International Journal of Education and Development using Information and Communication Technology (IJEDICT), 2018, Vol. 14, Issue 1, pp. 91-102 Feasibility of cloud computing implementation for eLearning in secondary schools in Tanzania Kennedy F. Mwakisole, Mussa M. Kissaka, and Joel S. Mtebe University of Dar es Salaam, Tanzania ABSTRACT This article assessed the feasibility of implementing eLearning systems in a cloud-based infrastructure for secondary schools in Tanzania. The study adopted questionnaire and document reviews as data collection tools. A total of 820 students successfully returned the questionnaire from seven secondary schools in Tanzania. The study found that 11% of government secondary schools had computer labs with 20.1% of them connected to the Internet. Moreover, more than half of surveyed students (56.6%) had access to mobile phones at home with 53.5% using the phones to access the Internet. However, the study found that the cost of the Internet had remained unaffordable to many students. This study shows that eLearning implementation in the cloud for secondary schools in Tanzania is feasible. This research will serve as a base for future studies that plan to implement eLearning systems to enhance teaching and learning in secondary schools in Tanzania. Keywords: Cloud computing, eLearning, Internet, Secondary schools, eLearning Tanzania INTRODUCTION In the recent years, there has been an increasing interest in the development of eLearning systems to enhance the quality of teaching and learning in secondary schools in Tanzania. This was due to continued improvement of Information and Communication Technologies (ICT) infrastructure and penetration of mobile phones in Tanzania. The cost of Internet service has declined by more than 50% (MWTC, 2016) due to rolling out of two of the optical fibre network namely Eastern Africa Submarine Cable System, and the African Cable System (Mtebe, 2015). The cost of mobile devices such as smartphones and tablets has dropped to as low as US$ 30; and are affordable to the majority of Tanzanians (Mtebe & Kondoro, 2016). Additionally, Internet users have increased from 29% in 2014 to 40% in 2016 while mobile phone penetration has increased to 80% of the total population by 2017 (TCRA, 2017). Given these developments, the Government of Tanzania and development partners have been equipping schools with ICT facilities to improve the quality of education in secondary schools in Tanzania. A recent report by the Ministry of Education, Science and Technology (MoEST), indicates that approximately 31.4% of government secondary schools (out of 3,601) have computers ranging from 1 to 68 computers (MoEST, 2017). The report further states that 20.1% these schools are connected to the Internet. Although the report did not focus on private owned secondary schools, it is estimated that the number of private schools with computers connected to the Internet is high. Apart from ICT facilities, there exist some initiatives to develop eLearning systems in order to facilitate sharing of digital content and provide interaction between teachers and students synchronously and asynchronously via the Internet. The first effort to implement eLearning system can be traced back in 2006 when an eLearning system was developed and piloted in two secondary schools in Kibaha and Dar es Salaam (Kalinga, Bagile and Trojer, 2006). Since then, a number of similar systems such as Retooling, Shuledirect, Halostudy, and Brainshare have been developed and implemented. Shuledirect ,for instance, consist of 8 subjects benefiting more than 10,000 92 IJEDICT students countrywide (Mtebe & Kissaka, 2015). Similarly, Halostudy has multimedia-enhanced content of science and mathematics subjects from Form I to Form IV. More than 427 schools connected to the Internet by Halotel have been accessing the content benefiting more than 50,000 students in Tanzania. The majority of implemented eLearning systems are hosted in-house in schools or in providers’ computer servers surrounded by several challenges. One of these challenges is poor accessibility of these systems due to unpredictably power interruptions. The majority of computer servers hosting these systems tend to be off at some point and therefore affecting users accessing the systems. Moreover, schools are burdened with hiring technical staff who are responsible for maintaining and managing computer servers at their premises. Generally, hosting eLearning systems at schools premises is unreliable and costly. Mtebe and Raisamo (2014) computed the cost of hosting an eLearning system at a school premise and found that the school needs to invest approximately US$ 25,441 to host eLearning system for 3 years. Similarly, the University of Education, Winneba, in Ghana spent an estimated amount of US$ 300,000 per year to implement eLearning system to service 15,000 students (Unwin et al., 2010). Therefore, hosting eLearning systems in-house requires substantial investment that many secondary schools cannot afford. However, cloud-computing technology that has been adopted and implemented worldwide to overcome these challenges has not been explored. Cloud computing comprises virtualized servers, networks, database storage, applications and services which are delivered over the Internet. In this computing model users rent and consume computing and storage resources as needed and pay per usage similar to water and electric bills (Laisheng & Zhengxia, 2011; Carroll et al., 2011). By gaining support from largest ICT companies such as Google, Amazon, and Microsoft, cloud computing is being widely embraced by many organizations (González-Martínez et al., 2015). In education, the adoption of cloud computing will enable schools to implement eLearning systems without procuring and hosting ICT infrastructure in their premises. By using this approach schools can save substantial capital costs for purchasing hardware and software, administration and operational costs associated with hardware maintenance, software licensing, electric power, cooling system and wages for ICT personnel (Mokhtar et al., 2013; Sultan, 2010; Carroll et al., 2011). As a result, schools will concentrate on improving students learning rather than managing ICT infrastructure and services in their premises (Chandra & Borah, 2012). Schools will also obtain substantial cost savings by paying only services they use due to pay-as-you-use pricing mechanism offered by cloud computing firms. It should be noted that cloud computing is Internet based technology; and its success depends on readiness of schools, students, and teachers to have access to the infrastructure that enable them to access Internet services. Therefore, this article aimed at assessing the feasibility of implementing eLearning system in cloud-based infrastructure for secondary schools in Tanzania. The study adopted questionnaire and document review data collection methods. A total of 820 students successfully returned the questionnaires from seven secondary schools in Tanzania. LITERATURE REVIEW According to the National Institute of Standards and Technology (NIST), “cloud computing is a distributed computing paradigm that enables access to virtualized resources including computers, networks, storage, development platforms or applications via the Internet” (Mell & Grance, 2011). In terms of ICT resources proprietorship, cloud computing technology can have four types of deployment models namely public clouds, private clouds, hybrid clouds, and community clouds (Jin, et al., 2010). A private cloud model enables organization to have full control of the cloud underlying Feasibility of cloud computing implementation for eLearning 93 infrastructure, data, applications, services, and resources that are provided to their users. It may be hosted on the premises of an organization or by a third party provider (Despotović-Zrakić, et al., 2013). Community cloud is managed by a limited number of organizations that have shared interests and form a community of practice in which the operations are managed by the community with the distribution of responsibilities (Selviandro, et al., 2014) . A hybrid cloud is a combination of two or more individual clouds (private, community, or public) that remain exclusive entities but are bound together by uniform or proprietary technology that enables data and application movability. Applications that access less sensitive data can be outsourced to the public cloud, while keeping business services and sensitive data in a secured private cloud (Carroll, et al., 2011). Similarly, cloud computing has three types of service models: infrastructure as a service (IaaS), platform as a service (PaaS) and software as a service (SaaS). IaaS aims to deliver over the network computing resources and storage as a service to users. Users install and manage operating systems and software applications on virtualized machines. PaaS model provides an environment for programmers to develop, test and execute software applications via the Internet. SaaS model delivers application software such as eLearning systems to the user’s client software via Internet. This model allows end users to use their browser or client software installed on their mobile phone to access application software that runs on a cloud infrastructure. The users of these services do not control or manage underlying hardware, network and application platforms (González-Martínez et al. , 2015; Bora & Ahmed, 2013; Pocatilu et al., 2010). Several cloud computing service providers support to enhance eLearning in various contexts. The cloud can provide schools with tools to deploy computing resources on-demand for class materials and activities according to their learning needs (González-Martínez, et al., 2015). The benefits of cloud computing require the availability of ICT infrastructure within the schools and readiness of users to use various Internet services. Nonetheless, available studies in Tanzania have been assessing the availability of computers and Internet connectivity without assessing how prepared students are to use Internet services. Malero, Ismail, and Manyilizu (2015), for example, assessed 89 schools in Dodoma region for the readiness of schools in using ICT for teaching and learning. The study found that many teachers and students indicated their willingness to use ICT to support teaching and learning. However, the majority of studied schools did not have computers or Internet. Therefore, the findings were based on their perceptions of use rather than experience of using ICT facilities and equipment for teaching and learning. Similarly, Kafyulilo (2014) explored the access, use and perceptions of teachers and students towards mobile phones as a tool for facilitating teaching and learning using a sample 29 teachers and 40 students from Kibasila secondary school in Tanzania. The study showed that 60 % of students owned mobile phones, or had access to mobile phones; and they were in favour of the use of mobile phones for learning. A similar study was conducted in 10 secondary schools in Kilimanjaro Tanzania with a sample of 294 students (Chambo et al., 2013). The study found that 86.2% of students owned mobile phones with 65.8% of them having access to Internet connectivity. Tarimo and Kavishe (2017) conducted a study to investigate Internet access and usage by secondary school students in Morogoro Tanzania using a sample of 120 students. Interestingly, 82% of students indicated that they were using Internet for searching academic information while 87.6% of them were using it for playing and downloading music. The study did not specify if the selected schools had computer labs with Internet connection. No data was provided to whether students used other devices such mobile phones to access the Internet. Generally, there are limited studies that have investigate the accessibility and usage of Internet services in schools already connected to the Internet in Tanzania. Many studies tend to focus on perceptions of teachers and students on the use of ICT facilities regardless of whether they have 94 IJEDICT access to them or not. While these studies provide baseline for understanding the perceptions