ABSTRACT. The paper presents research findings from a simulation study on technical power loss reductions in selected outgoing feeders of 15-kV to distribution transformers in one Sub-station of Ethiopia’s growing power system. In the simulation process for modeling the power losses in the medium voltage to low voltage (MV/LV) transformation sub-networks, the well-known software DIgSILENT has been employed. Combinations of power loss sources were considered for testing the performances of four selected feeders from Sebeta I Substation to parallel distribution transformers supplying electric power to consumers’ centers mainly in south west Addis Ababa. Maximum loading of the outbound feeders from the Substation to the MV/LV distribution transformers located at different load centers were examined during peak load hours. Along the selected routes, with all transformers loaded to their 60% capacities, the total nominal power losses in the feeders and the transformers were also appraised. Alternative sources of technical power losses have been studied by considering uses of conductor segments and power factor improvements with different loading considerations. Further simulation studies supported by power loss measurements and relevant applications of information communications technologies for sustainable development are also strongly recommended.

ABSTRACT. With the world progressively heading towards an energy crisis, renewable energy sources and technologies have become the focus of research. Harvesting the sun's energy through solar thermal power plants (STPPs) is an attractive alternative to the use of energy supplied through the grid. However, installation and maintenance of cabling for communication and control of a heliostat array for STPPs is costly. In this paper, a wireless heliostat system based on a wireless sensor network is proposed. The system is able to accurately track the sun’s position, and cooperatively focus the sunlight onto a target plane using wireless communication for signaling. The latency introduced by wireless communication has no significant impact on system performance, and the proposed system is robust to tilt and wind disturbances.

ABSTRACT. This paper presents design, simulation and study of different noise minimization techniques that should be used in chopper stabilization circuits for CMOS-MEMS sensor readout interface circuit applications. Chopper stabilization technique is widely used in CMOS-MEMS sensors, as it is effective in tackling low frequency noises including the most dominant, flicker noise (1/f). In CMOS-MEMS sensors where noise is one of the main design targets, it is necessary to minimize noise arising from the chopper stabilization circuit itself to boost the overall noise performance. Chopper circuits contribute part of the total input referred noise and appropriate techniques need to be deployed in place to minimize the overall effect. Here in this work, different techniques which should be considered to reduce noise from the chopper circuits are discussed and analyzed. The demonstrated work is designed in MIMOS 0.35 μm AMS CMOS 3.3 V CMOS process technology and Cadence Spectre circuit simulator is used to verify and analyze the circuit performances.

ABSTRACT. This paper investigates the dynamic power requirements for controllable energy storage systems used for stabilization of standalone three-phase low-voltage microgrids with high penetration of PV energy. It uses recorded data of fluctuating Photo-voltaic (PV) sources and fluctuating loads to determine how an ideal controllable energy storage can be used to stabilize a microgrid with 100% PV energy penetration. Using the data recorded for a typical bad day in Johannesburg, the sampling time required to capture the dynamics in PV sources was found to be 2 seconds. However this report uses sampling time of 1 second to calculate the ramp rate required for the energy storage to stabilize a PV microgrid. For the case considered in this paper, the ramp rate was found to be 0.4 pu/second.

ABSTRACT. Micro-grid for villages using renewable energy is becoming a trend for developing countries like in Africa. Commonly, ac microgrids are used for the simple reason of ac household equipments and protection abilities. DC microgrid is a better choice in terms of efficiency, control simplicity and converter prices. However, whether it is an AC or DC, the demand-supply management is the most basic problem in developing countries which causes more outage due to overloads and/or equipment failures. This paper proposes, a weakly cou- pled dc grid, a systematic approach to building electrical grid with the consideration of social behavior, future power system considerations and social participation in decision making.

ABSTRACT. In this paper, the authors have carried out fault analysis in Power Transmission Lines using Wavelet Multi-Resolution Technique. Different Realistic fault cases at different locations in the transmission lines are considered for this fault investigation. Adequate details and explanations are given for each assumption made for more clarity and better understanding. The authors explain various assumptions made by many other different authors in their work with the aid of enough MATLAB simulation results. In this paper, a two terminal transmission line fed by two generators is used for the fault investigation. The authors compare the results obtained with that from the travelling wave method. The authors believe that the proposed technique here is generally applicable to determine fault in any multi terminal transmission lines, if the voltage and current value at each bus is known – these are now readily available with the help of phasor measurement units.

ABSTRACT. This paper deals with a protection circuit against an overvoltage caused by a sudden battery removal from the dc link terminals. The system can be protected against this overvoltage by opening the neutral point of the generator and by engaging a dissipating resistor across the dc link terminals. This method is successfully tested and validated.

ABSTRACT. several experimental tests were conducted to investigate the performance of the solar tunnel dryer for drying of ginger. The dryer is arranged to supply hot air to the drying chamber using two axial flow fans, with a power rating of 28W, powered by a 50W PV module. No-load and load tests on the dryer, with solar PV operated fans have been conducted. The drying parameters like solar radiation, drying air temprature, ambient temprature, relative humidity and air velocity are recorded at five solar tunnel dryer positions by dividing the 8.50 meter long solar tunnel dryer in to four equal parts in every thirty minutes interval. In the no-load condition, the hot air temperature at the collector outlet for 8 hour operation raised from 34.00C to 65.50C when the solar radiation was varying between 540 and 820 W/m2. On the other hand, an average average maximum temperatures of300C (morning) and 770Cat 13:00pm was achieved at While, open sun drying has taken 40 hours (5 days).The solar tunnel dryer showed a net saving in drying time of 40% over open sun drying with solar collector area of 6m2. As the collector area increases, the drying time decreases and the drying efficiency and the saving in drying time increases. In addition, the gingers being dried in the solar tunnel drier were completely protected from rain, insects and dust, and highquality dried gingers were obtained. As the fans are powered by a solar module, the drier could be used in rural areas where there is no supply of electricity from grid.

ABSTRACT. It is very critical for solar racers to achieve real time tracking of their solar vehicle during any race. The system is made of real time visualisation of the video feedback of the car from the escort vehicle, the location of the car on the maps as it is racing and its velocity. The real time video feedback of the race could be retrieved from the cloud for broadcasting purposes. The main advantage of this system is the wireless link communication which was achieved for distance transmission, coupled to a very reliable and robust desktop application for data sensor display purposes.

ABSTRACT. The design of innovative technological instruments and frameworks for smart energy management is a challenge for countries across the world, and the creation of a 'Smart Grid' still has unresolved research and development (R&D) problems. The design of optimal distributed energy resources (DER) is a promising approach for creating more renewable energy. In order to successfully implement DER, one must first consider several technical performance issues that would benefit from applied research. In this paper, a distributed autonomous energy management system (DAEMS) will be introduced. The DAEMS is envisaged to be self-organising, scalable and a reliable component for energy trading, client-side usage optimisation and DER management. Therefore the focus of this paper will be on creating a technical platform for facilitating energy trading, thereby creating an energy market and empowering consumers within the green economy. The creation of such technical platform can be achieved by formulating algorithms for client-side energy efficiency, and finding suitable methods for real-time DER management in terms of energy consumption, generation and connectivity

