Hybrid Cloud and Fog Computing for SMEs in Ethiopia

By: Denekew A. Jembere

Licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Introduction

Reliable information communication technology (ICT) infrastructure enables small and medium enterprises (SMEs) to collect and process transactional and operational data to generate insights used for better decision making. Lack of affordable and reliable ICT infrastructure, on the other hand, causes SMEs to operate in a dark and drive their businesses using commonsense, with no knowledge about their current performance and predict the next competitive market, which is the case for SMEs in most developing countries, like Ethiopia. The lack of reliable ICT resources has a significant impact both on public and privately owned businesses in Ethiopia and requires to consider both technical and strategic options. Given the low-bandwidth & unreliable internet connection and the high-cost of using cloud computing for most SMEs in Ethiopia, the use of hybrid cloud and fog computing would have long-term cost-saving benefits for businesses and, increasing the headcount & quality of skilled workforce for the country. This article will, therefore, outline the background and impact of the lack of ICT resources, and propose a hybrid cloud and fog computing for enabling SMEs in Ethiopia to generate business intelligence insights from their business operations and customer transaction.

A brief overview of ICT in Africa

According to the World population review (Africa Population, 2017), Africa is the second-largest and second-most-populous continent on earth with an estimated population of 1.29 billion, in 2018. Ethiopia is one of the 54 African sovereign nations, located at the horn of Africa. Although Africa has such a huge population and is considered as an untapped market destination for the rest of the world, the ICT infrastructure is not as well developed as the rest of the world.  As a result of this, leveraging the latest ICT technologies, such as cloud computing, is a far-fetched reality.  

According to Jimenez-Castellanos (2012), although Africa is a continent with a wide spectrum of computational resources and infrastructures, businesses and institutions in the continent suffer a lack of resources and a poor Internet connection. Although the business dynamics and speed of technology are fast in most parts of the world, Africa trails behind. As a result of this, businesses and SMEs in Africa fail to gain an equal competitive landscape and wine markets. The success of SMEs a key factor in a nation’s economy and its employment rates.  

To alleviate the infrastructure problem and empower small and medium enterprises (SMEs) through the use of ICT, there are initiatives both at continental and national levels. In relation to this, the development activities being undertaken by the African Development Bank (AfDB) Group (n.d.) is believed to enable the nations in particular and Africa in general to leverage global ICT resources and services. However, until the multi-year huge national and continental AfDB projects get completed, each African country should develop a stop-gap solution to reduce the impact of the lack of well-developed ICT infrastructure. 

ICT in the case of Ethiopia

Understanding the global market landscape and implementing creative and custom technology solutions not only enables the business and SMEs to thrive but also ensures economic growth and healthy employment rates of a country.  In relation to this, according to the Embassy of Ethiopia in Brussels (2016), the Ethiopian government is undertaking a huge investment on the development of 11 industrial parks throughout the country. These parks are for both domestic and foreign SME investors to create employment opportunities for the Ethiopian youth, in manufacturing, and export goods and generate foreign currency for Ethiopia.

However, enabling these SMEs to be globally competitive, through automated business operations and services depends on the level and sophistication of ICT services provided in the parks. In relation to this, due to the poor internet service, the issue of leveraging technologies like cloud computing by SMEs in Ethiopia is not any different from most of the nations in Africa. Therefore, considering the global, continental and national ICT dimensions, this article attempts to propose a potential solution that would enable SMEs to maintain their competitive advantage in the context of Ethiopia.

The state of ICT infrastructure in Ethiopia

In the African context, the shortage of ICT resources in many African nations affects SMEs and costs them their global competitive advantages. Apart from pushing the SMEs out of the market, the impact will eventually increase the unemployment rate and affects the economy of the countries. While businesses in most developed nations leverage the latest technologies, such as cloud computing for their data management and business intelligence needs, SMEs in most nations in Africa have been suffering from the slow internet connection with limited bandwidth.

In the Ethiopian context, although the Ethiopian economy is the fastest growing economy in Africa (Wikipedia, n.d.), the youth which is the majority of the 108mln. Population (Worldometers, 2018) needs a sustainable employment opportunity. So, like any other emerging economy, maintaining a healthy economy in Ethiopia can be realized by providing the required resources and environment for SMEs.  In this regard, a cost-effective and efficient ICT infrastructure and data management solution could play a major role in the success of SMEs. Such solutions, for countries like Ethiopia with a poor internet connection, could be put together by amalgamating cloud and fog computing.

