading scheme using D2D-enabled mobile edge computing / Ramtin Ranji. Masters thesis, University of Malaya.

[img] PDF (The Candidate's Agreement)
Restricted to Repository staff only

Download (219Kb)
    PDF (Thesis M.A)
    Download (1302Kb) | Preview


      Energy efficient operation of mobile/Internet of Things (IoT) devices is a major challenge due to the limited capacity of their batteries. Also, because of their limited processing power, many of them cannot perform computationally intensive applications like face recognition in a timely manner. Device to Device (D2D) communication and Mobile Edge Computing (MEC) are two technologies to mitigate these limitations by offloading the computationally intensive tasks. With D2D, mobile/IoT devices can cooperate directly without the intervention of the Base Station. MEC provides computing services at the edge of network, that is close to the user. Most of the current studies on the energy efficient offloading, put the focus either on the offloading to the edge server, or to a device in the proximity. The problem of MEC solution is scalability, because, edge servers would be overloaded in dense networks. On the other hand, to find a proper offloading destination in D2D networks, devices must consume excessive amount of energy. Recently, a few numbers of integrated schemes were proposed to mitigate those problems. However, there is lack of study to propose both MEC and D2D, as the target of offloading tasks for execution while considering the energy required for offloading and its delay. In this work, we study Energy-Efficient and Delay-aware Offloading Scheme (EEDOS). In the proposed scheme, energy constraint devices and those with low computational power, have two options to offload their work. They can either use MEC, or D2D, and the computational power of the edge server is leveraged to find a proper candidate. For EEDOS network topology, we integrate D2D communication capability in the user layer of mobile networks, so that, mobile users can communicate with MEC layer. The research problem is formulated with consideration of the required energy for task offloading and completing the task execution, under the required deadline. The EEDOS, MEC and D2D offloading schemes have been simulated to evaluate the proposed scheme and to validate the findings with the existing schemes. The numerical results showed that EEDOS was capable to save the energy of mobile devices up to 95 % in comparison to local task execution, as well as reducing the execution delay. The proposed EEDOS, outperformed the existing schemes in terms of task offloading energy consumption and execution delay. This is due to the integration of the computational capability of MEC and idle devices in the network. The resource-limited mobile devices can save more energy, because, the proposed EEDOS, used edge servers to find the proper offloading destination and took into account the high computational power of edge servers and computational resources of large number of idle devices in the network. In this scheme, the load on the edge server was decreased dramatically in comparison to current MEC offloading schemes, because of participating idle devices in the network through D2D communication.

      Item Type: Thesis (Masters)
      Additional Information: Dissertation (M.A.) – Faculty of Computer Science & Information Technology, University of Malaya, 2019.
      Uncontrolled Keywords: D2D communication; Network topology; Mobile/IoT devices; EEDOS; Server; High computational power
      Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
      Q Science > QA Mathematics > QA76 Computer software
      Divisions: Faculty of Computer Science & Information Technology
      Depositing User: Mr Mohd Safri Tahir
      Date Deposited: 06 Jan 2021 06:55
      Last Modified: 06 Jan 2021 06:55

      Actions (For repository staff only : Login required)

      View Item