ABSTRACT. This paper presents the temperature based models for predicting the daily average global solar radiation of Ibadan. The data used in developing the models consist of daily average global solar radiation, minimum and maximum daily average temperature collected over a period of 9 years at International Institute of Tropical Agriculture (IITA) located in Ibadan. Different variation of temperature model are derived and compared. The suitability of the models is ascertained using standard statistical tests: Mean Bias Error (MBE), Root Mean Square Error (RMSE), Mean Percentage Error (MPE), and Correlation Coefficient (R). One of the key results shows that a quadratic temperature model present the overall best model for predicting the global solar radiation of Ibadan with MBE of 1.86 , RMSE of 2.7, MAPE of 9.34% and R of 0.68

ABSTRACT. Broadband connectivity is Africa is estimated to be less than 10%. There is also a digital divide gap between the rural areas and urban areas in Africa. TV white spaces (TVWS), together with dynamic spectrum access (DSA), offers a great promise in bridging the digital divide gap in Africa and making broadband connectivity affordable and ubiquitous. TVWS offers a great promise because it has low risk regulation, there is plenty of unused spectrum in Africa and TVWS is a great rural technology. Some of the application areas of TVWS and DSA are delivery of government services, education and telemedicine. Trials have been conducted across Africa to demonstrate that TVWS technology can be used to deliver low cost broadband services, to increase awareness of the potential of TVWS technology and to provide information about costs involved in the technology. TVWS trials in Kenya and South Africa have demonstrated that the performance of a TVWS network compares to that of a Wimax network. In order for TVWS networks to be operational, TVWS and DSA regulations are required. The success of the use of TVWS in Africa depends on the national authorities across the continent in enacting the necessary regulations so as to enable the use of this promising technology.

ABSTRACT. Television white spaces (TVWS) refers to a portion of the ultra-high frequency (UHF) band (between 470 and 690 MHz) which is assigned for television broadcast but not being used in a given location at a specific time. There is a growing evidence that broadband TVWS networks can co-exist on the same radio frequency spectrum (RF) band with licensed broadcasting services for as long as interference to licensed users is minimized efficiently. This paper introduces the Limpopo TVWS trial network, which is one of the TVWS trials in South African aimed at providing Internet connectivity to five secondary schools in rural areas through a point-to-point network configuration. A 24 hours RF monitoring results of the TVWS trial network are presented which shows the behaviour of white space devices with respect to incumbent TV band services. The paper also presents TVWS trial network performance with respect to average throughput, average latency and jitter for both the uplink and downlink connections. The overall TVWS trial network performance shows good throughput for each TVWS link which is in the magnitude of 4 to 8 Mbps and low latency of less than 10 ms for typical packet sizes ranging from 32 to 1500 bytes.

ABSTRACT. This paper presents a gap mechanism for evaluating the performance of three energy minimisation models in wireless multicast networks. The first model is the existing coded packet that is based on simulation technique. The second and the third models are the recently proposed techniques that are based on data envelopment analysis (DEA) methodology. The input-oriented Banker, Chanes and Cooper (BCC) and Chanes, Cooper and Rhodes (CCR) approaches are considered for DEA model development. It is fund that the proposed input-oriented BCC and CCR models provide better energy gap in wireless networks compared to the coded packet model.

ABSTRACT. We explore the use of ultra-capacitors as backup storage for solar powered electronic devices. We discuss two use cases: routers in off-grid broadband networks consuming a few Watt and wireless sensor network nodes using a few mW when operating and a few μ W in quiescent mode. We have been testing batteries with ultra-capacitor based cells, both EDLC and LIC, in the range from 40 to 3000F. We found them viable alternatives in some applications and started long term field tests to sort out some unclear aspects regarding maximum discharge levels. Besides the traditional function of a backup battery storage in low-power applications, the very low internal resistance of capacitors make them useful also in hybrid solutions to accommodate transient needs for high currents that traditional batteries cannot deliver. Due to the low resistance, charging/discharging can be very fast. We also show how a very simple design requiring only one LIC capacitor can power a WSN mote sending at 1 packet/minute for 6 weeks on a single charge.

ABSTRACT. The ability to estimate positions of sound sources is one that gives animals a 360deg awareness of their acoustic environment. This helps complement the visual scene which is restricted to 180deg in humans. Unfortunately, deaf people are left out on this ability. Smart phones are rapidly becoming a common tool amongst mobile users in developed and emerging markets. Their processing ability has more than doubled since their introduction to mass consumer markets by Apple in 2007. Top-end smart phones such as the Samsung Galaxy Series; 3, 4, and 5 models, have two microphones with which one can acquire stereo recordings. The purpose of this research project was to establish a feasible Sound source localization algorithm for current top-end smart phones, and to recommend hardware improvements for future smart phones, to pave way for the use of smart phones as advanced auditory sensory devices capable of acting as avatars for intelligent remote systems to learn about different acoustic scenes with help of human users.

ABSTRACT. In brain tumor diagnosis, clinicians integrate their medical knowledge and brain magnetic resonance imaging (MRI) scans to obtain the nature and pathological characteristics of brain tumors and to decide on treatment options. However, manually detecting and segmenting brain tumors in today’s brain MRI, where a large number of MRI scans taken for each patient, is tedious and subjected to inter and intra observer detection and segmentation variability. As result a number of methods have been proposed in recent years to fill this gap, but still there is no commonly accepted automated technique by clinicians to be used in clinical floor due to accuracy and robustness issues. In our approach, an automatic brain tumor detection and segmentation framework that consists of techniques from skull stripping to detection and segmentation of brain tumors is proposed with fuzzy Hopfield neural network as its final tumor segmentation technique. Through preprocessing, image fusion and initial tumorous slice classification, the final hybrid intelligent fuzzy Hopfield neural network algorithm based tumor segmentation, and tumor region detection and extraction is achieved. The performance of the proposed framework is evaluated on various MR images including simulated and real, normal and tumorous. Quantitatively the method is validated against available ground truth using commonly used validation metrics. The final segmentation mean and standard deviation result in Jaccard similarity index, Dice similarity score, sensitivity and specificity are 0.8569+/-0.0896, 0.9186+/-0.0638, 0.9480+/-0.0402 and 0.9917+/-0.0387 respectively. Quantitative and qualitative segmentation result indicates the potential of the proposed framework.

ABSTRACT. Hybrid reasoning systems are increasingly used to improve the quality of medical services. Physicians can use the hybrid reasoning systems to diagnose a particular patient’s disease. Local languages can be integrated with hybrid reasoning systems to allow end-users communicate with the system in a simpler and easier way. This study presents localized hybrid reasoning systems that uses a combination of rule based reasoning (RBR) and case based reasoning (CBR) techniques using Ethiopian national language to achieve TB disease diagnosis. To develop the localized hybrid reasoning system, knowledge is acquired from documented and non- documented sources. The localized hybrid reasoning system is developed using SWI Prolog version 6.4.1 programming language. The system is tested and evaluated to ensure that whether the performance of the system is accurate and the system is usable by physicians and patients. The localized hybrid reasoning system (LHRS) has an average accuracy of 85.5%.

ABSTRACT. This paper presents the design of a smart geyser controller that can relay information between two functional units making use of low power IEEE 802.15.4 radios, with the objective of lowering the power consumption of a household geyser. The first operational unit measures the environmental data of the geyser, i.e. water temperature, voltage, current and power measurements, and also switches the geyser on and off. The other functional unit is used to display information to the user and also connects to a PC where environmental data is stored in a local database for historical review. This database is used to create a priority time table, which forms the basis of the operational optimization algorithm.