To understand the ICT challenges of SMEs in Ethiopia, it is worth assessing the state of the internet service in Ethiopian; and the attempts for the adoption of cloud. Moreover, it important to note the Ethiopian government’s effort in the development of intertrial parks in Ethiopia.

Internet service in Ethiopia

According to Wikipedia (n.d.), the sole Internet service provider in Ethiopia is the national telecommunications firm, Ethio Telecom. Ethio Telecom. (n.d.) claims that it has 65.7 million mobile customers; 17.87 million Internet & Data customers; and, 1.23 million fixed-line customers. However, this Ethio Telecom service coverage doesn’t breakdown the rate by distinct customers.  According to FreedomHouse (2018), despite being one of Africa’s fastest-growing economies, Ethiopia currently has a low internet penetration rate of 15 percent, which is greatly improved from a 3 percent penetration rate in 2015. 

The latest and current internet speed test averages (Dospeedtest, 2018), in the Ethiopian capital – Addis Ababa, shows a download speed of 14.17Mbps and upload speed of 5.59Mbps, with a latency of 508milliseconds. As per Dospeedtest, the Ethiopian government has plans to improve the internet speed and expand its access to rural areas in the country, using fiber-optic cables (broadband line). In relation to this, thousands of kilometers of fiber-optic cables are reported to have been installed in Ethiopia.

Ethiopia is part of the East African Submarine Cable System (EASSy) project funded by the African Development Bank (AfDB) Group (n.d.). This project involves the installation of a 10,000km submarine cable system along the east and southeast coast of Africa, to address the internet connection issue in most African countries. According to AfDB, with more than 10 Tbps of capacity, EASSy will act as a medium of internet connectivity carrying telecom traffic for all African operators from the Eastern and Southern African markets to connecting Cable networks in Europe, Asia, and the Americas. 

Although there is a country- and continental-wide efforts to improve Ethiopia’s internet access and connectivity issues, there is no recent official report that shows whether or not Ethiopia has been connected to the EASSy. As a result, the terrestrial backbone network remains limited and becomes a bottleneck to the use of cloud services, in Ethiopia.  Therefore, there should be a custom solution for SMEs, especially for those who are involved in manufacturing, to support and improve their business operations.

Cloud Adoption in the Ethiopian context

As a result of Microsoft’s donation of 250,000 laptops powered by Windows Azure, for schoolteachers nationwide (Chan, 2009), Ethiopia was reported to be the first to use Azure.  According to Chan, the laptops were to enable the teachers to download curriculum, keep track of academic records and securely transfer student data throughout the education system, without the need to build a support system of hardware and software to connect them. Apart from being Microsoft’s media hit of the time, before the Worldwide Partners Conference in which Microsoft was expected to announce more detail about Azure, there are no official follow-up reports on the positive or negative impact of the donation in Ethiopia’s cloud adoption. 

Multiple studies were carried out to investigate the cloud adoption challenges and provide guidance on readiness (Alemeye & Getahun, 2015) and architectural considerations (Hiran et al., 2018) in the Ethiopian context. These studies agree to the fact that the rolling blackouts and slow response times in the Ethiopian Internet backbone being the challenge hindering cloud adoption in the country. 

Until recently countries in Africa were never targeted to host Cloud data centers, as shown in Alguacil’s (2016) global cloud data centers map. Alguacil’s map shows the geographical coverage of Amazon Web Service (AWS), Microsoft Azure and Google Cloud with no cloud data center in Africa. In addition to this, the map of Gattuso (2018) showing the new and existing locations of Azure Content Delivery Network (CDN) shows no CDN location in Africa. Being at the horn of African, Ethiopia is in close proximity to some of the Azure datacenters and Azure CDNs in UAE and Europe. However, coupled with its communication infrastructure issues and the associated high cost of deploying cloud services on any of the cloud data centers has hindered the adoption of cloud in Ethiopia. 

Industrial Parks Development in Ethiopia

According to the Embassy of Ethiopia in Brussels (2016), the Government of Ethiopia has set Growth and Transformation Plan (GTP) policies and strategies to achieve an annual growth rate of 24 percent in the manufacturing sector and increase the sector’s contribution to export revenues from the current 10 percent to 25 percent. As a result, as per the Embassy, the Ethiopian government has been building 11 industrial parks on the land of size between 8.7 hectares (21.5 acres) and 1500 hectares (3706.6 acres), across the country.  These industrial parks have been intended to house textile, leather, food-processing and other labor-intensive factories, run by both domestic and foreign private investors.