of students and teachers in ICT integration in teaching and learning, assessing usage of Internet and its services is important for future plan of cloud computing implementation in secondary schools in Tanzania. Therefore, this study investigated the feasibility of implementing eLearning systems in cloud through assessing usage of Internet and its services in seven secondary schools in Tanzania. METHODOLOGY Study Design The study adopted questionnaire and document reviews as data collection tools. A total of 150 questionnaires were distributed to each school: Benjamin Mkapa, Turiani, Kibasila, Makumbusho, Kambangwa, Chang’ombe, and Kibamba making a total of 1,050 distributed questionnaires. The schools were conveniently selected focusing on government owned schools with computer labs connected to the Internet. The questionnaires were distributed to students from Form I to Form IV. This research design was chosen because it is relatively quick and easy to conduct as they do not need long periods of follow-up and data on all variables that can easily be collected at once. A total of 820 questionnaires were collected making a response rate of 96.5%. The document review was conducted to assess ICT infrastructure in secondary schools and relevant policies to support cloud computing in secondary education. The study was conducted between July and September 2017. Demographic information The study shows that the majority of respondents were from Benjamin Mkapa (21.8%), followed by Turiani (15.9%) and Kibamba (15.0%) secondary school. However, Kibasila secondary school had the smallest number of respondents (10.1%) who participated in the study as shown in Figure 1. Figure 1: Respondents by Schools in Percentage Moreover, the majority of respondents were Form II students followed by Form I and Form VI students. The minority of respondents (5%) were Form V students (13.2%) followed by Form III (13.2%) and Form IV students (14.3%). Figure 2 shows the distribution of respondents by the level of classes. Feasibility of cloud computing implementation for eLearning 95 Figure 2: Distribution of Respondents by Their Class of Study in Percentage FINDINGS Relevant Policies The study reviewed relevant documents to assess the availability of polices supporting cloud computing in education. The study revealed that the government has developed several policies that create conducive environment for cloud computing implementation in secondary education in Tanzania. One of the important policy is the ICT Policy for Basic Education which sets the guidelines desired to transform Tanzania to information and digital driven society (MoEVT, 2007). This was planned to be achieved through the application of ICT in all levels of education. This policy was followed by the Education Training Policy in 2014 that stressed the use and application of ICT in education and training at all levels in order to improve quality education provision. Recently, the government has formulated the National ICT Policy of 2016 replacing the ICT policy of 2003 under the funding of Finnish Government (MWTC, 2016). The policy emphasizes effective integration of ICT in education while calling for increased broadband access and ICT Infrastructure development. ICT Facilities and Basic Infrastructure Documentary reviews and various reports (MoEST, 2017), revealed that out of 3,601 government schools in Tanzania, only 396 (11%) schools have been equipped with computer laboratories. Moreover, approximately 20.1% of government schools are connected to the Internet. The data for private schools could not be established but it is expected that many private schools will have ICT facilities compared to government schools. Use of Internet at Schools Since the surveyed schools were those with computers connected to the Internet, we were interested to know if students use the Internet to access learning materials. It was interesting to find out that about two thirds (65.6%) of respondents indicated that they do not use Internet to access learning materials using computers installed at their schools. On the other hand, more than half of students 96 IJEDICT (58.3%) indicated that they use Internet to access other content that are not related to learning materials as shown in Figure 3. Figure 3: Respondents on Internet Usage at Schools in Percentage Access to Computer and Internet Outside of School Environment We were also interested to find out if students had access to Internet outside school premises. As shown in Figure 4, approximately 42.3% of respondents indicated that they use computers or laptops outside school environment. Out of those students, 39.8% of them used computers or laptops connected to the Internet. The results further showed that 57.7% of the respondents did not have access to computers or laptops outside the school environment. Feasibility of cloud computing implementation for eLearning 97 Figure 4: Students’ Access of Internet Outside of the School Environment in Percentage Access and Use of Smartphones for Studies Regarding access and use of smartphones, the analysis showed that more than half of the respondents (56.6%) indicated that they use smartphones at home while 53.5% used smartphones to access the Internet. The study also revealed that 58.8% of respondents use smartphones to search for learning materials at home as shown in Figure 5. Figure 5: Respondents on Access and Use of Smartphone in Their Studies in Percentage Expenditure on Internet Bundles Figure 6: Students’ Expenditure on the Internet per Week in Percentage 98 IJEDICT Students were asked to indicate how much they spend for buying Internet bundles via smartphones per week. The study found that the majority of students (48%) spend less than Tsh. 1,000 (equivalent to US$ 0.5) to pay for the Internet bundles, while 44% of students spend between Tsh 1,000 and Tsh 2,000 per week. A small number of students spent more than Tsh 2,000 for Internet bundles per week (See Figure 6). Students were further asked to indicate if they use students’ special bundles provided by the majority of mobile phones operators in Tanzania, and 36% of students said YES while more than half of students (63.2%) indicated that they do not use them. Awareness of the Existing eLearning Initiatives There are many eLearning initiatives in Tanzania which have developed content for secondary schools and shared them via various platforms accessible via the Internet (e.g. Shuledirect, Halostudy, etc). Therefore, we were interested to know if students are aware of these platforms and if they have been using them. The study found that more than half of the students (53.8%) are aware of these initiatives, and 46.2% accessed to them at least once (See Figure 7). Figure 7: Awareness of the Existing eLearning Initiatives When they were asked to mention at least one eLearning platform they have been frequently accessing in the last 2 months, the majority of students mentioned Halostudy (61%) followed by Shuledirect (30%), and Elimuyangu (9%). DISCUSSION The effective use of eLearning systems can potentially improve the quality of teaching and learning in secondary schools in Tanzania. Teachers can share and exchange information and knowledge with their students easily using these systems. With this in mind, the government and development partners have been creating conducive environment to enable smooth adoption and implementation of eLearning systems in Tanzania. However, very few schools have managed to implement eLearning systems due to costs as well as lack of reliable ICT infrastructure in school premises. The majority of existing eLearning initiatives such as Halostudy, Shuledirect have been implemented by NGOs or mobile firms. Feasibility of cloud computing implementation for eLearning 99 To enable many schools to be able to implement eLearning the need to find a reliable and costeffective solution is important. Cloud computing, which has been used to implement eLearning systems worldwide, has not been considered. Nonetheless, successfully implementation of cloud computing requires availability of ICT infrastructure and users’ readiness in using Internet services. This study investigated the feasibility of implementing eLearning systems in cloud through assessing ICT infrastructure and usage of Internet services in seven secondary schools in Tanzania. The study found that use of eLearning in the cloud for secondary schools in Tanzania is feasible. The study also found that the majority of existing formulated policies: ICT Policy for Basic Education (2007), Education Training Policy (2014), and the National ICT Policy (2016), support the ICT integration in education and the use of Internet services in education. In addition to these policies, the Government has been capacitating schools with computers and Internet connectivity to increase accessibility of eLearning systems and other Internet based services. The study also found that many students had access to the Internet both at school and at home. The finding are similar to other studies conducted in secondary schools in Dodoma (Malero, et al., 2015), Morogoro (Tarimo & Kavishe, 2017b), and Kilimaro (Chambo, et al., 2013). Despite having access to the Internet, many students use Internet to access content that is not related to learning materials (66%). It seems that many students in secondary schools tend to use Internet for accessing content that is not related to studies. This evident is similar to the finding in Morogoro where Tarimo and Kavishe (2017) found 87.6% of students used Internet for playing and downloading music. A possible explanation for this might be due to the fact that students are not aware of existing eLearning systems with content that can be accessed to enhance their learning activities. For instance, this study found that only 53% of students were aware of these systems. There is a need to promote awareness of the existing eLearning systems such as Halostudy and Shuledirect so that students can use Internet to access these systems rather than downloading music and other content that is not related to learning. The fact that students have access to the Internet and can use various Internet services provide a strong base for implementing eLearning in the cloud for secondary schools in Tanzania. This is to say; even those schools without computer labs students will still use eLearning systems using Internet access at home or elsewhere. Another important finding was that more than half of surveyed students (56.6%) had access to mobile phones at home with (53.5%) of students using mobile phones to access the Internet. This finding corroborate with that of Chambo, et al., (2013) conducted in ten schools in Kilimanjaro which found that 86.2% of students had access to mobile phones; and 65.8% of them students used them to access the Internet.. Similarly, Tarimo and Kavishe's (2017) study of schools in Morogoro found that 97.8% of students had access to the Internet via mobile phones. These findings provide an evidence that many students have access to mobile phones connected to the Internet which provides a strong base for implementing eLearning in the cloud. It seems, therefore, that schools do not need even to have computer labs to be able to implement eLearning systems. The system can be implemented in the cloud and enable students to access learning materials via mobile phones. The study also found that the cost of the Internet was unaffordable to many students. This is evident from the fact that many students (48%) were paying less than Tsh 1000/= (US $ 0.5) for Internet per week. There is a need for the government to subsidise Internet costs, especially those dedicated for learning and teaching in secondary schools. Currently, many mobile firms have special bundles for students to enable them access Internet at a special rate. The findings from this study have shown that 63.2% of students did not use those special bundles provided by mobile firms. The possible reason for this could be that even special bundles are still expensive for secondary school students. 