ABSTRACT. As demand for mobility continues to rise accompanied by environmental and energy problems, electric vehicles (EVs) have been heralded as imminent long-term solutions to sustainable transport electrification. Battery electric vehicles in particular offer zero tailpipe emissions. However, they can only be considered as true GHG emissions mitigating option if the electricity used to charge their batteries is generated through low carbon technologies. This paper deals with the influence of the energy mix for the effective contribution of EVs in carbon emission reduction. Moreover, it compares the life cycle of the CO2 emissions in electric vehicles powered with coal dominated and renewable energy mix. The same assessment has been done to comparable conventional petrol cars. The factors affecting the energy consumption of electric vehicles and the impacts of vehicle electrification on an existing grid and cost have also been discussed. The discussion on this paper is based on the energy mix of 15 African countries in general and Ethiopia in particular.

ABSTRACT. ABSTRACT—In this paper, we investigate the design and simulation of an electronic system for automatic controlling of water pumps that are used for agricultural fields or plant watering based on the level of soil moisture sensing. The detected signal from the soil moisture sensor is processed by a conditional comparator circuit corresponding to different levels of actual soil moisture content. A logic circuit follows the conditional circuit with its output signals used to activate a system of relays that control the power circuit of the motors used for water pumping. The speed of the motor is varied according to the level of the soil moisture content; the motor is OFF during maximum wet and is running with HIGH speed during dry soil conditions respectively. The duration of water pumping is controlled by a timer circuit where the timer can be designed according to the desired watering time. The different stages of the overall electronic system are simulated and tested using NI MULTISM simulation software. The technique is useful to save human-power, enhances crop or plant productivity, saves water and provides proper water management to ensure efficient and viable farming industry especially in those places that receive less rain or mostly dry areas.

ABSTRACT. As a contribution to human safety and survival, the design, operation and experimental implementation of a fire detection system with a smart audio-visual emergency response interface is presented. This paper describes how real-time scientific data and instructions can be relayed to multiple computer devices in the case of a fire emergency. The experimental results showing the effectiveness and reliability of the system are reported.

ABSTRACT. Cognitive radio networks (CRN) has been depicted as one of the most important driving forces in achieving next generation wireless communication capabilities. To this end, several works are being done in describing and designing this new wireless communication paradigm. In this work, a study on how dynamic resource allocation can be used in achieving the desired quality of service (QoS) in an heterogeneous CRN environment is carried out. An underlay CRN model is developed where, in the resource allocation formulation, the demands of secondary users differ but must be met concomitantly. The objective is to maximize the weighted data rate of the different categories of users in the CRN, and by so doing, achieve some required QoS. The developed problem is both non-linear and also non-deterministic polynomial-time (NP) hard. However, the problem is smartly re-formulated as an integer linear programming problem and the branch-and-bound approach is employed in obtaining optimal solutions. The QoS metrics evaluated are average data rate, throughput and outage probability. The results obtained show that optimal solutions with QoS satisfaction are achievable for the heterogeneous CRN.

ABSTRACT. In a cognitive radio network, secondary unlicensed users are opportunistically allowed to utilize licensed spectrum band when the primary user or licensed user is not present in the cognitive radio environment. Energy is consumed by these secondary users in utilizing the licensed spectrum band before giving up access to the primary user. In this paper, we study the total amount of energy consumed by these secondary users in a cognitive radio network. The energy consumption in each state of the secondary user’s activity is analyzed and presented. The results reveal the total energy consumed at each stage of activity of the secondary users in the network and also the implication of increasing the secondary network size.

ABSTRACT. Combining spectrum sensing (SS) and primary user (PU) traffic forecasting provides a cognitive radio network (CRN) with a platform from which informed and proactive operational decisions can be made. The success of these decisions is largely dependent on prediction accuracy. Since individual SUs may suffer from SS and prediction inaccuracies due to poor channel conditions, allowing secondary users (SU) to perform these predictions in a collaborative manner allows for an improvement in the accuracy of this process. A collaborative approach to forecasting PU traffic, that combines SS and forecasting through SU cooperation, was proposed in this paper. A sub-optimal cooperative forecasting algorithm was presented to minimise cooperative prediction error. The algorithm was used to investigate the cooperative prediction performance of a group of ten SUs experiencing different channel conditions. Simulation results indicated that cooperative prediction lead to a significant improvement in prediction accuracy and illustrated how diversity, both in terms of SS accuracy and individual prediction performance, can positively impact the prediction process.

ABSTRACT. In this paper, a queue scheduling technique of evaluating the incidence of rainfall over radio links is presented. This technique proposes the synthesis of rainfall events from decomposed rain spike components generated from queueing analysis. In this study, rainfall spikes generated during rain events over Durban (29o52’S, 30o58’E) – are assumed as finite – with each overlapping one another, according to a specified queue discipline. Over this subtropical location, the M/Ek/s queue discipline is adapted such that spike arrivals are Markovian (M), and their service time, Erlangian (Ek) with multiple servers, s. From results, the rainfall statistical distributions generated from this technique are observed to compare well with the actual data in South Africa.

ABSTRACT. The deployment of small cells to address capacity and coverage problems in urban environments has resulted in the need for high capacity short range backhaul systems. In this paper, a NLOS small cell backhaul system based on diffracted signals, operating mmW technology is proposed. An analysis of diffraction loss using the Single Knife Edge diffraction model for mmW frequencies is presented. The results obtained show the viability of the use of the model for predicting diffraction loss at mmW frequencies. Furthermore, experimental results of a prototype small cell backhaul based on LoS operation are presented. Based on signals levels obtained, the results show the suitability of the system for LoS communications. However, innovative solutions are required to enable NLOS communications.

ABSTRACT. Abstract—while rainfall is a major cause of communication impairment at microwave and millimeter bands, the effect of rain attenuation in the Ethiopian microwave links, has so far remained as studied. Therefore, in this paper, the rain rate distribution at 15 minutes integration time is predicted for different percentage of time of the year based on two years of rain intensity data obtained from Ethiopian National Meteorological Agency (ENMA).The rainfall rate for one minute integration time not exceeded 0.01% of the time is calculated using ITU-R and Chebil designations, it is found that the highest rainfall rate at 0.01 % of time for any location in Ethiopia, does not exceed 64 mm/h . The ITU-R model is then applied to estimate and analyze rainfall attenuation over terrestrial radio links for ten sites in Ethiopia. The attenuation is predicted for frequencies from 1-150 GHz, with variable distances. The results show that Bahir Dar is predicted to experience largest effect of rain attenuation whereas, Dubti, has the lowest attenuation.

ABSTRACT. This research considers the application of Statistical method to automatic Machine Translation (MT) from English to Amharic. The research focuses on improving the translation quality by applying phonemic transcription on the target side, which is Amharic. Accordingly, the BLEU score results for the phoneme-based EASMT system is 37.53 a gain of 2.21 BLEU point from another baseline phrase-based EASMT with a BLEU score result of 35.32. This clearly shows that phoneme-based translation outperforms the baseline system.

ABSTRACT. This paper presents an automatic recognition system for Ethiopian license plates. The proposed system has three major components: plate detection, character segmentation and character recognition. We used Gabor filters for plate detection, connected component analysis for character segmentation, and correlation based template matching method for character recognition. In addition to the correlation value, the recognition process is supported by color analysis techniques and location information of characters. To test the performance of the system, a dataset of 350 car images of RGB color format were collected from moving cars under different angle, distance, motion and illumination conditions. Test results on the dataset showed plate detection rate of 88.9%, character segmentation accuracy of 83.9% among detected plates, and character recognition rate of 84.7% among correctly segmented characters. We have also achieved an overall plate recognition rate of 63.1%, and the whole recognition processes takes 2-5 seconds depending on whether post processing operations are needed or not.

