The Future of Mobile Cloud Computing: Integrating ... - IEEE Xplore

2016 23rd International Conference on Telecommunications (ICT)

The Future of Mobile Cloud Computing: Integrating Cloudlets and Mobile Edge Computing Yaser Jararweh,1 Ahmad Doulat,1 Omar AlQudah,1 Ejaz Ahmed,2 Mahmoud Al-Ayyoub,1 and Elhadj Benkhelifa3 1 Jordan University of Science and Technology, Irbid, Jordan 2 University of Malaya, Kuala Lumpur, Malaysia 3 Staffordshire University, Stafford, UK

Abstract—Extending the coverage area of mobile cloud computing services will allow new services to be provisioned to the mobile users. The main obstacle for achieving this goal is related to the deployments challenges and limitations of the Cloudlets system. Mobile Edge Computing (MEC) system emerged recently providing an opportunity to fill the gap of the Cloudlets system by providing resources-rich computing resources with proximity to the end users. In this paper, we are proposing a hierarchical model that is composed of MEC servers and Cloudlets infrastructures. The objective of the proposed model is to increase the coverage area for the mobile users in which the users can accomplish their requested services with minimal costs in terms of power and delay. An extensive experimental evaluation is conducted showing the superiority of the proposed model.

I. I NTRODUCTION Next generation Mobile Cloud Computing (MCC) systems will require a paradigm shift in how they are constructed and managed. Current deployment and management platforms are facing considerable challenges regarding flexibility, coverage, dependability and security that next generation systems must handle. Recently, Mobile Edge Computing (MEC) technology emerged as a viable solution to provide services to the mobile users within their access range [1, 2]. This enables a seamless access to a resources-rich system with high bandwidth network connections. MEC is perceived as a natural evolution to the previously emerged technologies for deploying mobile cloud services such as Cloudlets [3]. Cloudlets provide cloud services to the mobile users within the coverage of their Wi-Fi access point. The increasing demand of the mobile applications for high computing and storage capacities with free user mobility made Cloudlets an inefficient solution for end user workload offloading. This is due to the limited resource capabilities of the Cloudlet servers and the short coverage range of the Cloudlet Wi-Fi connection [2, 4]. The seamless mobile task offloading and execution is crucial to the application latency and the quality of services provided to the users [5]. Many new applications can benefit from the MEC system such as compute-intensive video encoding [6] and local streaming service [7]. These The authors would like to thank the Deanship of Research at the Jordan University of Science and Technology for funding this work, grant number 20150050.

978-1-5090-1990-8/16/$31.00 ©2016 IEEE

applications argue that MEC systems provide the required seamless task execution on the mobile cloud system. The proposed system in this paper is an attempt to integrate the Cloudlet capabilities into the MEC paradigm. This is achieved by integrating a large scale Cloudlet system [8, 9], deployed in proximity to the users, with the MEC servers, deployed at the base stations of the mobile network. The following two points summarize the limitations of Cloudlets (as discussed in [2]) which serve as motivations. 1) The coverage regions of Cloudlets are typically very small as they rely on Wi-Fi accessibility. The need for continuous access to computing resources in wide regions including those not covered by Wi-Fi limits the use of Cloudlets. 2) According to [2], Cloudlets have serious issues when it comes to scalability. This is not limited to the provided services, but also to the provisioning of the resources. This second limitation contradicts the core advantage of modern mobile computing systems which are intended to be scalable in order to be able to serve the increasing demands of their large numbers of users. Hence, we, in our system, adopt the mobile edge computing (MEC) paradigm to overcome the above mentioned limitations of the Cloudlets. The rest of this paper is structured as follows. The mobile edge computing system is presented in section II. Section III presents and discusses the proposed integrated model for mobile edge computing architecture with Cloudlets. The proposed system prototype and experimental evaluation are then presented in Section IV. Finally, we conclude and present our future plans in Section V. II. M OBILE E DGE C OMPUTING Mobile Edge Computing (MEC) promises a paradigm shift in enabling efficient Mobile Cloud Computing (MCC) services by providing storage and processing capabilities within the access range of the mobile devices. In MEC, Mobile Edge (ME) servers are placed at the edge of the mobile networks eliminating the need to offload compute-/storageintensive tasks of the mobile devices to the core of the network (the centralized cloud data center). This reduces the network latency and enhances the quality of service provided