Although some of the 11 industrial parks are inaugurated and operational (Matengo, 2018; Semonegna, 2017) the lack of specialized services including utilities, banks, customs, and transport links was reportedly causing a failure to meet the targeted textile export amounting to USD 1.5billion. While textile is just one of the impacted export line items, the negative impact of the lack of specialized services such as a reliable internet connection and cloud services on all of the export line items is not estimated and reported.

Hybrid Cloud and Fog computing

Mell and Grance (2011) of NIST (National Institute of Standards and Technology) defined cloud computing as follows:

“Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.”.

It is a common scenario for SMEs not to fully trust the cloud platform for migrating their on-premises applications and services and, operate their business from the cloud. While security and data governance-related concerns are the main reasons for SMEs in developed nations, the high cost of cloud services is the main factor hindering cloud migration for SMEs in developing countries. In this regard, a lightweight usage of the cloud platform, with a hybrid architecture (Figure-1) would enable SMEs to selectively use only necessary services from in the cloud platform, in a cost-effective way, while keeping their core business and data elements in their on-premise environment. This would also provide a stop-gap solution to the poor internet connection, by only uploading and downloading necessary data to and from the cloud environment.

Fog Computing

According to Iorga et al. (2018), fog computing is a layered model for enabling ubiquitous access to a shared continuum of scalable computing resources. Figure 2 below represents a fog computing in a broader context of a cloud-based ecosystem connected with smart edge-devices. As Iorga et al. explain, in such an ecosystem, the cloud services (from Provider A & B) are not a mandatory layer to support the functionality of the smart edge-devices functionality.

According to Butler (2017), as a network of data centers, fog computing can be a micro-data center in a small area as 100 square feet and can process or store critical data within the fog nodes.  This micro-data center can also optionally connect to a central (cloud) data center and push all or some of the data to the central data center.

The National Institute of Standards and Technology (NIST) recommends the architectural service model of fog computing to be implemented in a similar fashion as the traditional cloud computing architectural service model (Iorga et al., 2018).  Using the recommended architectural service model is claimed to enable the deployment of distributed, latency-aware applications and services, with physical or virtual fog nodes between the edge-devices and the optional centralized cloud service. The recommended architectural service models are:

• Software as a Service (SaaS) allows the fog service customer to use the fog provider’s applications running on a cluster of federated fog nodes managed by the provider.

• Platform as a Service (PaaS) allows deployment onto the platforms of federated fog nodes forming a cluster, of customer-created or acquired applications that are supported by the fog service provider runtime.
• Infrastructure as a Service (IaaS) allows the fog service customer to provision processing, storage, networks, and other fundamental computing resources leveraging the infrastructure of the fog nodes forming a federated cluster.

According to Nebbiolo (n.d.), fog computing runs applications in a multi-layer architecture that decouples the hardware and software functions, allowing for dynamic reconfigurations for different applications while performing intelligent computing and transmission services.  Moreover, with its hierarchical architecture, fog computing tends to be limited to a small number of peripheral devices and addresses storage, control and data-processing accelerations.

Shared computing platforms in the Industrial Parks

By leveraging shared hybrid cloud and fog computing platforms hosted within the industrial parks in Ethiopia, SMEs residing inside and outside the parks could satisfy their data management and data processing needs. Some of the advantages that such a shared infrastructure could provide are:

• Cost-savings: With proper infrastructure governance and security practices, such a platform reduces the amount of data to move between the cloud and on-premise platforms, whereby reducing the cost need for egress and ingress of data.
• Security of core business data: Such a hybrid platform facilitates to split data based on business-critical criteria and keep the business-critical data locally.
• Minimal latency impact: Since the fog computing platform is going to be in much closer proximity than any of the cloud data centers, the latency of data transfer is only dependent on the performance of the WAN or LAN, which can be improved.