100 IJEDICT CONCLUSION Cloud computing can shape the way eLearning is implemented in secondary schools in developing countries. Once this technology is adopted, it can reduce the burden of purchasing ICT infrastructure for implementing eLearning systems in schools’ premises. Cloud computing resources and storage could be provided by service providers as a service to the schools for hosting their eLearning systems which will be accessible through the Internet anytime anywhere. Therefore, schools need to have supportive infrastructure and Internet to access and use these systems once they are hosted in the cloud infrastructure. This study looked at the feasibility of implementing eLearning in the cloud for secondary schools in Tanzania. The findings from this study have revealed that it is feasible to implement eLearning systems in the cloud for secondary schools, and students can continue accessing learning materials via the Internet. Many students have access to mobile phones and have been using these devices to access the Internet. With more than 31% of schools having computer labs connected to the Internet, this provides a strong base for implementing eLearning in the cloud. The study also found that the Government has formulated relevant policies that support the implementation of cloud computing in secondary schools. Many of the reviewed policies in this study have shown that the Government has been setting up conducive environment for ICT integration in teaching and learning in secondary education. Generally, the implementation of cloud computing is secondary schools in Tanzania is feasible. REFERENCES Bora, U. J. and Ahmed, M. (2013) ‘E-Learning using Cloud Computing’, International Journal of Science and Modern Engineering, (2), pp. 9–13. Carroll, M., Merwe, A. Van Der and Kotzé, P. (2011) ‘Secure cloud computing benefits, risks and controls’, in Information Security South Africa (ISSA). Johannesburg, pp. 1–9. doi: 10.1109/ISSA.2011.6027519. Chambo, F. et al. (2013) ‘Mobile Learning Model for Tanzania Secondary Schools: Case Study of Kilimanjaro Region’, in Pan African International Conference on Information Science, Computing and Telecommunications (2013), pp. 698–701. Chandra, D. G. and Borah, M. D. (2012) ‘Cost benefit analysis of cloud computing in education’, in 2012 International Conference on Computing, Communication and Applications (ICCCA), pp. 1–6. Despotović-Zrakić, M. et al. (2013) ‘Scaffolding Environment for е-Learning through Cloud Computing’, Educ. Technol. Soc. JSTOR, 16(3), pp. 301–314. González-Martínez, J. A. et al. (2015) ‘Cloud computing and education: A state-of-the-art survey’, Computers and Education, 80, pp. 132–151. doi: 10.1016/j.compedu.2014.08.017. Jin, H. et al. (2010) Handbook of Cloud Computing. Edited by A. E. Borko, Furht. Springer. Kafyulilo, A. (2014) ‘Access , use and perceptions of teachers and students towards mobile phones as a tool for teaching and learning in Tanzania’, Education and Information Technologies, 19(1), pp. 115–127. doi: 10.1007/s10639-012-9207-y. Kalinga, a E., Bagile, R. B. B. and Trojer, L. (2006) ‘An Interactive e-Learning Management Feasibility of cloud computing implementation for eLearning 101 System ( e-LMS ): A Solution to Tanzanian Secondary Schools ’ Education’, International Journal of Human and Social Sciences, 1(4), pp. 252–255. Laisheng, X. and Zhengxia, W. (2011) ‘Cloud computing: A new business paradigm for elearning’, in 2011 Third International Conference on Measuring Technology and Mechatronics Automation. Ieee, pp. 716–719. doi: 10.1109/ICMTMA.2011.181. Malero, A., Ismail, A. and Manyilizu, M. (2015) ‘ICT Usage Readiness for Private and Public Secondary Schools in Tanzania , a Case of Dodoma Municipality’, International Journal of Computer Applications, 129(3), pp. 2013–2016. doi: 10.5120/ijca2015906791. Mell, P. and Grance, T. (2011) The NIST Definition of Cloud Computing. USA. Available at: http://csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf (Accessed: 26 April 2013). MoEST (2017) Feasibility study exploring e-learning initiatives at secondary schools in Tanzania mainland. Dodoma, Tanzania. MoEVT (2007) ICT Policy for Basic Education. Dar es Salaam, Tanzania. Mokhtar, S. A. et al. (2013) ‘Cloud computing in academic institutions’, in Proceedings of the 7th International Conference on Ubiquitous Information Management and Communication ICUIMC ’13. New York, New York, USA: ACM Press, pp. 1–7. doi: 10.1145/2448556.2448558. Mtebe, J. S. (2015) ‘Learning Management System success : Increasing Learning Management System usage in higher education in sub-Saharan Africa’, International Journal of Education and Development using Information and Communication Technology (IJEDICT), 11(2), pp. 51–64. Available at: http://ijedict.dec.uwi.edu/viewarticle.php?id=2005. Mtebe, J. S. and Kissaka, M. M. (2015) ‘Heuristics for Evaluating Usability of Learning Management Systems in Africa’, in IST-Africa 2015 Conference Proceedings. Lilongwe, Malawi: Paul Cunningham and Miriam Cunningham (Eds), pp. 1–13. Mtebe, J. S. and Kondoro, A. W. (2016) ‘Using Mobile Moodle to Enhance Moodle LMS Accessibility and Usage at the University of Dar es Salaam’, in IST-Africa 2016 Conference Proceedings. Durban, South Africa: Paul Cunningham and Miriam Cunningham (Eds), pp. 1–11. doi: 10.1109/ISTAFRICA.2016.7530649. Mtebe, J. S. and Raisamo, R. (2014) ‘eLearning cost analysis of on-premise versus cloud-hosted implementation in Sub-Saharan Countries’, The African Journal of Information Systems, 6(2), pp. 48–64. Available at: http://digitalcommons.kennesaw.edu/cgi/viewcontent.cgi?article=1216&context=ajis. MWTC (2016) National Information and Communication Technology (ICT) Policy. Dar es Salaam, Tanzania. Available at: https://tanzict.files.wordpress.com/2016/05/national-ict-policyproofed-final-nic-review-2.pdf. Pocatilu, P., Alecu, F. and Vetrici, M. (2010) ‘Using Cloud Computing for E-learning Systems 2 Cloud Computing’, WSEAS Transactions on Computers, 9(1), pp. 42–51. Selviandro, N., Suryani, M. and Hasibuan, Z. A. (2014) ‘Open learning optimization based on cloud technology: Case study implementation in personalization e-learning’, in 16th 102 IJEDICT International Conference on Advanced Communication Technology, pp. 541–546. doi: 10.1109/ICACT.2014.6779019. Sultan, N. (2010) ‘Cloud computing for education: A new dawn?’, International Journal of Information Management, 30(2), pp. 109–116. doi: 10.1016/j.ijinfomgt.2009.09.004. Tarimo, R. and Kavishe, G. (2017) ‘Internet access and usage by secondary school students in Morogoro Municipality, Tanzania’, International Journal of Education and Development using Information and Communication Technology (IJEDICT), 13(2), pp. 56–69. Available at: http://ijedict.dec.uwi.edu/viewarticle.php?id=2338. TCRA (2017) Quarterly communications statistics report : April- June 2017. Dar es Salaam, Tanzania. Available at: https://www.tcra.go.tz/images/documents/telecommunication/TelCom-Statistics-June2017.pdf. Unwin, T. et al. (2010) ‘Digital learning management systems in Africa: myths and realities’, Open Learning: The Journal of Open and Distance Learning, 25(1), pp. 5–23. doi: 10.1080/02680510903482033. Copyright for articles published in this journal is retained by the authors, with first publication rights granted to the journal. By virtue of their appearance in this open access journal, articles are free to use, with proper attribution, in educational and other non-commercial settings. Original article at: http://ijedict.dec.uwi.edu/viewarticle.php?id=2381 Copyright of International Journal of Education & Development using Information & Communication Technology is the property of University of the West Indies Open Campus and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. http://wjst.wu.ac.th MiniReview A Survey on Communication Issues in Mobile Cloud Computing Dasari Naga RAJU1,* and Vankadara SARITHA2 1 Department of Computing Science and Engineering, Sri Venkateswara College of Engineering and Technology, Chittoor, India 2 Department of Computing Science and Engineering, Sri Vidyanikethan Engineering College, Tirupati, India (*Corresponding author’s e-mail: raj2dasari@gmail.com) Received: 27 March 2016, Revised: 31 July 2016, Accepted: 29 August 2016 Abstract Despite the expanding utilization of mobile devices, exploring their full resources is an issue due to their limited battery power, processing power and data storage. The integration of cloud computing with mobile devices solves these issues by offloading major computation in to the cloud. This paper provides a survey on Mobile Cloud Computing (MCC), which helps to understand the MCC architecture, communication issues and applications. An extensive survey is made of communication issues and different approaches are discussed to overcome the communication issues. Finally open research challenges are also provided which will be helpful for active researchers in the field of MCC. Keywords: QoS, offloading, mobile, cloud, communication Introduction In recent years, mobile devices (e.g., tablet pcs, smartphones, etc.) have become a part of human life. Mobile devices are rich in various applications like Google apps, iPhone apps which run on a remote server and/or mobile devices connected through wireless networks. The possible services provided by the mobile devices are service and computing due to the advanced inbuilt sensors like GPS, pressure sensors, light sensors, accelerometer, and magnetometer. The greater utilization of these sensor requires more analysis of data leading to an increase in computation. Mobile computing [1] is the emerging trend in IT, industry and commercial fields. Mobile devices are facing many challenges in communications (security and mobility) and their resources (battery, storage and bandwidth) [2]. The constrained resources fundamentally obstruct the improvement in Quality of Service (QoS). Cloud computing is treated as one of the emerging platforms for computations. Cloud computing offers services like infrastructure (servers, storage, and networks), platforms (Operating systems and middle-ware services) and software (application programs) to users. Cloud computing provides flexible and virtualized services based on user requirements. The scalability can be increased due to nonmaintenance of servers and other required infrastructure. The cloud service providers provide services to the clients with minimum cost and in on demand fashion. The richness of cloud computing and the drawback of mobile computing has led to the development of a new environment called Mobile Cloud Computing (MCC). This paper presents the survey on MCC with the discussion of the paper concentrated more on the communication issues and applications of MCC. Towards mobile cloud computing Mobile cloud computing is the term evolved from the combination of mobile computing and cloud computing. It didn’t take much time to introduce MCC after the introduction of cloud computing in mid- Walailak J Sci & Tech 2018; 15(1): 1-17. Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA 2007. Most industries are concentrating on MCC to reduce the cost of applications in mobile devices by introducing the cloud concept in to it. Researchers are focusing on saving the energy of mobile devices which it is commonly called green IT [3]. As per the mobile cloud computing forum, MCC is defined as [4]; “The mobile cloud computing is a paradigm where the data storage and computation is performed external to the mobile device. Mobile device take the advantages of cloud computing and process the data in to the cloud to save the resources in the internal environment.” The MCC is defined as the most efficient tool for accessing the applications and services over the internet and it is a combination of the mobile environment and cloud computing [5-7]. MCC architecture Figure 1 shows the overview of MCC architecture. The MCC architecture is divided in to 3 layers such as mobile environment, wireless medium and cloud environment. The mobile environment contains the mobile devices that are connected to the sink nodes to maintain the communication between the mobile devices and network. The AAA (for authentication, authorization and accounting) policy is applied to establish the communication. As a next step, the mobile users request the services from the cloud; the requests are handled by the cloud through the internet services. The cloud forwards the mobile user request to the cloud services (application, web and data servers). The architecture of cloud computing is defined with 4 layers with different contexts. The 4 layers are application, platform, unified resource and fabric [8]. The application layer is responsible for running the applications which are present in the cloud. The platform layer contains the specified tools for development software, middleware for supporting heterogeneous communication and is responsible for software development. The unified resource layer provides the virtualization of the resources. The fabric layer consists of the hardware environment such as storage and network. Aneka platform was introduced to build .NET applications in multi programming models to support developers [9,10]. Huang et al. [11], to support business models the authors introduced business oriented services to the cloud. They have developed a framework with a scalable, low cost and secure platform for web based services. But, the research is not concentrated on major QoS issues such as availability, security and reliability. Tablet Cloud Environment Figure 1 Architecture of mobile cloud computing. 