e-Infrastructure and e-Services for Developing Countries

Many SMMEs fail within their first year of operation in South Africa mainly because of the lack of proper financial management skills. A number of attempts have been made by way of software applications; however, these attempts fail. This paper intends to design and implement a system that addresses some of the financial management challenges faced by SMMEs. To achieve this, features of similar systems were studies intensively through related work then incorporated into our system design and then implemented. The main objectives of the new system, was to make it possible for SMME owners to have access to their financial information anywhere, have access to real time data, reduce the amount of time needed to enter data in the system and a system that is easy to use. Usability testing was done and the result was a system that was 20% better in keeping records compared to the user’s manual accounting system.

This paper identifies and unpacks a troubling phenomenon whereby Africans have historically been and currently are de-facto excluded from processes that set mobile network standards such as 3G, 4G, and (now) 5G. It combines technical and procedural observations and colonial discourses of computing, concluding that enshrined systems and processes that steer the changes in mobile technology disempower African inputs and represent a continuation of the single use of situated techniques, skills, methods and processes in the production of core mobile technologies, all conceptualised outside Africa.

During the last decade, the design of social software involving the mobile phone contact list has been an active field of research. Systems, mostly those based on awareness, addressed many issues of social interest such as group communication or social ties building and strengthening. Yet, the mobile devices contact lists, as they are designed until now, do not efficiently take into account the social connections. Besides, the social connections are the principal provider of contact entries. Moreover, the current contact lists do not allow to efficiently retrieve/remember a contact who is forgotten or concerned by homonymy. Focusing on African social practices and behaviors, this paper proposed to redesign the contact list by integrating social relations that link people. This redesigning is accompanied with novel functionalities that will facilitate contacts retrieval, homonymy resolving and contacts remembering.

Since the Arab spring, the social media have become more popular among young people in Africa. Their uses have been illustrated during political and social events that occurred recently in the continent. Not long ago, Burkina Faso has been a field of important political events, the insurrection followed by the presidential election. In this work, we study how Facebook has been used during the 2015 presidential campaign. We use a language model to analyze many Facebook pages of political parties and media. We find that the campaign debates have been focused on campaign slogans and on the candidates rather than their political programs. Our findings show that the Facebook pages content reflect the election result. The most quoted candidate has been the winner.

Social networks (e.g. Facebook) are becoming ubiquitous, especially in the developed world. However, they rely on two fundamental assumptions (i.e., reliable Internet connectivity, literacy) that exclude scores of potential end-users, especially in the developing world. Including these potential end-users will require lifting those two assumptions. This paper is a first step towards inclusive social networks in the developing world. It proposes and validates a two-layer system architecture. The proposed architecture allows end-users to access the social network with low-end cellular phones, using voice and Short Message Service (SMS), in addition to traditional access via a computer or smart phone, using keyboards. We have built a proof of concept prototype in which poorly literate end-users are included in two simplified dedicated social networks (a small farmers’ social network and a motorcycle taxis’ social network).

In this work, we study the influence maximization in multilayer social networks. This problem is to find a set of k persons, called seeds, that maximizes the information spread in a multilayer social network. In our works, we focus in the determination of the seeds by proposing a centrality measure called Multi-DiffusionDegree (denoted by $$C_{dd}^{MLN}$$) based on IndependentCascade model. We consider the $$top-K$$ persons as the most influential. This centrality measure uses firstly, the diffusion probability for each person in each layer. Secondly, it uses the contribution of the first neighbors in the diffusion process. To show the performance of our approach, we compare it with the existing heuristics like multidegreecentrality. With software R and igraphpackage, we show that Multi-DiffusionDegree is more performant than the benchmark heuristic.

Cloud Computing is a new technology, widely used in different areas, allowing easy and on-demand access through the internet to a shared set of configurable computing resources. The adoption of this new technology within companies is restricted by security constraints. In this paper, we present the most significant risks that could affect organizations intending to deploy Cloud technology and security measures to be in place to reduce the impact of risks based on a literature review.

In this paper, we try to clearly outline the opportunities of cloud computing in the context of developing countries, while helping stakeholders grasp the challenges ahead. In particular, we consider the case of the G-Cloud project engaged by the government of Burkina Faso to boost the ICT economy, improve administration-citizen relations, and strengthen the education system. Through this article, we try to show the actors, that the G-Cloud to be a vector of opportunities and sustainable development for Burkina Faso, pre-conditions and challenges are facing particularly in terms of security in the cloud environment and the challenge on the supply in sufficient quantity and quality of electric power.