ABSTRACT. When a game can be played at perfect-play level by an Artificial Intelligence engine, it is possible to reach the theoretical value of a position. If the opponent is a perfect player as well, there is nothing more that could be done. But, when playing against human player, we can try to achieve more than the theoretical value.What we propose here are some techniques, in particular an algorithm named WINIMAX, to increase the win probability of a perfect play engine against a human player.

ABSTRACT. Designing a vision system, which was motivated by that of the human eye, has been done since the introduction of digital computing devices. Its computational complexity hinders it from the required accuracy and flexibility achievable by these systems. A real-time Automatic Target Recognition system having the ability of detection, identification and tracking of pre-specified targets even with bad visual artifacts is developed. This work investigates an improved combination of detection and recognition algorithms to reach a better solution to the problem of target detection and recognition while aiming to fit the design of a minimal system. A speed up of 2.4 is achieved, by using a non-preeminent GPU, over brute force implementations.

ABSTRACT. The railway traffic system and its respective subsystems, especially station and timetable management are extremely large and complex systems comprising a multitude of complex structured data. The theoretical and pre-planned traffic control is calculating with idealized traffic situations, but the real life scenes often suffer from conflicts and deviations. The complex traffic conflict states include uncertain conditions and vagueness of information. Classical methods cannot work properly under these uncertain conditions. In this paper we propose a new approach to describe these complex conflict situations using fuzzy signatures for recognizing and describing traffic conflicts in a hierarchically structured manner.

ABSTRACT. In concentrated winding multi-phase permanent magnet machines, harmonic currents are intentionally injected to produce additional torque. However, harmonics result additional power losses, core temperature rise and reduction in efficiency. In this paper, analysis on the electrical power losses due to the harmonics in nine-phase permanent magnet synchronous motor is done. Analytical loss model and modified nine-phase machine, accounting for the power losses are derived. The analytical model is to generate the flux distribution in different regions of the machine and determine the losses as the result of the nonsinusoidal flux distribution. The modified model for nine-phase machine is presented to show an appropriate operating points which minimizes the total electrical loss (copper plus core loss). Comparison of power loss and efficiency in nine-phase machine with that of equivalent three phase machines is also explored.

ABSTRACT. The problem of detecting and locating shorted turns in round rotor windings has always been difficult. This is because the winding is completely embedded in the steel and the turn's insulation is not physically loaded until it is subjected to the centrifugal forces produced at speed. Thus this problem appears to be without practically known solution. However, efforts has been made by some researchers especially in the Power Industry to develop systems for detecting and analysing rotor faults which could possibly be caused by shorted turns. Repetitive Surge Oscillograph (RSO) and some other methods developed are not effective enough. A new method called Stray flux test can be used effectively to determine and locate an inter-turn short circuited fault on generator rotor. This paper describes the design and operation of a new shorted turn detector method called Stray flux test. This method provides data at operating speed to locate the coil(s) which have shorted turns and gives indications as to the number of turns shorted in the affected coils while the rotor is at speed

ABSTRACT. Geomagnetically Induced Currents (GICs) are low-frequency (typically < 100 mHz) currents that enter power transmission networks through the grounded-neutral points of the transformer windings. As early as in 1998, GICs studies within the Southern African region focused on demonstrating the existence of GIC in this low-middle latitude region. Recent literature has amply argued, proved and demonstrated the presence of GICs in the Namibian Transmission Network during a Geomagnetic Disturbance (GMD). Geomagnetic disturbances are perturbations in the magnetosphere which produce variations in the Earth’s geomagnetic field, hence leading to flow of GIC in the transmission networks. Since the GICs are DC-like currents they are also essentially referred to as “quasi-dc” currents. Consequently, conventional AC models used in power system studies cannot be used directly to analyse the GIC flow in the electric transmission system, and hence there is need for the development of a working DC model to carry out GMD impact studies on the Namibia Power Network. In this paper, a working DC model for the Namibia High Voltage (HV) Transmission network is presented. The model is tested via modelling and simulation of GIC flows in the network using the EPRI OpenDSS (Open Distribution Simulation Software). The model presented is based on the current existing network as presented to the Chamber of Mines in 2013. The partial results seem to suggest that the magnitudes of GICs in the network are increasing with the expansion and development of the network since early 2002. The effects of probable network contingencies, based on N-1 Criterion, on the flow of GICs are also presented. The simulated results are compared with the recent measurements of GICs in the network

ABSTRACT. Inductive coupling affects both aerial and underground metallic pipelines that run parallel to or in the vicinity of AC high voltage power lines, whereby voltage is induced in to the pipelines due to the time varying magnetic fields produced by the transmission line currents. We present computation of the planar magnetic field distribution on the ground below balanced AC double circuit power transmission lines of vertical and horizontal geometries for six different phase arrangements. The computations showed that the magnetic field distribution depends, among other factors, on the conductors’ phase sequence arrangement. Some arrangements show symmetrical lateral profile while some exhibit asymmetrical lateral profile. The field is also polarized with different polarization shapes at different locations which depend on the phase arrangements as well as other factors. We concluded that in placing pipelines in the vicinity of AC double circuit HVTLs, it is essential to consider the field distribution by the phase arrangement of the line. If possible, pipelines should be placed in a region of minimum field intensity within the allowable servitude of the line, in order to have minimal induction on the pipeline from the HVTL in normal operating conditions.

ABSTRACT. Alternating current (AC) induced interference corrosion has been a major threat to pipeline sharing right of ways with AC high voltage transmission lines (HVTL). Defects occur on these pipelines due to the dielectric breakdown of the pipe coating as a result of appreciable voltage being induced on the pipe from the nearby transmission lines. These defects cannot be easily assessed by inspection of the pipe’s surface as the pipeline of interest is buried below the soil surface. The value of AC current density through these coating defects has to be monitored not to exceed the threshold value that can cause corrosion damage to the pipe. In view of this, we present an assessment model given input data from field measurement, for monitoring of AC induced corrosion of an underground pipeline co-located with four single circuits HVTL of horizontal geometry. The results obtained showed that the pipeline is exposed to corrosion damage for soil with low electrical resistivity. AC current densities increase as the magnitude of the induced voltage on the pipe increases. The results also showed that the relative risk of corrosion damage also high at that area. We also provide a user friendly MATLAB-based graphical user interface to facilitate easy usage of the developed assessment model.

ABSTRACT. Modern power systems are highly interconnected and heavily loaded. The continuous load increase may drag the whole systems to a condition of voltage instability and this may result in voltage collapse. Prediction of the proximity of the system to voltage collapse is therefore an important task in power system operation and planning. Presented in this paper is the application of two techniques that can be used to determine the voltage collapse point in a power system network. At first, the conventional approach of voltage stability index, which is solely based on running repetitive load flow solutions while finding the point susceptible to system voltage collapse, is considered. The inherent structural characteristic of the power system network, which is the structural interconnections between the nodes governed by the impedances or admittances between them, is also presented. The formulation of which is done based on the basic circuit theory laws. Simulations are done using MATLAB software package. Results obtained show that voltage collapse point is easily and quickly detected and predicted with the approach of inherent structural characteristic without necessarily going through the rigor of a time consuming and repetitive load flow based voltage stability index (L-index).