2 Walailak J Sci & Tech 2018; 15(1) Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA Benefits of mobile cloud computing To overcome the issues of mobile computing (mobility, bandwidth and portability) [12], cloud computing is a promising solution. The major benefits of integrating cloud technology to the mobile computing are given as follows. Extending data storage: Data storage is one of the major constraints of mobile devices. MCC offers services to the users to utilize the data storage in the cloud. Amazon S3 is one example of cloud storage which provides storage as a service [13]. Facebook is one successful social networking sites that is utilizing cloud services in the mobile device [14]. Flicker [15] and ShoZu [16] are image storage application benefits from the MCC. Improving computational capacity: Computing-intensive applications need more computational capacity, but the mobile devices are lack in the required level of processing power. MCC helps in reducing the cost of executing the application by synchronizing the cloud to the mobile device in terms of processing capacity. For example, the cloud can provide multimedia services [17], online gaming [18] or E-banking [19] for mobile devices. The examples which were discussed earlier consume a lot of time and energy when they are executed in a mobile environment, but they consume less time in the cloud environment. Enhancing battery life: The major challenge in the face of mobile devices is limited battery power. Many solutions have been proposed to improve the battery life by enhancing the performance of the CPU and by managing the storage and screen resolution smartly [20-23]. These solutions require additional changes to the mobile devices and it results in new hardware and additional cost. Computational offloading is a technique which can be implemented for migrating the complex and large computations from mobile devices to cloud servers. This will reduce the load on the mobile device and increase the performance of the battery. Several experiments were conducted using offloading techniques. The results obtained from the experiments showed that the remote execution of the application can save energy [24,25]. A mathematical model was introduced for the reduction of energy in mobile devices. They obtained up to a 45 % reduction in energy consumption [24]. The MAUI (Mathematical Arithmetic Unit and Interface) architecture was proposed in Cuervo et al. [26], for reducing energy in MCC. Their approach was to offload the mobile game components in to the VMs of cloud; it saved almost 27 % in the energy of the mobile device. Enhancing reliability: The offloading of data and computation to the cloud makes the application more reliable. The data storage in servers creates backup and it will be helpful to reduce data loses on mobile devices. A comprehensive security model has been designed for the MCC for both users and providers. The model had proposed to control the unauthorized access of data from the MCC [27]. The cloud can provide services like authentication, malware detection and virus scanning to mobile users [28]. Communication issues in the mobile cloud computing The MCC is a combination of both the mobile environment as well as cloud computing, it has many challenges regarding the mobile communication and data provisioning in the cloud. This section describes several research issues, research solutions and future directions in MCC. Quality of service in MCC In MCC, the mobile users are able to access the resources in the cloud to reduce energy consumption, but the mobile users faced many communication issues regarding the connection to the cloud. The issues are limited bandwidth, network delay, signal attenuation and network disconnection. Figure 2 explains the QoS issues in MCC. Walailak J Sci & Tech 2018; 15(1) 3 Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA Computation Capacity Unrealistic Communication Limited Bandwidth QoS Issues Node Mobility Routing RoutingProtocols protocols Heterogeneity Figure 2 QoS issues in the mobile cloud computing. The computation capacities in the mobile devices are limited and very poor when compared to desktops, laptops, etc. The computation capacity always affects the services of mobile cloud computing. Limited bandwidth always creates major problems in the communication of MCC. Jin et al. [29] proposed architecture to solve the bandwidth problem in MCC. They concentrated on sharing of limited bandwidth to the mobile users which are requesting from the same location and the same application. But, the limitation of the solution is that it cannot address the unfairness in the distribution of bandwidth. Mishra et al. [30] proposed a model for bandwidth shifting and redistribution in MCC. They followed an auction based model for distribution of bandwidth. Canepa et al. [31] proposed a mechanism for unrealistic communication, the model has the capacity to search unavailable nodes when the link is in a failure state. To address the network heterogeneity, the author in [32] presented an intelligent radio network access for heterogeneous networks. To solve the network delay in MCC, Intel researchers proposed a model called CloneCloud [33]. This approach manages to send the data to the nearest server; this brings a lot of advantages to the mobile platform to speed up the computation. To address the mobility management in MCC, Rahimi et al. [34] proposed a heuristic algorithm MuSIC to convert the mobility patterns into mobile usage patterns. A detailed survey on QoS issues is shown in Table 1. 4 Walailak J Sci & Tech 2018; 15(1) Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA Table 1 Related Work regarding QoS issues in MCC. Issue Papers Techniques used Limited bandwidth [29,30,35,36] Availability [37,38] Heterogeneity [39,40] Mobility management [41-43] Service selection [44,45] power-efficient mobile P2P media streaming, Auction based mechanism, fmobile software, Machine to Machine (M2M) cloud Distributed Application Processing Frameworks, CloneCloud Multidimensional heterogeneity, Service-based arbitrated multitier M2C2, Mobility-Aware Optimal Service Allocation, OpenFlow MACSS, network-centric Barbera et al. [35] suggested architecture for MCC where each real device is connected with a software clone of the cloud environment. They considered 2 types of clones, one is an off-clone that supports computational offloading and the other is a back-clone used to restore the data of the mobile device. The architecture was evaluated based on the bandwidth and energy consumption of the mobile devices. Ravi Teja et al. [36] proposed a congestion network model for M2M devices by managing the network traffic using the 2 level mapping. The first level mapping is carried in between cluster head to the sink nodes and second level of mapping is carried in between sink nodes to the cloud gateway. The mapping is done based on the social choice mechanism. Shiraj et al. [37] extensively reviewed distributed application processing frameworks for mobile devices in MCC. They made the contribution towards the study of current offloading frameworks and analysed the critical aspects and implications. Zhang et al. [38] presented the research challenges in cloud computing. The clone-cloud is addressed for the efficient availability of the network bandwidth to the devices which are connected to the cloud services. They reviewed different challenges and important research issues in the direction of cloud computing. Sanaei et al. [39] discussed heterogeneity and different challenges in MCC. Heterogeneity in terms of hardware, software, platform and network were analysed. The impacts of heterogeneity in various environments were discussed and they also presented the handling approaches like middleware, virtualization and service oriented architecture for heterogeneity. Sanaei et al. [40] proposed a Servicebased Arbitrated Multitier Infrastructure (SAMI) for a service oriented platform. This architecture deals with 3 layers: cloud, Mobile Network Operators and MNO’s for dealers. This architecture concentrates on the arbitrator layer which has the functionality to classify the services and assign them to the available resources based on the latency, service resource requirement and security. Mitra et al. [41] concentrated on heterogeneous access of networks over MCC. They proposed an M2C2 model for supporting the multihoming, cloud and network probing, and cloud and network selection. The experimental results of the model support the efficiency of the M2C2 model. In the future, the usability of the mobile devices will increase rapidly. It is important to manage the mobility of the devices and handover processes. Ryu et al. [42] proposed fast handovers for MIPv6 (FMIPv6). FMIPv6 provided an efficient handover process and prevented the packet losses and handover latency through buffering and tunnelling. Kempf et al. [43] described the Software Defined Networks (SDN) with the help of mobile Evolved Packet Core (EPC). They proposed OpenFlow 1.2 for 2 vendors. One is for Walailak J Sci & Tech 2018; 15(1) 5 Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA encapsulation and de-capsulation of virtual points and another is flow routing through GTP Tunnel Endpoint Identifier (TEID). To address the service selection, Liu et al. [44] proposed a mobile-aware framework to Mobile Cloud Streaming Services (MACSS) for optimizing mobility management by using server selection. The MACSS significantly improved the user mobility management and channel selection and variation. Ejaz et al. [45] concentrated on the limitations of mobile devices, overcoming these limitations by integrating with cloud services. They analysed the task offloading process through the network centric mechanism. The proposed model was analysed with an application migration process with the impact of various parameters such as file size, the number of users in the LAN, the traffic load on the Wi-Fi network, the number of nodes in the network, message length and mobility speed. The proposed model was tested with migration time and packet delivery ratio. The main drawback of the proposed model was that it did not consider the energy consumption of the network. Operational issues in MCC The operational issues of MCC refer to technological matters such as computational offloading, cost benefit models and connection protocols used. Computational offloading The main operation of MCC is offloading of tasks from the mobile device to the cloud. Due the distance between the mobile and cloud a heterogeneous communication is needed for the underlying system. Different research has been carried out in heterogeneous communication in many ways. In this section, the review concentrates on the client server model, VM migration model and mobile agents. Client-server model: The communication with the network is carried with a client server model. The client