A new class of Virtual Private Networks (VPN), which supports both security and mobility, has recently emerged. Called mobile VPN, these systems provide not only secure tunnels but also session continuity mechanisms despite location change or connection disruptions. These mechanisms enable secure sessions to survive in dynamic/mobile environments without requiring a renegotiation of security keys during the session resumption phase. In this paper, we compare four open-source mobile VPNs in terms of functionality and performance.

In cloud computing service providers offer services to be used by customers. Given the increasing number of clients and also the variety of their needs, some users adjust services to their context. Also context may differ from those services. For this, we propose in this paper a set of tools (architecture and algorithm) for achieving this adaptation taking into account the specificities related to the context of the user. This set, knowing the services offered by suppliers, enables the user to make an adjustment using methods to construct the user context and the research service related to this context. The result obtained after the different tests carried out shows that our tool could be a better simulation environment of research and selection service adaption like in other environments.

Mobility and security are major features for both current and future network infrastructures. Nevertheless, the integration of mobility in traditional virtual private networks is difficult due to the costs of re-establishing broken secure tunnels and restarting broken connections. Besides session recovery costs, renegotiation steps also present inherent vulnerabilities. In order to address these issues, we propose a new distributed mobile VPN system called SEcured MObile Session (SEMOS). Based upon our CLOAK peer-to-peer overlay architecture, SEMOS provides security services to the application layer connections of mobile users. Secure and resilient sessions allow user connections to survive network failures as opposed to regular transport layer secured connections used by traditional VPN protocols.

In this paper, we propose an extension of the grammar of an application deployment constraints description language from a cloud computing platform. To do this, we draw a meta data model proposed by [1] for an application deployment in a cloud. This meta model, we extend j-ASD for the consideration of compatibility constraints or conformity between the virtual image data file formats used by the components of the virtual machines and those target sites that are deploying Smartphones, PC, etc. … Indeed for a full deployment of applications from a cloud environment on high mobility rate (Smartphone, PC, etc. …), it occurs to ensure compliance of the data formats of these environments to that of Cloud platform. This conformity proves a prerequisite for deployment on a device from a cloud. To address these compliance constraints, we formalize them in matrix form and propose the use of a constraint solver.

Long-range radio are promising technologies to deploy low-cost Low Power WAN for a large variety of IoT applications. There are however many issues that must be considered before deploying IoT solutions for low-income developing countries. This article will present these issues and show how they can be addressed in the context of African rural applications. We then describe the WAZIUP low-cost and long-range IoT framework. The framework takes cost of hardware and services as the main challenge to be addressed as well as offering quick appropriation and customization possibilities by third-parties.

Applying the capabilities of the Internet of Things holds particular promise for advancing certain of the United Nations Sustainable Development Goals. For instance, citizens in many developing countries lack access to energy, and suffer from a lack of energy independence, scarcity, and high energy costs. Efforts are underway to make energy generation, transmission, and consumption more efficient via the application of modern communications and control. In this paper we focus on a communications and demand-side control application for legacy appliances. Instead of replacing existing devices with new “smart” devices with better communications capability, we integrate legacy devices into a centralized control system. We designed and implemented a communications system that allows control of infrared-enabled appliances over a network using wired or wireless communications. We designed a web application and user interface that can function remotely over the Internet, and use microcontrollers and infrared transmitters to communicate with infrared-enabled devices. We demonstrate a prototype of the system to control air conditioning units in a commercial building for energy management. The system uses open source hardware, software, and protocols. While we use energy management as an example, we focus on the general communications and control architecture, which is easily scalable and applicable to other devices and applications beyond those demonstrated here.

Increasing numbers of pipeline breakdown experienced by utilities undoubtedly raise alarms concerning the anticipated failure consequences. Seemingly mild, these consequences can however, fluctuate to severe or fatal, especially in high risk locations. Utility personnel are therefore pressured to employ up-to-par operational policies in attempt to minimize possible fatalities. This however, may be overwhelming considering inherent uncertainties that make it difficult to understand and adapt these consequences into utilities’ risk management structure. One way of handling such uncertainties is through the use of Bayesian Networks (BNs), which can comfortably combine supplementary information and knowledge. In this paper therefore, we present an overview of the causes and impacts of pipeline failure. We aggregate and classify failure consequences in a select high risk zone into four indexes; and finally, we outline how BNs can accommodate these indexes for pipeline failure prediction modeling. These indexes function as effective surrogate inputs where data is unavailable.