ABSTRACT. This paper looks at the challenges facing Water Sanitation and Hygiene (WASH) services among the rural population and the urban poor who do not have access to piped water supply in Kenya and proposes the use of a simple technology that is supported by the application of mobile money transfer services to improve delivery of water to these segments of the population. We apply smart water metering using Advanced Metering Infrastructure (AMI) systems integrated with the popular mobile banking system in Kenya, M-Pesa to deliver efficient utilization of water resources and also employ a business model that ensure sustainability of the water supply system.

ABSTRACT. Water is a vital component in our daily lives, and a precious resource to human race. This paper describes the work carried out on the design of a water management system that can log pressure and pH readings wirelessly and determine if there is a water leak within the pipe system or if the pH levels of the water are safe enough for human consumption. A pipe monitoring system was designed, implemented and fitted with pressure sensors as well as pH sensors. The data received from the sensors was processed by a microcontroller and sent to a computer via a wireless communication module. A graphical user interface was also implemented to display the data received and to trigger an alert if any irregularities (pipe leak/burst or dangerous pH levels) are detected in real time, and leak locations can be determined.

ABSTRACT. Biodiversity monitoring is important in assessing the state of an ecosystem and determining if conservation actions are required. This is particularly important when conservation resources are scarce. However, traditional methods of biodiversity monitoring are labour intensive and cannot be applied in every ecosystem where there is need. In order to expand the application of biodiversity monitoring, there is need to automate this important task. In this work we present an application of audio diarization methods for biodiversity monitoring and show how these methods can be used to measure the abundance of indicator taxa in areas of interest. The use of audio recordings has the potential to reduce the time and effort spent in biodiversity monitoring. The experiments are performed on a freely available dataset of bird song recordings with the birds serving as indicator taxa in the ecosystem of interest. We are able to estimate the number of bird species in the recordings and this information can be used to estimate the species richness in an ecosystem.

ABSTRACT. The number of mAgriculture apps have steadily increased over the last 10 years in Kenya. The availability of mobile technology among a large portion of the population has created an opportunity to deploy mobile-based services among small holder farmers. These platforms have been deployed in different areas of the country targeting small holder farmers practicing various forms of agriculture including crop and livestock farming. These services range from agriculture advisory information, record keeping, diagnosis services, market information and veterinary services among others. These services are availed on a range of mobile devices including basic phones, feature phones, smart phones and other hand-held computing devices. Different technologies have been used to develop these innovations such as SMS, USSD, IVR, mobile web and downloadable apps. We set out to establish the uptake and use of mAgriculture among the small holder farmers through FGDs and interviews with small farm holder farmers with a diverse and representative sample. Here we report our findings. We identified the challenges and lessons learnt at different stages of their implementation. It was discovered that only a few of these implementations have been widely adopted, as various challenges have hampered their growth beyond the pilot stage. However, some of the mAgriculture innovations have been successful and offer key lessons that other stakeholders should adopt for similar attempts.

ABSTRACT. Nowadays climate change is expected to affect society in a number of ways ranging from food security to water resources. Water plays an important role in the socio-economic development of any society. Thus, this study mainly deals with assessing climate change impact on water availability of Bilate watershed, Ethiopian Rift Valley Basin. The watershed is situated in Ethiopian Rift Valley Basin and covers an area of about 3643 km2 in which Bilate river is the main river flowing throughout the year. To project the probable impact of climate change on the available water, HadCM3 coupled atmosphere-ocean GCM model was used since it is the only GCM model that has grid box containing the study area for SDSM. The output of HadCM3 coupled atmosphere-ocean GCM model for the A2a and B2a SRES emission scenarios were used to produce future scenarios of precipitation and temperature. Climate change scenarios of precipitation and temperature were developed at two upper and lower stations of the watershed for three periods namely; 2020s (2011-2040), 2050s (2041-2070) and 2080s (2070-2099 and their respective changes were determined as deltas (for temperature) and as percentages (for precipitation) from the base period values. The future climate variable such as daily precipitation, maximum and minimum temperature found as an output from the GCM model and downscaled by the SDSM model and likely change in land use were given directly as an input to the SWAT model. The SWAT simulation of future average annual flow shows a decreasing trend in 2011-2040 periods and an increasing trend in 2041-2070 periods. The average total annual flow at outlet of the watershed might decrease up to 3.7% for A2a scenario and 1.5% for B2a scenario for the 2011-2040 periods but for 2041-2070 periods it might increase up to 2.6% for A2a scenario and 3.7% for B2a scenario. The decrease in the future flow of 2011-2040 periods might be insufficient in some months to meet future demands for water of the ever increasing population with in and around the watershed. Adopting water storage options to store the excess water flowing during the rainy period is crucial.

ABSTRACT. Modern wireless communication systems such as Long-term Evolution (LTE) are expected to deliver higher data rates. However their performances are heavily impacted by the emergence of Phase noise (PN) in the transceivers. In fact PN has become an integral part of the Frequency Synthesizer design. In the process of generating carrier waves, the local oscillators of the frequency synthesizers generate an unwanted phase noise that limits the performance of the transceivers. Despite several research advances that sought to mitigate and suppress phase noise errors in radio transceivers, phase noise remains a major problem that limits the performance of wireless communication systems. Phase noise is mainly caused by the oscillators’ inability to produce a single and clean frequency without spurious tones and harmonics. In this paper a modified Fractional Phase Locked Loop Frequency Synthesizer (FPLL FS) architecture that is capable of suppressing unwanted spurs is presented. Simulation results show that the proposed model achieves significantly low phase noise level at about -145dBc/Hz with minimal spurious tones in the frequency of interest.

ABSTRACT. Channel estimation based on Pilot Symbol Assisted Modulation (PSAM) has been investigated for OFDM systems on time-varying frequency-selective channels in different literature. The varying properties of the channel degrades the performance of the OFDM system which is perceived as special case of multicarrier transmission. This paper investigates the performance of PSAM polynomial interpolation based channel estimation for OFDM systems on time-varying frequency-selective channels. The PSAM polynomial interpolation channel estimation is based on block type pilot arrangement. The SER performance of the OFDM system, and computational time complexities of the PSAM polynomial channel estimation are analysed for different fading rates, frame lengths and polynomial orders (P-orders). The SER performance is documented assuming 16-QAM as the modulation scheme in the OFDM system. From the computer simulation results, we summarize that the SER performance of polynomial based channel estimator is highly dependent on the SNR and the fading rate. Although, low P-orders exhibit good SER performance most significantly at low SNR. Also, the lower the P-orders the better SER performance and computational time complexity trade-o because high P-orders exhibit more computational delay and time complexity.

ABSTRACT. The Multi-Carrier Interleave Division Multiple Access (MC-IDMA) scheme, which is a combination of Orthogonal Frequency Division Multiplexing and IDMA, has gained wide acceptance in wireless communication. This is due to the combined individual advantages from both schemes which include: highly efficiency in data transmission, ability to combat multiple access interference and its flexibilities. A major challenge affecting wireless communication performance, is the ability to have an accurate and correct Channel State Information (CSI) required to enhance transmitted symbol recovery at the receiver. This challenge is tackled by employing channel estimation techniques which include the training based, blind and semi-blind techniques. This paper presents an efficient semi-blind estimation technique. We proposed the modified version of the conventional Minimum Mean-Square Error (MMSE) algorithm called Modified Minimum Mean-Value Estimator (MMMVE) for implementation of channel estimation for MC-IDMA systems. Complexity analysis and computer simulation results, which show the performance of the MMMVE algorithm, are presented. It can be deduced from the results that MMMVE algorithm outperform the conventional MMSE algorithm.