for the mobile end users. Different applications can benefit from the large scale deployments of ME servers such as smart grid applications, content delivery networks, content sharing, traffic management, and E-health applications [10]. This promising paradigm comes with its own downside related to the management complexity of large scale deployments that offers hundreds of applications to millions of users. In this paper, we introduce a hierarchical framework to enable efficient MCC services through the integration of different system components with the MEC system. MEC comes to accelerate applications and data streaming in mobile networks through caching and/or providing the required processing capabilities at the edge of the mobile network, i.e., as close as possible to the mobile end user. In MEC, instead of relying on a centralized platform to serve all mobile devices, a decentralization platform is suggested. This can be achieved by moving cloud servers to the edge of the mobile network. According to [11], this platform was first proposed by Akamai Technologies as they describe the topology of their content delivery network. This topology was used to cache some of the frequently requested contents at the network edge. Recent studies proposed integrating MEC with cloud computing principles to provide more complex services at the edge of the network aiming at (i) reducing the load on the centralized cloud and (ii) avoiding bottlenecks and single points of failure. Traditionally, devices at the mobile network edge merely represent mobile access points responsible for forwarding traffic coming from base stations almost without any analysis and/or intervention of the user requests. MEC introduces new network elements at the network edge that provides computational and storage capabilities. Therefore, these new devices can serve user requests locally instead of just forwarding these requests. In the rest of this paper, these devices are referred to as MEC servers. Typically, a MEC system can provide lowlatency on-premises services to its users by accessing local resources. It also facilitates users location awareness. According to [12], there are many proposed sites for the deployment of ME servers. One examples is the LTE base station, which is referred to as evolved NodeB (eNB). Other potential locations include the 3G Radio Network Controller (RNC) site and the multi-technology (3G/LTE) cell aggregation site. The last option, which is the multi-technology (3G/LTE) cell aggregation site, can be located indoor within an enterprise (e.g., a hospital, the headquarters of a large corporation, etc.), or indoor/outdoor for a special public coverage scenario (e.g., a stadium, a shopping mall, etc.) in order to control the number of local multi-technology (3G/LTE) access points providing radio coverage on-premises. Figure 1 shows a possible MEC deployment. Applications that may incorporate MEC include smart grids, smart transportation or smart traffic, video streaming, mobile big data analytics, mobile gaming, edge health care, and sensor networks application.

Fig. 1. A possible MEC framework

III. T HE P ROPOSED MEC-C LOUDLET I NTEGRATED A RCHITECTURE As stated in [13, 14], one of the main objectives of deploying large scale Mobile Cloud Computing (MCC) systems using the Cloudlet based infrastructure is to reduce the total power consumption and the network delay while satisfying the Service Level Agreement (SLA). The limited capabilities of the Cloudlet system makes this objective challenging to achieve as they limit the number of available services on each Cloudlet system. Such limited capabilities will force the cooperatation between different Cloudlets in order to meet the users demands for different types of services. Moreover, the cooperation might require moving some user requests from the local Cloudlet to a remote Cloudlet that provide the requested service using a backbone network. In this paper, we consider large scale cooperative Cloudlets deployments and propose a model to address the problem of power consumption optimization for such systems. Being evaluated under different realistic scenarios, the results show that the proposed model can be used to accurately optimize power consumption in large scale MCC systems. Figure 2 illustrates the general framework architecture of the proposed MEC-Cloudlet System. The proposed architecture model has the following components: 1) Mobile Users. 2) Cloudlets Subsystem. 3) Mobile Edge Computing Subsystem. 4) The core cloud system. The MEC network consists of several domain environments. Each domain is controlled by a Cloudlet controller which is responsible for keeping smooth communication between the entities inside the same local domain. This Cloudlet controller is also needed because all the local resources in the domain should be manged and a decision whether a local request


Fig. 2. A hybrid Mobile Cloud Computing (MCC) framework

should be services locally or should be forwarded to the cloud is taken by this controller. Nevertheless, all the decisions which requires a higher level of control are enforced by a MEC controller. The main responsibility of the global centralized controller is the controlling of all the local controllers and ensuring that all the entities in the MEC network work in an efficient way. Moreover, any critical and unsolved issue faced by the local controller is transferred to the MEC controller where the proper action is taken. A. The advantages of the MEC-Cloudlet integrated Model The Cloudlet/MEC layering system has many advantages that can be summarized as follows. •

•

•

Reducing the overhead on the MEC controller: The overall control load is distributed among the local controllers and only the main critical issues are handled by the MEC controller. Taking real time decisions: When each local controller serves its nearby local entities, then the request does not need to go to the MEC controller, and, thus, the response time is reduced. Avoiding the single point of failure problem: The idea of the distribution of control load to the local controllers reduces the risk of single point of failure.

•

•

•

Reducing the delay: When the requests are served locally, the system performance is enhanced by reducing the delay. Scalability: Integrating several local controllers facilitates the process of expansion of the network and increases the number of supported domains. Reducing the global network traffic: The traffic is reduced through reducing global communication. This is achieved since only the main requests are transferred to the MEC controller. IV. E XPERIMENTAL R ESULTS AND E VALUATION

As a proof of concept, we evaluate our framework through two experiments with varying numbers of requests (5000, 10000, 20000, and 50000 ) and fixed numbers of both Cloudlets (i.e. 400) and MECs (i.e.200). Our evaluation metrics are total number of served request per each layer, total delay per layer and the total power consumption per layer. Four simulation scenarios are used. for each scenario, we change randomly the percentage of requests per each layer (P2P, Cloudlet, MEC and Core Cloud) in our framework. The objective of theses experiments is to show how the amount of offloaded tasks from the mobile devices to the upper layer will impact the total power consumption and the incurred delay on the system.