Crowdfunding has been growing rapidly as a new financing source for ventures. To support the application of crowdfunding for the agricultural sector, this study explored the characteristics and crowdfunding performance of agricultural projects. A total of 112 projects were selected from one of the largest crowdfunding platforms in Asia. The results showed most of the agricultural projects were grouped under the categories of leisure, public, or local. Within two months of the funding duration, each of these agricultural projects pledged an average of 6,880.9 USD from 123 contributors. But the total share and the overall success rate of agricultural projects remained low. Furthermore, project category, funding target, and number of contributors had significant effects on the crowdfunding performance of the agricultural projects. The findings suggest project initiators may need to diversify project categories, set a relatively low funding target, and expand social network to increase number of contributors.

Disaster and emergencies are sometimes difficult to manage in terrains where there are limited communications. Some geographical areas can prove to be difficult during disaster areas to reach. During disasters, lack of communications infrastructure can hinder operations involving coordination, effective reporting and communications of events on the grounds for delivery of services such as food, medicine, etc. In this paper we proposed and developed an integrated radio over Internet Protocol Network for easy integration into other networks for effective communications, deployment and management of information. We integrated the system over satellite, phone and other alert voice systems for effective disaster management. Results showed ease of usage was effective and it is suitable for easy deployment and management of disaster and emergencies.

TV White Space networks are gaining momentum worldwide as an important addition to the suite of wireless protocols available for connecting developing regions. However, there has been no thorough investigation of scenarios where TV White Space performs better or worse than alternative low-cost wireless technology such as WiFi. This paper analyzes the performance of 5 GHz WiFi links and TV White space links using down-converted WiFi, typically used as wireless backhaul for poorly connected regions, in different scenarios including line-of-sight links and links obstructed by trees and structures. The experiments make use of 802.11a/b/g WiFi and TV White Space equipment that downconverts standard 802.11 a/b/g WiFi from the 2.4 GHz band into the UHF band. The paper finds that 5 GHz links outperformed TVWS where clear line-of-sight is available and point-to-point links are required. TVWS however is a clear choice where there are obstructions and where wider coverage is needed. Some interesting observations on the negative effect of TV transmissions in adjacent channels a few channel-hops away from the channel being used for TVWS are also provided.

In this paper, we use the shooting and bouncing ray/image (SBR/Image) method to compute the path loss for different outdoor environments in the residential area. Their corresponding path loss of the outdoor environment for the non-line-of-sight (NLOS) cases are calculated. Numerical results show that the performance in reduction of path loss by DDE algorithm is better than that by GA. Besides, the path loss for the circular shape antenna is lower than that for the cross shape antenna array both in NLOS case.

The ever increasing number of users in Wireless Mesh Networks (WMNs) setups consequently represents an upsurge in competitions for available services. Consequently, services are clogged and ran over WMNs, which further leads to poor Quality of Service (QoS). Quick and timely discovery of available services becomes an essential parameter in optimizing performance of WMNs. In this paper therefore, we present a Priority-based Service Discovery Model (PSDM) using Swarm Intelligence in WMNs. We use the Particle Swarm Optimization (PSO) algorithm to dynamically define and prioritize services supported by the network. Additionally, the Ant Colony Optimization (ACO) algorithm is used to choose the shortest path when each transmitter has to be searched to identify if it possesses the requested services. We have designed and implemented the PSDM using Network Simulator 2 (NS-2) tool. Consequently, we realized throughput of 80%, service availability of 96% in some instances, and an average delay of 1.8 ms.