server model is a traditional technique where the communication is taken care of by remote procedure calls and remote method invocations. These methods are well-supported APIs for developers to offload the task. But, these 2 methods have to be pre-installed on the mobile devices. It is a drawback of this model when it is participating in Ad-hoc networks. In [46,47], they used the RPC for communication in the offloading process. The RPC are pre-installed into the device which invokes the functionality in the cloud and mobile SPECTA servers. The servers have the RPC pre-installed methods. The spectra clients consult the database server for information regarding the CPU, memory, availability etc., when the task is to be offloaded. Developers will partition the application manually and take the decision on which part of the application has to be offloaded and which part has to be executed locally. Marinelli [48], has proposed a model called Hyrax for applications of smartphones. They used the Hadoop framework for both data and computation in android. Hyrax investigates the likelihood of utilizing a group of mobile phones as resource providers and demonstrates the possibility of such a portable cloud. They introduce a ‘HyraxTube’ application; which is a search and sharing multimedia mobile application. The goal of HyraxTube is to permit clients to look through multimedia applications in relation to quality, location and time. Huerta-Canepa and Lee [49] presented another Hadoop framework for mobile devices considering the set of mobile devices as cloud resource providers. This method argues that the location of the user plays a major role in deciding the task offloading. The manager in the offloading process manages the sending and receiving of the task from the other devices and creates VMs on the other mobile devices. They tested this environment in a Korean OCR application. The results for the application are not satisfactory in terms of speedup. But, it showed noticeable results in energy saving. The Mobile Message Passing Interface (MMPI) framework is proposed for mobile devices, it uses Bluetooth as the communication medium for creating a connection with other mobile devices [50]. This model follows the mesh network procedure so that each mobile device can communicate with other mobile devices. They implemented the model using the BlueCove [51]. Kemp et al. [52], the Cuckoo framework was proposed for the offloading of tasks to the cloud using the Java model. The server which runs Java instances is eligible for executing the offloading tasks. The Amazon EC2, commercial cloud provider is selected to evaluate the framework. 6 Walailak J Sci & Tech 2018; 15(1) Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA Virtual machine migration: As per Clark et al. [53], virtual machine migration is defined as the process of transferring source server memory images to the destination server. This process duplicates the memory pages without the involvement of the operating system and other installed software. This technique guarantees the secure execution and no code changes are essential when tasks are offloaded, since the VM limit protects the surrounding mobile devices. VM migration is tedious to a particular level and the workload could end up being substantial for mobile devices. The CloneCloud is presented and it utilizes the virtual machine migration policy to offload the task to the cloud server through Wi-Fi or 3G network [33]. Since they utilize mobile device clones, the applications are unmodified and there is no need for decision making for example, as followed in MAUI [18]. The CloneCloud proposed the cost model for examining the cost of VM migration and execution in the cloud. The android platform mobile devices were selected to manage the clones. Sathyanarayanan et al. [54] proposed cloudlets as a solution for connecting distant clouds. The cloudlet is like a small data center which is present nearer to the devices and connected to the cloud through the internet. The mobile devices have the flexibility to connect the cloudlets for offloading of tasks. The computation power of the cloudlets is minimal when compared to the cloud. The major drawback of this process is reliability and energy consumption. MobiCloud [55] is a mechanism for integrating the cloud computing with Mobile Ad-hoc Networks (MANET). In this model, the general architecture of the MANET is considered as the service oriented architecture and each node in the architecture is considered to be a service node. The serve broker in the architecture takes care of incorporating the service nodes in to the cloud. Extended Semi-Shadow Images (ESSIs) are used to clone the environment in to the cloud. Mobile agents: Kristensen and Scavenger [56], Scavenger suggested the framework for cyber foraging which utilizes Wi-Fi as a communication medium. It uses the mobile agent approach for partitioning and execution of tasks. It also introduced the cost assessment policy with the help of a scheduler. It is possible to offload the tasks and execute at multiple servers using the framework. Apart from the advantages, the limitation of the model is that it does not consider fault tolerance. Table 2 Analysis of offloading approaches. Computational offloading Client-Server Model VM Migration Mobile Agents Frameworks Spectra [46] Chroma [47] Cuckoo [52] Hyrax [48] CloneCloud [33] MAUI [18] Cloudlets [54] MobiCloud [55] Scavenger [56] Advantages Stable and supported by APIs Disadvantages Require pre-installation, network congestion No code modification is required VM migration takes time and compatibility issues Dynamic execution and suitable for mobile devices which are not connected. Security and agent management The analysis of offloading frameworks is given in Table 2. In spite of the fact that a comparison of results have been given in some papers, looking at them against one another is difficult since the energy consumption and performance rely upon the application too. Actually, while utilizing the same framework, execution differs for distinctive applications. The communication medium (whether 3G, LTE or Wi-Fi) and size of the task plays a crucial role. Walailak J Sci & Tech 2018; 15(1) 7 Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA Cost benefit model The cost is a major issue in MCC, the offloading of the task in to the cloud considers different issues like time, energy, and economical execution. To calculate the cloud cost, Li et al. proposed a model with a set of rules [57]. They considered utilization cost and the Total Cost of Ownership (TCO) as the deciding factors for cost analysis. Utilization cost is calculated in terms of resource utilization with respect to particular users according to the dynamic demand; resources are VMs, power, computational resources and software. The TCO is the cost estimation for installing the IT infrastructure. In terms of the cloud, the TCO is the infrastructure cost, software cost, network cost, maintenance cost and much more. Walker et al. [58] presented an analysis of electrical cost, the memory utilization cost and infrastructure maintenance cost of the commercial cloud service provider Amazon EC2. They made a decision model whether to buy the services or to lease. The decision model calculates the Net Present Value (NPV) of the services. The value of the NPV ≥ 0 represents to buy the service otherwise to lease the service. Eq. (1) represents the calculation of NPV. CY − EY + LY S + −C Y (1 + R F ) (1 + R F ) n Y =0 n NPV = ∑ (1) where CY, EY and S represent the disc controller unit cost, operating cost per year and disc lifetime salvage value. LY represents the expected lease payment per year and RF represents the interest rate per year. Kim et al. [59] proposed a Luyapunov drift-plus-penalty technique for dual side control algorithms in mobile device and cloud services. They suggested an NC-UC (Non-Cooperation) algorithm for the mobile device by concentrating on the delay factor and NC-CC algorithm for the cloud. In MCC, due to the dynamic nature of mobile devices, the resources may change at any movement. Therefore, the cost analysis model is required to get benefit from the offloading process. Figure 3 illustrates the user specific requirements of cost analysis. User Requirements Decision Offloading Available Resources Required Resources Figure 3 Cost analysis for MCC. Connection protocols The MCC research has been carried with many communication protocols like 3G, LTE, Wi-Fi and Bluetooth, but the majority of users have utilized Wi-Fi as a communication medium. An overview of the communication protocols used in MCC frameworks is shown in Table 3. 3G: Third Generation mobile telecommunication is the mobile technology which is used for communication in MCC [60]. The data rate of 3G is much slower when compared to LTE and Wi-Fi. LTE: Long Term Evolution is the technology for the communication that increases bandwidth for mobile users. The capacity of the LTE is up to 100 Mbps. Moreover, LTE provides additional advantages 8 Walailak J Sci & Tech 2018; 15(1) Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA such as quicker handoff, wide coverage area and varied services [61]. However, it has some drawbacks related to the access of protocols, QoS, network architecture [62]. Wi-Fi: Wireless Fidelity (802.11b) is operated on unlicensed 2.4 GHz bandwidth. Initially, Wi-Fi was introduced to replace the wired network for sharing of data among the computers; later on it was used by mobile devices for data. The range of Wi-Fi is 100 m and the typical data rate is 11 Mbps [63,64]. Bluetooth: Bluetooth was introduced for wireless devices like mobile phones, laptops, tablets and is designed with low cost transceiver chips. The transmission range of Bluetooth is 10 m and data rate is up to 24 Mbps [65]. The analysis of communication protocols in MCC is given in Table 3. Table 3 Comparison between Communication Protocols. Communication protocols MCC models Advantages Disadvantages 3G [18,33,52,60] Near pervasive coverage LTE [18,61,62] Higher data rate Wi-Fi [46-48,52,63,64] Bluetooth [50,52,65] Better performance, less energy consumption when compared to 3G and LTE Low energy consumption, availability [66],compared to the other protocols [67] Bandwidth is limited, higher round trip time, high energy consumption Higher energy consumption Security threats, limited operational issues Limited range Security issues The computational offloading of tasks from mobile devices poses some questions regarding privacy and security. The data and the user programs are sent to the servers that are not under the control of the user. Hence it raises a privacy issue. The third party is introduced to store the user’s data so there is no security. Major research has been carried out in the area of protecting outsourced data [68-70] the solutions include hardware based execution [71], homomorphic encryption [72-74] and steganography [76,77]. These techniques have their own limitations due to the size of the encryption key and the approaches the researches followed for encryption. Context awareness Context awareness is a mechanism in MCC that identifies the state of the mobile user and surroundings of the device and infers the context information. This is important for MCC for offloading decisions because the performance of the offloading may vary based on the users location and context; the advancement of the research is carried out in the area of adaptive mechanism based on context awareness [78-81]. Applications of MCC MCC has gained popularity in the global mobile market. Different applications in the android platform support MCC. In this section MCC applications are presented briefly. Walailak J Sci & Tech 2018; 15(1) 9 Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA M-learning Mobile learning is developed with the help of E-learning and with the combination of mobility. The conventional M-learning process has some restrictions, for instance, low network transmission rate, limited educational resources, and the high cost of the device [82,83]. To overcome those issues the cloud based m-learning has been introduced. Utilization of services like