1XPEHURI8VHUV

33 &ORXGOHW 0(& &RUH&ORXG

1XPEHURI5HTXHVWV

Fig. 3. Total Number of Served Request per each Layer

&RPPXQLFDWLRQ7LPH6HF



1XPEHURI5HTXHVWV

Fig. 5. Total Processing Time per each Layer (Sec).

3URFHVVLQJ7LPH6HF

Figure 3 shows the total number of served requests per each layer with varying mobile workloads. Its obvious that the MEC layer is serving most of the requests. An interesting finding from Figure 3 is the poor performance of the Cloudlet layer which support the claim of Cloudlet limitations. The share of the core cloud is showing an increase with increasing number of requests to support the ubiquitous coverage for the services. Both Figure 4 and Figure 5 show the total processing time and communication time needed by the system with different scenarios. It is obvious that increasing the percentage of requests fulfilled within the mobile devices will reduce the total power consumption in contrast to the scenarios with requests that need the core cloud to fulfill.

1XPEHURI5HTXHVWV

Fig. 4. Total Communication Time per each Layer (Sec).

V. C ONCLUSION AND F UTURE W ORK This paper argued that the emergence of MEC is an inevitable result of the paradigm shift from traditional MCC models to ubiquitous MCC. MEC provides the ability to fulfill the computing and storage needs of mobile end users with a

resource-rich ME servers available at the access point of the mobile networks, which allows a secure, efficient, and low latency execution. It also enables aggregating the resources across multiple ME servers to handle large-scale applications and bursty data rates. This increases the complexity of resource management for the various workloads generated form different applications with different processing requirements. There is also a need to automate this resource management in order to enable precise resource provisioning for individual applications locally or globally. R EFERENCES [1] M. T. Beck, M. Werner, S. Feld, and S. Schimper, “Mobile edge computing: A taxonomy,” 2014. [2] A. Ahmed and E. Ahmed, “A survey on mobile edge computing,” in 10th IEEE International Conference on Intelligent Systems and Control, (ISCO 2016). IEEE, 2016. [3] M. Satyanarayanan, P. Bahl, R. Caceres, and N. Davies, “The case for vm-based cloudlets in mobile computing,” Pervasive Computing, IEEE, vol. 8, no. 4, pp. 14–23, Oct 2009. [4] U. Shaukat, E. Ahmed, Z. Anwar, and F. Xia, “Cloudlet deployment in local wireless networks: Motivation, architectures, applications, and open challenges,” Journal of Network and Computer Applications, vol. 62, pp. 18 – 40, 2016. [5] E. Ahmed, A. Gani, M. K. Khan, R. Buyya, and S. U. Khan, “Seamless application execution in mobile cloud computing: Motivation, taxonomy, and open challenges,” Journal of Network and Computer Applications, vol. 52, pp. 154 – 172, 2015. [6] M. Beck, S. Feld, A. Fichtner, C. Linnhoff-Popien, and T. Schimper, “Me-volte: Network functions for energyefficient video transcoding at the mobile edge,” in Intelligence in Next Generation Networks (ICIN), 2015 18th International Conference on, Feb 2015, pp. 38–44.


[7] O. Makinen, “Streaming at the edge: Local service concepts utilizing mobile edge computing,” in Next Generation Mobile Applications, Services and Technologies, 2015 9th International Conference on, Sept 2015, pp. 1–6. [8] L. Tawalbeh, Y. Jararweh, F. Dosari et al., “Large scale cloudlets deployment for efficient mobile cloud computing,” Journal of Networks, vol. 10, no. 01, pp. 70–76, 2015. [9] Y. Jararweh, F. Ababneh, A. Khreishah, F. Dosari et al., “Scalable cloudlet-based mobile computing model,” Procedia Computer Science, vol. 34, pp. 434–441, 2014. [10] M. Quwaider and Y. Jararweh, “A cloud supported model for efficient community health awareness,” Pervasive and Mobile Computing, pp. –, 2015. [11] A. Davis, J. Parikh, and W. E. Weihl, “Edgecomputing: extending enterprise applications to the edge of the internet,” in Proceedings of the 13th international World Wide Web conference on Alternate track papers & posters. ACM, 2004, pp. 180–187. [12] M. Patel, B. Naughton, C. Chan, N. Sprecher, S. Abeta, and A. Neal, “Mobile-edge computing introductory technical white paper,” White Paper, Mobile-edge Computing (MEC) industry initiative, 2014. [13] M. Al-Ayyoub, Y. Jararweh, L. Tawalbeh, E. Benkhelifa, and A. Basalamah, “Power optimization of large scale mobile cloud computing systems,” in Future Internet of Things and Cloud (FiCloud), 2015 3rd International Conference on. IEEE, 2015, pp. 670–674. [14] Q. Althebyan, Q. Yaseen, Y. Jararweh, and M. AlAyyoub, “Cloud support for large scale e-healthcare systems,” Annals of Telecommunications, pp. 1–13, 2016.