The advanced routing protocol not only operates on lower levels of a network protocol, but it also operates on upper layers such as the application layer of OSI model. The routing task can be operated on a wider scale. It can process based on results of some other protocols for example service based protocol can be operated based on the service discovery protocol. So this kind of routing protocols may be determinized as upper layer routing protocols. Such as the service based routing protocol can operate based on content based protocol and combines some service filters. In service based routing protocol as well as content based routing protocol, subscriber and publisher can communicate with each other but they don’t know the other’s address. So it is more flexible in processing and more comfortable for mobile ad-hoc network. In mobile ad-hoc networks, nodes usually move, so bandwidth of connection between them may be not stable. Therefore transmission delay, overhead and packet loss may be larger than other kinds of networks. The paper aims at purpose to increase QoS of routing by hierarchical clustering routing by using $$ R^{ + } $$ tree in addition with some advanced techniques such as multicast routing, multiple paths, use can ACO to optimize routes to transmit data. By using R tree structure, the network topology are managed by bottom-up model from leaf level to root of the tree. All the leaf nodes, inner nodes and root of this tree have two roles: (i) Manage a cluster that consisting all nodes that have direct connections with this node; (ii) Operate as a normal node. The paper introduces and analyzes: (i) Establish hierarchical clustering network by using R tree structure; (ii) Make multicast tree from some cluster heads for fast routing; (iii) Make optimized route by Ant Colony Optimization. The paper also uses the Artificial Neural Network to choose optimal cluster head and members for cluster of network.

The DNS, as one of the oldest components of the modern Internet, has been studied multiple times. It is a known fact that operational issues such as mis-configured name servers affect the responsiveness of the DNS service which could lead to delayed responses or failed queries. One of such misconfigurations is lame delegation and this article explains how it can be detected and also provides guidance to the African Internet community as to whether a policy lame reverse DNS should be enforced. It also gives an overview of the degree of lameness of the AFRINIC reverse domains where it was found that 45% of all reverse domains are lame.

An investigation into the Received Signal Strength Indicator (RSSI) dependency on receiver antenna height in UHF band is conducted. The results show a high correlation between RSSI and height on channels with high signal strength. There is approximately 2.5 dBm RSSI gain per 1 m increase in height above ground up to 8.5 m. From 8.5 m to 12 m, there is no consistent observable increase in RSSI. Furthermore, the geolocation spectrum database’s (GLSD) view of white space in the television band is compared with the ground truth. Results show signal presence on some of the channels indicated free by the spectrum database. These findings imply that an increase in transmission range of UHF links can be achieved by increasing receiver height. White space devices using A GLSD should additionally require spectrum scanning to determine clear channels.

Online platforms are becoming common and a way to enhanced interaction and wider reach. However, micro-entrepreneurs have generally not embraced online services for business development and support. Micro-entrepreneurs have largely failed during infancy owing to factors like technology use. Even with the business advisory service that is widely preached by government and non-government providers, micro entrepreneurs are still challenged. Concerns of business advisory outreach are largely raised. Using a quantitative approach the study sought to identify the advisory services needed, those accessed online and the reasons for using on not using the existing online platforms to access business advisory. The study was narrowed down to micro-entrepreneurs operating in the textile industry. The study provides insight about the state of internet-based business advisory in the micro-entrepreneurial society of Uganda. It was generally noted that micro-entrepreneurs prefer physical access to online access of business advisory although they acknowledge the benefits associated with online platforms. Limited knowledge and skills on how to use these platforms coped with the busy daily schedule limits the time available to search for advice online.

Knowledge is recognized as the most important resource in organisations including public organisations and its management is considered critical to organizational success. The literature suggests the development of indigenous knowledge as one of characteristics of the successful GIS implementation in developing countries. The topic of knowledge has been discussed extensively in the information system and organisation literature but much is written about why managing knowledge is important to organisations and little on how knowledge is identified, captured, shared, and used within organisations. As a contribution to ‘how’, this paper discusses opportunities of sharing knowledge in the GIS implementation in health management through some initiatives in Malawi. We can confidently say that there are a number of GIS implementation activities in the health sector in Malawi which are important for knowledge sharing but they are not utilised as expected.

In large classes of enrolled students, Polytechnic University of Bobo Dioulasso (PUB) can not complete one academic year in 12 months. The major obstacle is the lack of infrastructure and qualified local teachers for conducting parallel coaching tutorial groups. These tutorials are carried out in group of up 50 students. The academic year can last 18 or 24 months. To address this problem, we propose a techno-pedagogical system.Technically, this project consists in strengthen the capacities of the PUB in IT resources and optimize their use. A Learning Management System (LMS) is implemented on the local network. The campus network, accessible from all sites of the PUB and student residences, allows access to the server at a rate of at least 10Mbps. The LMS is also available on Internet. On the pedagogical and organizational level, we conducted training for content experts and tutors in design and modular structure of courses, tutoring in distance learning. A team of two techno-pedagogues are empowered to manage this training system. Theoretical courses are run-face. Group works are driven remotely with the help of remote experts. This project would enable, in two years, to reduce the duration of an academic year to 12 months.