networks, data, and storage from the cloud is much cheaper when compared to conventional M-learning [84]. Yu-Shan et al. [85] made a study on cloud based m-learning and designed a model called a non-equivalent pretest-posttest. The research showed positive results on creative performance of the students in engineering. The proposed model also improves the overall performance of the designed products. Aftab et al. [86] studied whether the design of E-learning is suitable for the MCC environment. They made several comparisons regarding the MCC architectures and estimated the performance of the MCC. M-commerce In a recent trend, M-Commerce is one of the developing areas concerned with the business market. M-commerce generally has some task with requirement for mobility. For example, online purchasing, mobile messaging, mobile banking, mobile ticketing etc. M-commerce faces some issues regarding the low bandwidth, security, high complexity, and mobile device configuration. These issues can be resolved by integrating m-commerce with the cloud. Yang and Lin [87], the authors proposed a mobile payment mechanism with the anonymity of cloud computing. The model concentrated on reducing the computational cost and non-repudiation requirement at the mobile device. The results of the model showed better results in terms of security. Figure 4 explains the payment gateway model in MCC. Banking Private Network Issuer (I) Acquirer (A) Payment Gateway Client Merchant Figure 4 Payment gateway for M-commerce adapted from [87]. Turban et al. [88] have explained the infrastructure that supports m-commerce, applications of mcommerce in banking and financial services, and value added attributes of the m-commerce. M-health To overcome the limitations of the traditional health care, mobile health (M-health) using MCC has been introduced. The limitations are physical storage, medical errors, privacy and security [89,90]. Mhealth provides convenient access to the health records of patients without any complications. Besides 10 Walailak J Sci & Tech 2018; 15(1) Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA this, M-health offers multiple on demand services to the hospitals with the help of the cloud [91-94]. For instance, 4 modules in [90] to deal with M-health. Emergency management: It can manage emergency situations by responding to calls from accidents and incidents. Daily heath monitoring: This will be used for continuous health monitoring of patients and can be useful for the doctors to observe the patients daily activities. Context aware mobile devices: This will monitor the blood pressure, pulse rate, heart rate and alerts the health system. Ubiquitous access: It allows the patient to check their health status daily and past data will be accessed from the data base. Mobile gaming The application of MCC in the field of gaming has made a drastic change to the gaming market. The computation needed for gaming is completely offloaded to the cloud and it will ultimately reduce the computational cost at the mobile device. Wang and Dey [95], the authors proposed an adaptive mechanism for a gaming platform that considers the users parameters and communication parameters and dynamically adjusts to the environment. They concentrated on the rendering an adaptive mechanism which minimizes the objects in the display screen of the mobile device and gives the user a fair play mode. The objective of this model is to increase the user experience and reduce the communication and computational cost. Other practical applications The MCC is used for evaluating multimedia [96-98], text editors [99-101], vision and voice recognition [102]. Text editors consume less data i.e., computations like spell check. Multimedia, vision and voice recognition consumes large amounts of data in the form of videos and images [102]. Open challenges for research The research which has been carried out up to now in the field of MCC is discussed in the earlier sections. Though there are several open challenges which have to be discussed. In this section, some research directions related to the MCC are presented. Network management The network management plays a crucial role in the performance of task execution over MCC. The better network manages to improve the link performance and bandwidth usage. Cognitive radio networks improve the spectrum utilization of mobile users. When the cognitive radio networks are integrated with the MCC, it saves thousands of dollars to the network providers. Low bandwidth Already a lot of research is going on in the field of improving bandwidth efficiency. But, it is continuous because of the drastic increase in the number of mobile and cloud users. LTE can be suggested as a solution as it is a promising technology to overcoming limitations and for improving bandwidth efficiently. Moreover, LTE also has some operational issues like network architecture, QoS, protocols, and much more. Quality of service The QoS improvement is the major research area in the field of MCC. Mobile users can access the services from the cloud service providers. Moreover, mobile users face some challenges regarding the Walailak J Sci & Tech 2018; 15(1) 11 Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA access of the services from the cloud such as network congestion at the time of mobility, time delay at the time of establishing the connection with the cloud. So, better QoS mechanisms need to be developed to overcome the research gaps. Compatibility Compatibility plays a major role while the mobile user is connected with the cloud interface. The present interface between mobile users and cloud platform is web services. The web services are not particularly designed for mobile devices. Hence, the web interface becomes overhead. There is a need for an efficient programming interface to address the compatibility issues in MCC. Cost In MCC, the mobile services as well as cloud service providers are converged to provide the services. Therefore, cloud service providers have different cost policies for utilization of resources. So, cost is a major issue in MCC i.e., how to decrease the cost of cloud computing and how the resources will be utilized efficiently is a major research area. Conclusions MCC aims to enable mobile users by giving them consistent and rich functionality, apart from the limitations of the mobile devices. MCC provides versatile support for mobile applications in the future. This paper surveys and classifies large bodies of research regarding communication issues and applications of mobile cloud computing. Various types of architectures for computational offloading and virtualization are examined. Classification of different types of applications that are used by the MCC is also presented. Finally the open issues and research directions in the field of MCC are presented. References M Satyanarayanan. Mobile computing: The next decade. In: Proceedings of the 1st ACM Workshop on Mobile Cloud Computing & Services. San Francisco, USA, 2010, p. 5-11. [2] M Satyanarayanan. Fundamental challenges in mobile computing. In: Proceedings of the 5th Annual ACM Symposium on Principles of Distributed Computing. Philadelphia, USA 1996, p. 1-7. [3] M Ali. Green cloud on the horizon. In: Proceedings of the 1st International Conference Cloud Computing. Beijing, China, 2009, p. 451-8. [4] Cloud Computing Forum. Available at: http://www.mobilecloudcomputingforum.com, accessed May 2015. [5] White Paper. Mobile Cloud Computing Solution Brief. AEPONA, 2010, p. 1-5. [6] JH Christensen. Using RESTful web-services and cloud computing to create next generation mobile applications. In: Proceedings of the 24th ACM SIGPLAN Conference Companion on Object Oriented Programming Systems Languages and Applications. 2009, p. 627-34. [7] L Liu, R Moulic and D Shea. Cloud Service Portal for Mobile Device Management. In: Proceedings of IEEE 7th International Conference on e-Business Engineering. 2011, p. 474-9. [8] I Foster, Y Zhao, I Raicu and S Lu. Cloud computing and grid computing 360-degree compared. In: Proceedings of Grid Computing Environments Workshop. 2008, p. 1-10. [9] C Vecchiola, X Chu and R Buyya. Aneka: A software platform for .NET-based cloud computing. High Speed Large Scale Sci. Comput. 2009; 18, 267-89. [10] R Buyya, CS Yeo, S Venugopal, J Broberg and I Brandic. Cloud computing and emerging IT platforms: Vision, hype, and reality for delivering computing as the 5th utility. Future Generat. Comput. Syst. 2009; 6, 599-616. [11] Y Huang, H Su, W Sun, JM Zhang, CJ Guo, MJ Xu, BZ Jiang, SX Yang and J Zhu. Framework for building a low-cost, scalable, and secured platform for Web-delivered business services. IBM J. Res. Develop. 2010; 6, 535-46. [12] GH Forman and J Zahorjan. The challenges of mobile computing. Computer 1994; 4, p. 38-47. [1] 12 Walailak J Sci & Tech 2018; 15(1) Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] Dasari Naga RAJU and Vankadara SARITHA Amazon S3, Available at: http://aws.amazon.com/s3, accessed July 2015. Facebook, Available at: http://www.facebook.com, accessed July 2015. Flicker, Available at: http://www.flickr.com accessed July 2015. Shozu, Available at: http://www.shozu.com/portal/index.do accessed July 2015. L Li, X Li, S Youxia and L Wen. Research on mobile multimedia broadcasting service integration based on cloud computing. In: Proceedings of the Multimedia Technology. 2010, p. 1-4. DN Brian and P Wiemer-Hastings. Addiction to the internet and online gaming. Cyberpsychol. Behav. 2005; 8, 110-4. HA Garcia and H Kalva. Cloud transcoding for mobile video content delivery. In: Proceedings of the IEEE International Conference on Consumer Electronics, 2010, p. 379-82. R Kakerow. Low power design methodologies for mobile communication. In: Proceedings of IEEE International Conference on Computer Design: VLSI in Computers and Processors, 2003, p. 8-13. LD Paulson. Low-power chips for high-powered handhelds. Computer 2003; 1, 21-4. JW Davis. Power benchmark strategy for systems employing power management. In: Proceedings of Electronics and the Environment, 1993, p. 117-20. RN Mayo and P Ranganathan. Energy consumption in mobile devices: Why future systems need requirements-aware energy scale-down. In: Proceedings of the International Workshop on PowerAware Computer Systems. 2003, p. 26-39. A Rudenko, P Reiher, GJ Popek and GH Kuenning. Saving portable computer battery power through remote process execution. In: Proceedings of the ACM SIGMOBILE Mobile Computing and Communications Review. 1998; 1, p. 19-24. A Smailagic and M Ettus. System design and power optimization for mobile computers. In: Proceedings of the VLSI. 2002, p. 1-15. E Cuervo, A Balasubramanian, DK Cho, A Wolman, S Saroiu, R Chandra and P Bahl. MAUI: making smartphones last longer, with code offload. In: Proceedings of the 8th International Conference on Mobile Systems, Applications, and Services. 2010, p. 49-61. P Zou, C Wang, Z Liu and D Bao. Phosphor: A cloud based DRM scheme with sim card. In: Proceedings of the Web Conference. 2010. p. 459-64. J Oberheide, K Veeraraghavan, E Cooke, J Flinn and F Jahanian. Virtualized in-cloud security services for mobile devices. In: Proceedings of the 1st Workshop on Virtualization in Mobile Computing. 2008, p. 31-5. X Jin and YK Kwok. Cloud assisted P2P media streaming for bandwidth constrained mobile subscribers. In: Proceedings of Parallel and Distributed Systems. 2010, p. 800-6. S Misra, S Das, M Khatua and MS Obaidat. QoS-guaranteed bandwidth shifting and redistribution in mobile cloud environment. IEEE Trans. Cloud Comput. 2014; 2, 181-94. G Huerta-Canepa and D Lee. A virtual cloud computing provider for mobile devices. In: Proceedings of the 1st ACM Workshop on Mobile Cloud Computing & Services: Social Networks and Beyond. 2010, p. 6-7. A Klein, C Mannweiler, J Schneider and HD Schotten. Access schemes for mobile cloud computing. In: Proceedings of 11th International Conference on Mobile Data Management. 2010, p. 387-92. BG Chun, S Ihm, P Maniatis, M Naik and A Patti. Clonecloud: Elastic execution between mobile device and cloud. In: Proceedings of the 6th conference on Computer systems. 