ABSTRACT. This paper investigates orthogonal frequency division multiplexing (OFDM) based cooperative communication system for frequency flat and selective channel conditions. Carrier frequency offset (CFO), which creates inter-carrier interference (ICI), is one known problem in OFDM-based systems. The CFO problem becomes worse when OFDM is integrated into the cooperative systems as identical OFDM signals suffer from several offsets at different uncorrelated channels. To minimize the multiple CFO effects, a two step technique is proposed in this work. Firstly, a time-domain windowing is used at the transmitter to reduce the effect of frequency offset due to Doppler shift. Secondly, Maximum Likelihood (ML)-based CFO estimation is used in each of the receiving nodes to reduces the effect of synchronization errors. The performance of the time-domain windowing and the ML-based CFO estimation is investigated, separately and jointly. Then, comparison of these systems with ICI self-cancellation and data conjugate techniques is performed. The combined technique has a signal-to-noise ratio (SNR) advantage of 7dB for a sample standard bit error rate (BER) value of 10-3.

ABSTRACT. This article presents a segmental analysis of the transmission in colour shift keying (CSK). The Euclidean theorem of distance is fractionally studied to find the boundaries of the distance between the observed and the expected points for an accurate communication. Practical segmental characterisation of the CSK receiver is presented to verify the Voronoi segmentation over the CSK channel and confirm a correlation between the red, green and blue channels based on the threshold selection.

ABSTRACT. This paper describes the use of a Nonlinear Loop Mirror to achieve wavelength and power stabilization in a three wavelength Erbium doped fiber ring laser. The laser uses three fiber Bragg grating reflectors as the oscillation wavelength selecting filters. The influence of the length of the Nonlinear Loop Mirror on the laser stability both in terms of wavelength and laser output power was investigated. The laser performance was improved by changing the length of the Loop Mirror to an optimal length and three simultaneous wavelength oscillations with acceptable power and wavelength stability were achieved.

ABSTRACT. In order to prevent various types of attacks to ICT systems, techniques of path diversity and data encryption are used. In multi-Gbps channels propagation, delay and transmission errors are always present. These introduce difficulties in protected systems operation, especially in compensation of delay variation and system synchronization. The authors have developed laboratory environment for analyzing system stability in the presence of variable propagation delay and bit error rate. Bit error occurrence is emulated by customized error generator, while delay variation is emulated by variable delay lines. This environment consists of: hardware channel simulator based on FPGA platform, PC control software, and QoS measurement instrument. Test environment emulates fiber optic or digital radio-relay channel. Experimental results of testing emulated SDH STM-16 and Ethernet 1000BaseFx transmission channels are presented.

ABSTRACT. We investigate both experimentally and by simulation the Noise Figure (NF) and pump reflection power of a Distributed Fibre Raman Amplifier (DFRA) as a function of pump power and fibre length, at different pump configurations. A DFRA of lengths 25 km and 50 km has been experimentally designed using Single Mode Reach Fibre (SMF-Reach) with an attenuation of 0.20dB/km. Two pumping techniques namely Co-pumping and counter pumping have been used. It was found that the longer the fibre length, the higher the NF irrespective of the pumping technique used. A NF of -2.2 dB and -1.9 dB was achieved experimentally for Co- and Counter pumping schemes, respectively, for 25 km fibre at 23 dB pump power. For a 50 km fibre, a NF of -1.8 dB and -0.7 dB was obtained for the two pump configurations respectively, under the same pump power. Pump reflection power varied inversely with gain and directly with fibre length. This work recommends Co-pumping technique in signal transmission due to its low NF and pump reflection power.

ABSTRACT. Content-based Image retrieval systems extract and retrieve images using their low-level features, such as color, texture, and shape. Nevertheless, these visual contents do not allow a user to formulate semantically meaningful image query. Image annotation systems are a solution to solve the inadequacy of CBIR systems and allow text based image retrieval. There have been several studies on automatic image annotation utilizing machine learning techniques and images’ representation with low level features extracted using either global or local methods. However, typically, these approaches suffer from low correlation between the globally assigned annotations and the visual features used to obtain annotations automatically. In this paper, we present an approach to enhance the effectiveness of CBIR using learning based automatic images annotation based on bag of visual word images representation that is created automatically using a set of manually annotated training images. The experimentation is performed with 4,000 annotated images for training, 1000 images for testing from ImageNet. The result has shown 77.5% of performance accuracy. The result of this work is believed to be one step towards enhancing the performance and effectiveness of existing CBIR and minimizing the semantic gap.

ABSTRACT. Electrical Impedance Tomography (EIT) is an imaging modality that discriminates among impedances of internal constituents of objects to create a map or image of such contents. EIT has been dominated by the use of Time Division Multiplexing (TDM) and Frequency Division Multiplexing (FDM) as methods of achieving orthogonal injection of excitation signals. Recently, an alternative: Code Division Multiplexing (CDM) has been proposed for the same purpose. To the authors knowledge no fully functional system or simulation platform using simultaneous injection of CDM currents has been reported. In this paper we formulate and simulate a fully functional pseudorandom code driven EIT system with 16 excitation and 16 separate measurement electrodes. In the work we verify the use of Gold codes in CDM as a multiplexing modality for EIT. The simulations are performed using the Electrical Impedance and Diffuse Optics Reconstruction Software (EIDORS). The results of this study show that: (a) Gold codes can successfully be used, in simultaneous injection as excitation current patterns in a Code Division Multiplexed EIT system. A simple image processing approach is proposed, to extract measures of reconstructed image quality. This method does not require the raw reconstruction data but uses the created images.

ABSTRACT. In this paper, fast Chan-Vese (CV) with simple shape model and connected component analysis (CCA) is proposed to segment kidneys in Magnetic Resonance Imaging (MRI). The performance of the CV model is improved by considering the sign of the level set in single layer (gray level) intensity image taking only the fitness term of the energy functional. Connected component analysis is performed on the mask obtained from final level set to earmark the possible candidate kidney. Once the candidate objects are identified, shape model is imposed to obtain the kidney. For the same size and placement of initial contour on the MRI image the fast CV has shown superior performance to that of the original CV model. In addition the paper also compares Otsu’s thresholding algorithm with fast CV method. It then established that both Otsu’s global thresholding and the proposed fast Chan-Vese model with single level set method achieve comparable performance for kidney segmentation. Further, the paper shows that Otsu’s adaptive thresholding model is superior to fast CV model in terms of speed.

ABSTRACT. Chronic Obstructive Pulmonary Disease (COPD) is characterized by progressive and partially reversible airway obstruction. Gait is often used to analyze and monitor the progression of diseases that interfere with the posture and locomotion such as COPD, however, due to dyspnea and skeletal muscle weakness presented by some patients with COPD, activities related to gait are avoided. In this case, the evaluation can be made by analyzing the movements related to other daily activities such as stand-to-sit and sit-to-stand. The aim of this study is to automatically identify a group of nine anatomical points used by a postural assessment protocol for lateral view of the stand-to-sit and sit-to-stand activity, allowing the extraction of information necessary for the protocol at any time of the execution of the activity. The proposed algorithm employs Digital Image processing techniques such as morphological filtering and image segmentation to identify anatomical points marked in the COPD patients with reflective markers. Comparing the anatomical points marked on the patient and the points detected by the proposed algorithm in 30 videos demonstrated that only two points were not identified in some frames and in 70% of the videos the algorithm detected all of the marked points. The automatization of the postural assessment protocol performed by the proposed algorithm assists the physiotherapist in obtaining a quantitative method for monitoring the evolution of the patient’s posture and enables periodic reviews to be made more quickly, accurately and throughout the physical therapy treatment.