Cloud computing has provided tremendous opportunities for a lot of organizations in cutting operating cost, making data available to distributed units, provision of easy interoperability etc. Health information systems forms a critical parts of one’s countries information technology infrastructure due to the sensitivity and nature of data processed over time with regards treatment history, medical records etc. And medical images form dominant part of the sensitive patient data. Hence privacy and security needs to be guaranteed for such images stored in the cloud. Most of the access security and encryption approaches are left for the cloud owners to manage and these poses a lot of insecurity if the system is compromised. In our wok, we proposed a hybrid spatial cryptographic and watermarking technique for authentication and security of medical images before storage in the cloud. Due to the sensitive nature of medical images, we were able to achieve full recoverability of the plain image after decryption and dewatermarking without pixel loss. Our results showed to be very effective and reliable for fully recoverable images.

Security in internet of things is very critical. The computer systems that use image acquisition and processing from onboard or remote systems need to secure image data over the internet. Security of multimedia image data transmitted over secured and unsecured communications channel in today’s cyberspace is of paramount consideration due to malicious activities over these channels and there has been a high demand especially for end to end security applications for mobile devices. This is to ensure safety and security as well as privacy to transmitted data and stored data in the clouds. In this paper, we proposed a hybrid approach for securing digital images for devices by engaging lossy compression using 2-D discrete cosine transform and Visual Cryptographic Technique. The DCT was used for the lossy compression process and the visual cryptography was used to encrypt the image for confidentiality. The results showed to be very effective and the implementation was done using MATLAB.

The modernization of African traditional medicine (TM) using IT faces to illiteracy of most of the domain stakeholders. In order to assist traditional medicine practitioner (TMP) in theirs activities, we have propose an icon-based system to visually use plants and recipe in the drug preparation process. Therefore, traditional physicians can easily combine icons for medical prescription. For that, ontoMEDTRAD is an ontology including formal description for knowledge related to iconic representation of plants and recipes. Structurally, ontoMEDTRAD includes two modules: ontoConcept_term and ontoIcone denoting respectively the terms and the icons of concepts in this domain. Thus, avoiding any semantic issues, TMP can be free from language barriers, textual writing and reading in their work of healer. More specifically, this work aims to model plants and recipes in TM and propose compositional iconic language for plants and sketches for recipes.

Deploying NREN infrastructures is considered a crucial step for fostering Science and Technology in Africa as it will allow to support advanced services for sharing science. For managing science components and players, the Current Research Information System (CRIS) has been proposed and adopted internationally. Understanding these concepts and applying them in the science ecosystem of developing countries will allow to foster science management in these countries, saving years of progress.Facing the above, this paper discusses the challenges and strategies for the implementation of technological platforms for accessing science, taking Mozambique and its NREN as the primary goal. Answering questions such as “What is the current state of Science and Technology in Mozambique? What are the challenges and strategies to put Mozambique a step forward in science development?” is the aim of this research work. The proposals envisioned in this paper are sustained by the analysis of worldwide initiatives and recommendations regarding open access to science, CRIS inputs, the Portuguese and African context, and the Mozambican science and education ecosystem. This work also aims to provide insights for other developing countries by identifying good practices and international cooperation opportunities.

Network protocols and applications have mostly been developed in and for a Western context and usually have an embedded set of assumptions about network performance and availability. As a result web-browsing, cloud-based services, live voice and video over IP, desktop applications and software updates often fail or perform poorly in (rural) areas of Southern Africa. This paper uncovers some of the reasons for this poor performance such as Windows TCP failing to reach capacity in high-delay networks, long DNS delays or time-outs and applications such as Office365 assuming constant connectivity to function, and describes them, set in the Southern African contexts. We address the issue of colonisation in ICT context and show the extend of such in the area of networking. These observations provide strong motivation for Africa-based engineering research to ensure that future network protocols and applications are context-sensitive, adaptive and truly global.