2011, p. 301-14. MR Rahimi, N Venkatasubramanian and AV Vasilakos. Music: Mobility-aware optimal service allocation in mobile cloud computing. In: Proceedings of the IEEE 6th International Conference on Cloud Computing. 2013, p. 75-82. M Barbera, S Kosta, A Mei and J Stefa. To offload or not to offload? The bandwidth and energy costs of mobile cloud computing. In: Proceedings of INFOCOM. 2013, p. 1285-93. PVSR Teja, S Chatterjee, SN Das and S Misra. Two-level mapping to mitigate congestion in machine to machine (M2M) cloud. In: Proceeding of Applications and Innovations in Mobile Computing. 2015, p. 104-8. Walailak J Sci & Tech 2018; 15(1) 13 Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA [37] M Shiraz, A Gani, RH Khokhar and R Buyya. A review on distributed application processing frameworks in smart mobile devices for mobile cloud computing. IEEE Comm. Surv. Tutor. 2013; 15, 1294-13. [38] Q Zhang, L Cheng and R Boutaba. Cloud computing: State-of-the-art and research challenges. J. Int. Ser. Appl. 2010; 1, 7-18. [39] S Zohreh, S Abolfazli, A Gani and R Buyya. Heterogeneity in mobile cloud computing: Taxonomy and open challenges. IEEE Comm. Surv. Tutor. 2014; 16, 369-92. [40] Z Sanaei, S Abolfazli, A Gani and M Shiraz. SAMI: Service-based arbitrated multi-tier infrastructure for Mobile Cloud Computing. In: Proceedings of the Communications in China Workshops. 2012, p. 14-9. [41] K Mitra, S Saguna, C Ahlund and DG Lulea. M2C2: A mobility management system for mobile cloud computing. In: Proceedings of the Wireless Communications and Networking Conference. 2015, p. 1608-13. [42] S Ryu, K Lee and Y Mun. Optimized fast handover scheme in Mobile IPv6 networks to support mobile users for cloud computing. J. Supercomput. 2012; 59, 658-75. [43] J Kempf, B Johansson, S Pettersson, H Lüning and T Nilsson. Moving the mobile evolved packet core to the cloud. In: Proceedings of the IEEE 8th International Conference on Wireless and Mobile Computing, Networking and Communications. Barcelona, 2012, p. 75-84. [44] R Liu, X Yuan, J Xu, Y Zeng, M Cao, J Liu, L Xu and Q Fang. A novel server selection approach for mobile cloud streaming service. Simulat. Modell. Pract. Theor. 2015; 50, 72-82. [45] E Ahmed, A Akhunzada, M Whaiduzzaman, A Gani, SHA Hamid and R Buyya. Network-centric performance analysis of runtime application migration in mobile cloud computing. Simulat. Modell. Pract. Theor. 2015; 50, 42-56. [46] J Flinn, S Park and M Satyanarayanan. Balancing performance, energy, and quality in pervasive computing. In: Proceedings of the 22nd International Conference on Distributed Computing Systems. 2002, p. 217-26. [47] R Balan, M Satyanarayanan, S Park and T Okoshi. Tactics-based remote execution for mobile computing. In: Proceedings of the 1st International Conference on Mobile Systems, Applications and Services. 2003, p. 273-86. [48] EE Marinelli. 2009, Hyrax: Cloud Computing on Mobile Devices using MapReduce, Masters Dissertation, Carnegie Mellon University. [49] G Huerta-Canepa and D Lee. A virtual cloud computing provider for mobile devices. In: Proceedings of the 1st ACM Workshop on Mobile Cloud Computing & Services. 2010, p. 1-5. [50] DC Doolan, S Tabirca and LT Yang. Mmpi a message passing interface for the mobile environment. In: Proceedings of the 6th International Conference on Advances in Mobile Computing and Multimedia. 2008, p. 317-21. [51] BlueCove. Available at: http://www.bluecove.org, accessed May 2015. [52] R Kemp, N Palmer, T Kielmann and H Bal. Cuckoo: A computation offloading framework for smartphones. Mobile Comput. Appl. Serv. 2010; 76, 59-79. [53] C Clark, K Fraser, S Hand, J Hansen, E Jul, C Limpach and I Pratt. Warfield. Live migration of virtual machines. In: Proceedings of the 2nd Conference on Symposium on Networked Systems Design & Implementation. 2005, p. 273-86. [54] M Satyanarayanan, P Bahl, R Caceres and N Davies. The case for VM-based cloudlets in mobile computing. IEEE Pervas. Comput. 2009, 8, 14-23. [55] D Huang, X Zhang, M Kang and J Luo. Mobicloud: Building secure cloud framework for mobile computing and communication. In: Proceedings of the 5th IEEE International Symposium on Service Oriented System Engineering. 2010, p. 27-34. [56] MD Kristensen and I Scavenger. Transparent development of efficient cyber foraging applications. In: Proceedings of the IEEE International Conference on Pervasive Computing and Communications. 2010, p. 217-26. 14 Walailak J Sci & Tech 2018; 15(1) Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA [57] L Xinhui, L Ying, L Tiancheng, Q Jie and W Fengchun. The method and tool of cost analysis for cloud computing. In: Proceedings of the IEEE International Conference on Cloud Computing, 2009, p. 93-100. [58] E Walker, W Brisken and J Romney. To lease or not to lease from storage clouds. Computer 2010; 43, 44-50. [59] K Yeongjin, J Kwak and S Chong. Dual-side dynamic controls for cost minimization in mobile cloud computing systems. In: Proceedings of the 13th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks. 2015, p. 443-50. [60] M Rubens, J Araujo, D Oliveira, P Maciel and K Trivedi. Sensitivity analysis of a hierarchical model of mobile cloud computing. Simulat. Modell. Pract. Theor. 2015; 50, 151-64. [61] X Sun and N Ansari. Energy-optimized bandwidth allocation strategy for mobile cloud computing in LTE networks. In: Proceedings of the Wireless Communications and Networking Conference. 2015, p. 2120-25. [62] A Mushtaq, JM Zain, MF Zolkipli and G Badshah. Mobile cloud computing & mobile's battery efficiency approaches: A Review. J. Theor. Appl. Inform. Tech. 2015; 79, 153-75. [63] L Yu, WT Tsai, X Wei, J Gao, TT Hildebrandt and G XQ. Modeling and analysis of mobile cloud computing based on bigraph theory. In: Proceedings of the Mobile Cloud Computing, Services, and Engineering. 2014, p. 67-76. [64] E Ahmed, A Gani, MK Khan, R Buyya and SU Khan. Seamless application execution in mobile cloud computing: Motivation, taxonomy, and open challenges. J. Netw. Comput. Appl. 2015; 52, 154-72. [65] S Abolfazli, Z Sanaei, A Gani, F Xia and WM Lin. RMCC: Restful mobile cloud computing framework for exploiting adjacent service-based mobile cloudlets. In: Proceedings of the Cloud Computing Technology and Science. 2014. p. 793-8 [66] JJ Dong and HC Xu. A distributed online test system based on Bluetooth technology. In: Proceedings of the 2nd World Congress on Software Engineering. 2010, 1, p. 15-7. [67] S Cherry. Update: WiFi takes on Bluetooth. In: Proceedings of the IEEE Spectrum. 2008, 45, p. 147. [68] R Chow, P Golle, M Jakobsson, E Shi, J Staddon, R Masuoka and J Molina. Controlling data in the cloud: outsourcing computation without outsourcing control. In: Proceedings of the ACM workshop on cloud computing security. 2009, p 85-90. [69] Q Wang, C Wang, J Li, K Ren and W Lou. Enabling public verifiability and data dynamics for storage security in cloud computing. In: Proceedings of the European Symposium on Research in Computer Security. 2010, p. 355-70. [70] W Wang, Z Li, R Owens and B Bhargava. Secure and efficient access to outsourced data. In: Proceedings of the ACM Workshop on Cloud Computing Security. 2009, p. 55-66. [71] S Bajikar. Trusted platform module (TPM) based security on notebook pcs-white paper. White Paper, Mobile Platforms Group-Intel Corporation. 2002, p. 20-6. [72] C Gentry. Fully homomorphic encryption using ideal lattices. In: Proceedings of the ACM Symposium on Theory of Computing. 2009, p 169-78. [73] M Louk and H Lim. Homomorphic encryption in mobile multi cloud computing. In: Proceedings of the Information Networking. 2015, p. 493-7. [74] Z Wang, G Sun and D Chen. A new definition of homomorphic signature for identity management in mobile cloud computing. J. Comput. Syst. Sci. 2014; 3, 546-53. [75] C Gentry. Computing arbitrary functions of encrypted data. Commun. ACM 2010; 3, 97-105. [76] J Liu, K Kumar and YH Lu. Tradeoff between energy savings and privacy protection in computation offloading. In: Proceedings of the ACM/IEEE International Symposium on Low Power Electronics and Design. 2010, p. 213-8. [77] C Vorugunti, M Sarvabhatla and G Murugan. A secure mutual authentication protocol for cloud computing using secret sharing and steganography. In: Proceedings of the Cloud Computing in Emerging Markets. 2014, p. 1-8. Walailak J Sci & Tech 2018; 15(1) 15 Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA [78] S Goyal and J Carter. A lightweight secure cyber foraging infrastructure for resource-constrained devices. In: Proceedings of the Mobile Computing Systems and Applications. 2004, p. 184-95. [79] T Guan, E Zaluska and D Roure. A grid service infrastructure for mobile devices. In: Proceedings of the IEEE International Conference on Semantics, Knowledge, and Grid. 2005, p. 42-8. [80] H Viswanathan, EK Lee, I Rodero and D Pompili. Uncertainty-Aware autonomic resource provisioning for mobile cloud computing. IEEE Trans. Parallel. Distr. Syst. 2015; 26, 2363-72. [81] J Wan, D Zhang, S Zhao, LT Yang and J Lloret. Context-aware vehicular cyber-physical systems with cloud support: Architecture, challenges, and solutions. Commun. Mag. 2014; 8, 106-13. [82] SM Butt. Cloud centric real time mobile learning system for computer science. In: Proceedings of the International Conference on Computer and Emerging Technologies. 2014, p. 1-4. [83] M Wannous, M Mahfuri, H Nakano and T Nagai. Facilitating access to course contents during war situation with M-Learning and cloud computing technologies. In: Proceedings of the Teaching, Assessment and Learning. 2014, p. 93-6. [84] F Ozdamli and H Bicen. Effects of training on cloud computing services on M-learning perceptions and adequacies. Procedia Social Behav. Sci. 2014; 116, 5115-9. [85] YS Chang, SY Chen, KC Yu, YH Chu and YH Chien. Effects of cloud‐based m‐learning on student creative performance in engineering design. British J. Edu. Tech. 2017; 48, 101-12. [86] A Alam, AM Ansari, MM Barga and A Albhaishi. Is mobile cloud computing efficient for elearning? In: Proceedings of the 2nd International Conference on Computer and Communication Technologies. 2016, p. 109-23. [87] JH Yang and PY Lin. A mobile payment mechanism with anonymity for cloud computing. J. Syst. Softw. 2016; 116, 69-74. [88] S Kazaki and F Mendez. Mobile Commerce and Ubiquitous Computing. Electronic Commerce, Springer International Publishing. 2015, p. 257-308. [89] LT Kohn, JM Corrigan and S Donaldson. To Err Is Human: Building a Safer Health System. National Academies Press, Washington DC, 1999, p 34-67. [90] D Kopec, MH Kabir, D Reinharth, O Rothschild and JA Castiglione. Human errors in medical practice: Systematic classification and reduction with automated information systems. J. Med. Syst. 2003; 4, 297-313. [91] U Varshney. Pervasive healthcare and wireless health monitoring. J. Mobile Netw. Appl. 2007; 3, 113-27. [92] C Fernandez-Llatas, SF Pileggi, G Ibañez, Z Valero and P Sala. Cloud Computing for ContextAware Enhanced m-Health Services. In: C Fernández-Llatas and JM García-Gómez (eds.). Data Mining in Clinical Medicine. Springer, New York. 2015, p. 147-55. [93] B Xu, L Xu, H Cai, L Jiang, Y Luo and Y Gu. The design of an m-Health monitoring system based on a cloud computing platform. Enterprise Inform. Syst. 2017; 11, 17-36. [94] AM Mostafa, AE Youssef and G Alshorbagy. A framework for a smart social blood donation system based on mobile cloud computing. Health Inform. Int. J. 2014; 3, 1-10. [95] S Wang and S Dey. Rendering adaptation to address communication and computation constraints in cloud mobile gaming. In: Proceedings of the IEEE Global Telecommunications Conference. 2011, p. 1-6. [96] W Zhu, C Luo, J Wang and S Li. Multimedia cloud computing. IEEE Signal Process. Mag. 2011; 28, 59-69. [97] A Khan and K Ahirwar. Mobile cloud computing as a future of mobile multimedia database. Int. J. Comput. Sci. Commun. 2011; 1, 219-21. [98] S Wang and S Dey. Adaptive mobile cloud computing to enable rich mobile multimedia applications. IEEE Trans. Multimed. 2013; 4, 870-83. [99] P Simoens, FD Turck, B Dhoedt and P Demeester. Remote display solutions for mobile cloud computing. Computer 2011; 44, 46-53. [100] LF Chu, MJ Erlendson, JS Sun, HL Alva and AM Clemenson. Mobile computing in medical education: Opportunities and challenges. Curr. Opin. Anesthesiol. 