ABSTRACT. This paper proposes a fast and robust procedure for sensing and reconstruction of sparse or compressible magnetic resonance images based on the compressive sampling theory. The algorithm starts with incoherent undersampling of the k-space data of the image using a random matrix. The undersampled data is sparsified using Haar transformation. The Haar transform coefficients of the k-space data are then reconstructed using the orthogonal matching Pursuit algorithm. The reconstructed coefficients are inverse transformed into k-space data and then into the image in spatial domain. Finally, a median filter is used to suppress the recovery noise artifacts. Experimental results show that the proposed procedure greatly reduces the image data acquisition time without significantly reducing the image quality. The results also show that the error in the reconstructed image is reduced by median filtering.

ABSTRACT. Power quality is one of the major concerns in the distribution network nowadays. Many methodologies and equipment have been designed and developed to mitigate power quality issues that are encountered in an electrical system. All devices are susceptible to failure or malfunction when they encounter a power quality problem. These devices can be a home computer, the lighting equipment of a photography studio, a generator, a transformer, or a large machine used in a production process. This paper reviews the different types of power quality issues, their causes, and lists possible solutions for the mitigation of these issues, particularly the power electronics type solutions. The DVR and the DSTATCOM are investigated in detail, simulated using MATLAB/Simulink and PSCAD, and the simulation results are compared to determine which device is best suited for mitigating power quality issues pertaining to voltage sags in a distribution network. Conclusions are drawn based on the simulation results and recommendations are made for possibility of future work.

ABSTRACT. This paper presents the study of the behaviour of the distance protection relays on the series compensated transmission line under fault conditions. The insertion of series compensation presents the distance protection elements with unorthodox line impedance and as a result the relays may operate incorrectly. Sub-harmonic oscillations and non-linearity of the line impedance are some of the challenges introduced by the insertion of series compensation on transmission lines. The study is carried out in a digsilent power factory environment and results pertaining to the response of the distance relay on a 400 kV series compensated line under fault conditions are presented.

ABSTRACT. Contactors and adjustable speed drives are power devices that are commonly used in many industrial processes. A contactor is used for switching and/or protection while an ASD is used for process control. These power devices are vulnerable to voltage dips which can disrupt the whole process, and cause huge economic losses. The economic loss of a process depends on the number of process interruptions and the cost per process interruption. This paper discusses about the performance the two devices, whose sensitivity curves are experimentally tested to different types of voltage dips, to statistical data set of voltage dips monitored in the Dutch distribution networks. In the analysis, the transfer of voltage dips from the MV to the LV network is taken into consideration. This approach can help to get insight into the economic loss of industrial plants and the setting of limits for voltage dips in the distribution network.

ABSTRACT. This paper presents an attractive control scheme that decouples cell capacitor voltage regulation of the FB-MMC from the dc link voltage regulation. This is achieved by manipulating both ac and dc components of modulation functions. The dc component of modulation function is used to regulate the cell capacitor voltages independent of the dc link voltage, while the magnitude and phase of modulation function ac component are manipulated to control active and reactive powers FB-MMC exchanges with the ac grid. The significance of this control scheme is that it permits the FB-MMC to operate with variable dc link voltage, including negative dc link voltage and zero, while full control over the power FB-MMC exchanges with ac grid is retained. Additionally, this paper provides comprehensive discussion of MMC fundamental theory, including logical and mathematical derivations of the basic relationships that governed its operation and modulation. The validity of the presented control scheme is confirmed using simulations.

ABSTRACT. The voltage link back-to-back converter (VLB2B) is widely employed for integration of wind energy resources to the microgrid. However the dc-link capacitor of the VLB2B reduces the life span of the converter and also makes the entire converter bulky. Conventional matrix converters have been suggested as possible replacement for the VLB2B converter but the lack of voltage boost ability makes them unsuitable for grid integration applications. In this paper the unidirectional step-up ultra sparse matrix converter (SUMC) is proposed as direct competitor to the widely used voltage link back-to-back converter (VLB2B) for integration of wind energy resources to the microgrid. The principles of operation of the SUMC are established. A carrier based modulation scheme is applied to control the converter. It is shown that this modulation method is capable of providing sinusoidal output currents. A cascade control scheme is employed for the control of the input currents and output voltages of the converter. The common mode voltage for the converter derived from this carrier based modulation scheme is analyzed and shown to be less than the amplitude of the phase voltage. Simulation results are used to validate the suitability of the converter for integration of wind energy resources to the microgrid.

ABSTRACT. Wind energy conversion systems (WECSs) are currently the most matured and preferred as bulk renewable energy generation systems across the world. Hence extensive research is being done on investigating the impact of wind energy integration on voltage and frequency of electricity networks for proper planning of their grid-integration such that network performance is not deteriorated, rather enhanced in terms of power quality and stability. This paper presents explores through modeling and simulation in DIgSILENT Powerfactory, how wind power plant technology, capacity and point of integration affect the voltage profile and frequency profile of utility power grid. Ensuring enhanced power quality in electricity networks is important particularly in the current wind energy deployment in South Africa.

ABSTRACT. Mobile technology has huge potential to address challenges in the areas of tacit knowledge elicitation, capturing, sharing and integration in participatory agricultural innovation where farming communities are actively involving. Past studies emphasized the behavioral and managerial perspectives of tacit knowledge and limited studies have been conducted from ICT perspective. Those ICT related studies focused more on advanced technologies including Web 2.0, groupware, audio & videoconferencing tools which are beyond the reach of the rural communities. The purpose of this study is therefore, to identify various mechanisms of tacit knowledge elicitation and sharing in participatory agricultural research process and explore how mobile based application can enhance the process. The output of the study will contribute to the development of mobile-enabled communication system that can further enhance participatory agricultural innovation through facilitating articulation, capturing, sharing and integration of tacit knowledge of multiple stakeholders including farmers.

ABSTRACT. The floral industry has increasingly become one of the most important sectors for export earnings, especially in developing countries. However, during the cultivation process there may be a number of challenges that affect it, one of which is flower disease. This paper presents an automatic identification of of flower dieases based on image processing techniques. In view of this, normal and diseased flower image are used to create a knowledge base where images are pre-processed and segmented to identify the region of interest. Texture features of images are extracted using Gabor feature extraction, from which we computed seven different measures of dispersion and central tendency with the purpose of reducing the dimensionality of features. Then, an artificial neural network is trained using seven input features extracted from individual images and eight output nodes representing eight classes of diseases considered in this work. Unkown flower images are then tested based on the training model and we achieved an average accuracy of 83.3% in identification of the flower diseases.

ABSTRACT. In this paper, two model based approaches are applied to make GMM-UBM based speaker verification task noise robust. In the first approach, speaker model adaptation is implemented based on the noise condition observed during verification. In the second approach, multi-condition training is adopted in which multiple speaker models are trained using multiple noisy speech samples. In both approaches a range of signal to noise ratios are considered. The performance of the systems in clean and environmental noise conditions is tested for both target trials and impostor trials. For test utterance corrupted by additive noise, test results show that multi-condition based noise compensation approach achieve from 1.34 to 4.8 percentage improvement for GMM-UBM.