References

Africa Population. (2017, November 12). Africa Population. Retrieved from Worldpopulationreview: http://worldpopulationreview.com/continents/africa-population/

African Development Bank Group. (n.d.). EASSy: The Eastern Africa Submarine Cable System. Retrieved from African Development Bank Group: Data Portal: https://www.afdb.org/en/projects-and-operations/selected-projects/eassy-the-eastern-africa-submarine-cable-system-156/

Alemeye, F., & Getahun, F. (2015). Cloud readiness assessment framework and recommendation system. AFRICON 2015, 1

Alguacil, P. (2016, November 24). Comparing the geographical coverage of AWS, Azure and Google Cloud. Retrieved from Atomia: https://www.atomia.com/2016/11/24/comparing-the-geographical-coverage-of-aws-azure-and-google-cloud/

Chan, S. P. (2009, July 13). Microsoft cloud computing gets down to earth. Retrieved from The Seattle Times: http://old.seattletimes.com/html/microsoft/2009458942_microsoftazure13.html

Dospeedtest. (2018, May). Broadband speed test Ethiopia. Retrieved from Dospeedtest: http://www.dospeedtest.com/statistics/speedtest-Ethiopia/

Embassy of Ethiopia. (2016, May 3). Industrial Parks Development in Ethiopia. Retrieved from Medium: https://medium.com/@EthiopiaEU/industrial-parks-development-in-ethiopia-f09eb704d741

Ethio Telecom. (n.d.). Bringing new possibilities. Retrieved from EthioTelecom: http://www.ethiotelecom.et/

FreedomHouse. (2018, June 1). Freedom on the Net 2018, Ethiopia. Retrieved from FreedomHouse: https://freedomhouse.org/report/freedom-net/2018/ethiopia

Gattuso, M. (2018, August 9). New locations for Azure CDN now available. Retrieved from Microsoft, Azure Networking:

New locations for Azure CDN now available

Hiran, Kamal K., Henten, Anders, Shrivas, Mahendra K., Doshi, Ruchi. (2018). Hybrid EduCloud Model in Higher Education: The case of Sub-Saharan Africa, Ethiopia. 2018 IEEE 7th International Conference on Adaptive Science & Technology (ICAST)

Iorga, M. F. (2018, March). Fog Computing Conceptual Model: Recommendations of the National Institute of Standards and Technology. Retrieved from NIST: https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.500-325.pdf

Jimenez-Castellanos, A., de la Calle, G., Alonso-Calvo, R., Hussein, R., Maojo, V. (2012). Accessing advanced computational resources in Africa through Cloud Computing. 2012 25th IEEE International Symposium on Computer-Based Medical Systems (CBMS)

Matengo, D. (2018, October 7). Ethiopia PM opens industrial park in Oromia region. Retrieved from CGTN Africa: https://africa.cgtn.com/2018/10/07/ethiopia-pm-opens-industrial-park-in-oromia-region/

Matt, C. (2018). Fog Computing: Complementing Cloud Computing to Facilitate Industry 4.0. Business & Information Systems Engineering: The International Journal of Wirtschaftsinformatik, 60(4), 351.

Matt, C. (2018). Fog Computing: Complementing Cloud Computing to Facilitate Industry 4.0. Business & Information Systems Engineering: The International Journal of Wirtschaftsinformatik, 60(4), 351.

Mell, P., Grance, T. (2011, September). Computer Security Resource Center. Retrieved from NIST: https://csrc.nist.gov/publications/detail/sp/800-145/final

Nebbiolo. (n.d.). Fog vs Edge Computing. Retrieved from Nebbiolo Tech: https://www.nebbiolo.tech/wp-content/uploads/whitepaper-fog-vs-edge.pdf

Plu, C. (2019, August 22). Everything you need to know about Hybrid Cloud to set foot in DevOps. Retrieved from PANDOK: https://www.padok.fr/en/blog/devops-hybrid-cloud

Rodriguez, S. (2017, July 10). Microsoft to offer a local version of Azure cloud service. Retrieved from Reuters: https://www.reuters.com/article/us-microsoft-azure/microsoft-to-offer-local-version-of-azure-cloud-service-idUSKBN19V1KQ

Sandoval, K. (2017, May 25). What is Fog Computing? Retrieved from https://nordicapis.com/what-is-fog-computing/

Semonegna. (2017, July 9). Kombolcha and Mekelle Follow Hawassa in Inaugurating Industrial Parks. Retrieved from Ethiopia Semonegna: http://semonegna.com/mekelle-kombolcha-industrial-park-inaugurated/

Strobel, J. (2017, October). Edge Computing: The Telco business models. Retrieved from Intel: https://builders.intel.com/docs/networkbuilders/edge-computing-the-telco-business-models.pdf

Worldometers. (2018, December 2). Ethiopia Population. Retrieved from Worldometers: http://www.worldometers.info/world-population/ethiopia-population/