2012; 6, 699-718. 16 Walailak J Sci & Tech 2018; 15(1) Communication Issues in Mobile Cloud Computing http://wjst.wu.ac.th Dasari Naga RAJU and Vankadara SARITHA [101] S Kosta, VC Perta, J Stefa, P Hui and A Mei. CloneDoc: Exploiting the cloud to leverage secure group collaboration mechanisms for smartphones. In: Proceedings of the IEEE Conference on Computer Communications Workshops. 2013. p. 19-20. [102] E Miluzzo, R Cáceres and YF Chen. Vision: mClouds-computing on clouds of mobile devices. In: Proceedings of the 3rd ACM Workshop on Mobile Cloud Computing and Services. 2012, p. 9-14. Walailak J Sci & Tech 2018; 15(1) 17 Copyright of Walailak Journal of Science & Technology is the property of Walailak Journal of Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. Hindawi Journal of Electrical and Computer Engineering Volume 2018, Article ID 8309450, 9 pages https://doi.org/10.1155/2018/8309450 Research Article A Security Monitoring Method Based on Autonomic Computing for the Cloud Platform Jingjie Zhang, Qingtao Wu , Ruijuan Zheng , Junlong Zhu , Mingchuan Zhang , and Ruoshui Liu Information Engineering College, Henan University of Science and Technology, Luoyang 471023, China Correspondence should be addressed to Qingtao Wu; wqt8921@haust.edu.cn Received 16 November 2017; Accepted 5 February 2018; Published 5 March 2018 Academic Editor: Vincent C. Emeakaroha Copyright © 2018 Jingjie Zhang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. With the continuous development of cloud computing, cloud security has become one of the most important issues in cloud computing. For example, data stored in the cloud platform may be attacked, and its security is difficult to be guaranteed. Therefore, we must attach weight to the issue of how to protect the data stored in the cloud. To protect data, data monitoring is a necessary process. Based on autonomic computing, we develop a cloud data monitoring system on the cloud platform, monitoring whether the data is abnormal in the cycle and analyzing the security of the data according to the monitored results. In this paper, the feasibility of the scheme can be verified through simulation. The results show that the proposed method can adapt to the dynamic change of cloud platform load, and it can also accurately evaluate the degree of abnormal data. Meanwhile, by adjusting monitoring frequency automatically, it improves the accuracy and timeliness of monitoring. Furthermore, it can reduce the monitoring cost of the system in normal operation process. 1. Introduction Resource monitoring in cloud computing environment is an important part of resource management of cloud computing platform. It provides the basis for resource allocation, task scheduling, and load balancing. With the extensive use of cloud computing services, users have made increasing demands on the security of cloud computing. Since the cloud computing environment has the characteristics of transparent virtualization and resource flexibility, it is infeasible for a traditional security program to protect the data security in the cloud platform, which hinders further development and application of cloud computing [1]. Therefore, it is of critical importance to develop new tools suitable for monitoring cloud platform data. However, the collection, transmission, storage, and analysis of a large number of monitored data will bring huge resource overhead, directly affecting system performance, timely detection of anomalies, and pinpoint accuracy of problem. In addition, because cloud computing is essentially developed on the basis of current technology, the existing security vulnerabilities will be inherited directly to the cloud computing platform, which may even bring greater security threat. It can be seen that, in the cloud computing environment, users basically lost the control of private information and data, which triggered a series of security challenges, such as cloud data storage location, data encryption mechanism, data recovery mechanism, integrity protection, third-party supervision and auditing, virtual machine security, and memory security. At present, there is not enough research on cloud computing resource monitoring, but there are a lot of researches on distributed computing and grid computing, for instance, DRMonitor [2], Ganglia [3], and MDS (Monitoring and Discovery System) [4]. They play important roles in distributed systems or grid systems. However, if the above methods are applied directly in the cloud computing environment, there will be some shortcomings. On the one hand, the resource in the cloud computing environment is highly virtualized and flexible. Moreover, cloud computing provides services such as IaaS, PaaS, and SaaS, in addition to monitoring the resources of the physical server [5]. Users need to monitor the virtual machine running on it. On the other hand, cloud computing is a business model, and the cloud service provider will charge the user for usage accordingly. Monitoring information in 2 Journal of Electrical and Computer Engineering existing resource monitoring system is not fine granularity, so it is unable to get to the process level of information and track consumption of CPU, memory, storage and other resources in real time during the user task execution process. Cloud computing environment is dynamic, random, complex, and open. Cloud providers need to collect user-related fees based on resource usage; as a result, original resource monitoring methods cannot fully meet the requirements of the cloud computing environment. Therefore, according to the characteristics of cloud computing itself, some resource monitoring methods for current distributed computing, and grid computing, cannot fully adapt to the cloud computing environment. In order to adapt to the cloud computing environment, combining with abnormal data mining algorithm, we propose a data monitoring method under cloud environment based on autonomic computing model. In order to address the security challenges for data on the cloud platform, the model uses autonomic computing mechanism and the abnormal data mining idea to transmit the monitoring information to each other. The model is mainly composed of five modules: network monitoring module, data analysis module, response strategy module, system implementation module, and knowledge base. In the network monitoring module, the system gathers the data by collecting the data stream and generates the original data. In addition, through the data preprocessing mechanism, the original data are formatted. The data analysis module evaluates these processed data, extracts useful data from it to determine whether they are abnormal, and then feeds the analysis result back to the response strategy module to adjust the monitoring period. The data collection and analysis of storage are the core parts of this model, which provide users with essential data monitoring information. In the local computer deployment monitoring framework, the cloud is connected to data monitoring. Our contributions are as follows: (i) We propose a safe and effective model that enables the data on the cloud platform to be monitored in time, and the system adjusts the monitoring cycle to autonomously protect the data. (ii) We design a data mining algorithm, in which, based on an improved chaotic algorithm, data mining method was proposed for the frequently appearing abnormal data in the cloud computing environment. We also design and implement abnormal behavior detection based on the Poisson process to obtain accurate test results. (iii) We formally analyze the capability of abnormal behavior monitoring and implement all of these data security monitoring models based on autonomic computing. A large number of experiments are carried out in the simulation environment using prepared dataset, and the results show that our system achieves the desired goals. This paper is organized as follows. Section 2 states the origin of autonomic computing theory, and its related work. Section 3 analyzes how to establish a security monitoring Sensors Analyse Monitor Effectors Plan Knowledge base Sensors Execute Effectors Managed resource Figure 1: MAPE autonomic computing model diagram. model based on autonomic computing for the cloud platform. Then we analyze the existing security model and safety monitoring method of autonomic computing to the cloud platform oriented metrics and the calculation method. Section 4 analyzes method of simulation and experiment and presents a security monitoring model based on autonomic computing analysis for the cloud platform. Section 5 gives the summary and points out the future research directions. 2. Related Work The concept of autonomic computing was proposed by IBM’s Paul Horn in 2001, which has self-configuring, selfoptimization, self-healing, self- protection, and other good features that have been accepted by the computer scientists [6, 7]. Autonomous computing refers to the computing environment with self-management capabilities, dynamically adapting to the increasingly complex environment, and selfdiscipline since calculation unit (Autonomic Computing Element) is an essential part of the autonomic computing system [8]. At present, the research on autonomic computing is based on a model of control loop proposed by IBM in 2003. It is called the MAPE-K (Monitor, Analyze, Plan, Execute, and Knowledge) cycle. Based on this model, a strategy module has been added to allow IT managers to facilitate the management of autonomic computing units. Its structure is shown in Figure 1. The monitor assembly means collecting information from the managed resource, that is, from the external environment. The analyze assembly is used to analyze the complexity of the internal environment of the system, so that the self-regulatory manager can understand the running state of the system in real time so as to predict the future situation and take right strategy for the future condition. The plan generates a sequence of actions that can be achieved based on the monitoring component and the analysis component’s data information from the external and internal environments, as well as previous policies. The execute is given to the effector to adjust the state of the managed resource. The actions generated by the above four components are all based on the knowledge in the knowledge base. Journal of Electrical and Computer Engineering An event classification method used in the fault monitoring of autonomic computing system was introduced by Liu and Zhou in 2010 [9]. In this scheme, the system monitors the status of heterogeneous resources failure. With the self-management system communicating internally, an appropriate strategy is activated to repair the fault for selfrepair system. However, there is a lack of research in the field of self-discipline for system performance failures. On the basis of studying the self-monitoring of self-discipline, the team proposed a multipoint detection method in 2011 [10]. Detecting the threshold cross-border and recovery to determine the system performance failure ensures the effectiveness of detection for the system, providing a useful strategy for the repairment failure. On the basis of studying the existing autonomic computing model, they proposed a self-discipline model which is suitable for distributed environment and formalizes the model elements of management resources, resource operation, state and action, and so on, to enhance the application of self-discipline model and practical value [11]. Among these advances of distributed computing, one must take into account the emergence of new paradigms such as cloud computing. In 2012, Yolanda et al. proposed a monitoring tool [12]. The goal of their work is to evaluate existing monitoring tools that can be used in cloud environments and are subsequently included in the monitoring component of the projects. In order to improve the efficiency of resource management in the cloud environment, Liu and Li proposed a selfregulatory model for the cloud environment [13], which uses the multi-autonomous manager’s hierarchical management model to solve the traditional auton...
Purchase answer to see full attachment