ABSTRACT. A new method of compressive sensing reconstruction is presented. The method assumes that the signal to be estimated is both sparse and clustered. These properties are modeled as a modified Laplacian prior in a Bayesian setting, resulting in two penalizing terms in the corresponding unconstrained minimization problem. In the implementation an equivalent constrained minimization problem is solved using quadratic programming. Experiments on images with noisy observations show a significant gain when including the clustered assumption compared to the traditional LASSO approach only penalizing for sparsity. Further more, the algorithm performs better than the algorithms like CluSS-MCMC, CluSS-VB for the types of signals we have considered in this research.

ABSTRACT. Pre-qualification of wire pairs of access networks is essential, as the estimation of the performance gives the possibility to optimize the offered ensemble of services for telecommunications service providers. Our previous performance evaluation method based on fuzzy rule bases at discrete values of the insertion loss at 6 characteristic frequencies is improved by wavelet analysis. We have also proved, that the fine structure of the insertion loss spectrum does not play role in the achievable data transfer rate of the lines.

ABSTRACT. An electronically regulated, gravity-feed, intravenous infusion set in which a length of intravenous tubing is attached to a drip chamber to deliver a fluid to be administered intravenously. A contraption for monitoring drop flow in the chamber of the intravenous set incorporates a sensor housing containing a reference light source located a fixed distance from a photocell to define a fixed optical sensing gap there between, with a reference light beam normally impinging upon the photocell. The casing can be selectively fastened upon the drip chamber with the drip chamber set within the sensing gap to intercept the reference light beam. The rate of fluid flow is controlled by a non-invasive actuator attached to the intravenous tubing. A round spherical tip attached to a linear actuator stepper motor (constituents of the actuator) is adjusted with reference to a calibrated scale in accordance to feedback from the drop flow sensor. The system regulates fluid flow with an agreeable degree of accuracy (±1 count).

ABSTRACT. This paper presents a monolithic, low-noise capacitive readout interface circuit for CMOS-MEMS resonator based gravimetric chemical gas sensors. In these sensor devices, where the sense capacitances are usually very weak, the readout interface circuit plays a crucial role in determining the overall sensor performance. Noise is observed in various active and passive devices affecting circuit performances. Particularly at low frequencies, flicker noise is dominant in degrading the quality of output signals. A dual stage, open-loop continuous time voltage sensing with chopper stabilization technique is proposed in this work to cope with it and improve the total output signal SNR. The proposed circuit is designed based on MIMOS 0.35 μm AMS 3.3V CMOS technology. Cadence Spectre circuit simulator simulation results show, the proposed circuit achieves an input inferred noise of 11.6 n V/√Hz, total gain of 48.1 dB and consumes a total power of 3.385 mW. The designed circuit is able to detect minute capacitance changes as low as 0.0365 aF with total sensitivity of 67.95 μV/aF.

ABSTRACT. The early-age shrinkage of concrete was investigated using optical Fibre Bragg Grating (FBG) strain sensors. Two methods were implemented – a tube and strut mechanical transducer using a FBG sensor and a surface embedded groove FBG sensor. The two sensor methods were compared with a conventional method (Venier caliper), showing that results correlate.

ABSTRACT. A prosthetic swing-phase control mechanism simulates the action of the upper leg musculature to aid in increased gait function. More specifically, swing-phase control mechanisms limit the maximum knee flexion and allow the shank to smoothly decelerate into full knee extension without excessive impact. In this work, a hydraulic damper is designed with the objective of controlling swing-phase damping in an above-knee prosthesis. A linear spring and damper model is used to represent the dynamic properties of the damper. Based on this model, three control parameters that govern the damping force and displacement of the damper have been identified. The parameters of the damper are determined through optimization of the prosthesis knee angle with a desired knee angle trajectory obtained from experimental data in normal level walking. Experimental data of thigh and hip motions are introduced as inputs into a dynamic system to find out a set of control parameters. A computer simulation is carried out. Comparison of the desired knee angle with that of the knee angle obtained from control parameters shows the effectiveness of the present design. Moreover, conditions of knee angle and shank velocity at the end of swing phase have been checked. The results obtained show a satisfactory performance of the system.

ABSTRACT. As smartphones and tablets have become more sophisticated, they now include General Purpose Graphics Processing Units (GP GPUs) that can be used for computation beyond driving the high-resolution screens. To use them effectively, the programmer needs to have a clear sense of their microarchitecture, which in some cases is hidden by the manufacturer. In this paper we unearth key microarchitectural parameters of the Qualcomm Adreno 320 and 420 GP GPUs, present in one of the key SoCs in the industry, the Snapdragon series of chips.

ABSTRACT. This paper presents the simplified modelling of a retrofitted house-hold geyser, powered by a small Photo-Voltaic (PV) array and Wind Turbine (WT) generator, to justify the feasibility thereof. An existing 200l geyser (originally grid-connected) is modelled, and is assumed to supply one 100l bath (or two 50l showers) at a user-desired temperature of 40C, daily each evening. A recorded set of weather data (primarily solar insolation and wind speed) from the University's campus is used to approximately size the required PV area and WT diameter, to satisfy the above hot water requirements. Using the recorded weather data from 2011, a PV array area and WT swept area of 6m2 each, appears to be sufficient to provide the heating requirements throughout the year, bar one week in April. The modelling presented is somewhat simplified and lacks optimisation but intends, successfully, to demonstrate the feasibility of the application.

ABSTRACT. Utilization of renewable energy resources is currently a global concern due to several reasons that include environmental pollution caused by combustion of conventional fuels, availability of resources in the vicinity, and continuous increase in cost of conventional fuels. Solar energy is the major renewable energy resource that is a basis for the other ones. Since the annual average daily radiation for Ethiopia is 5.2kWh/m2/day and the minimum annual average radiation ranges from 4.5kWh/m2/day in July to a maximum of 5.55kWh/m2/day in February and March, parabolic trough collector solar power plants, which are mature technologies among the solar thermal ones, can be feasible to be used in the country for large and small scale applications. An interactive simulation software has been developed by the authors to support designers and researchers who are willing to simulate flow and power prediction of parabolic trough solar power plants and optimisation of their components. The design of the software began from thorough understanding of the physics governing conversion of energy from solar radiation to electricity using the plant, sizing of plant components, code writing for solving the governing equations and designing an interactive graphical user interface (GUI) for easy communication with the software. Besides giving the basic function of simulation and optimisation, the software benefits its users in being freely accessible and open source. The open sourceness of the software package also motivates others in the energy fields to develop their own packages either by modifying the code of this software or by creating new one. Furthermore, it is a unique method of mapping and identifying conducive environmental localities that are suitable for implementation of solar energy. Over and above this, this project is a unique means of sharing global responsibilities of mitigating effect of climate change. The software has been validated by simulating existing power plants for which all the data are available.

ABSTRACT. Micro businesses (MB) are small informal businesses, which have become very important for the economy and the livelihood of many people in developing countries. MBs face a number of challenges that include access to credit and funding. Micro-finance institutions (MFI) are gaining increasing foothold in developing countries as providers of credit for micro, small, and medium sized businesses. This research explores the interactive relationship between MFIs and MBs in Dar es Salaam, Tanzania. In-depth interviews were conducted with 16 informants from MFIs, and 20 informants from MBs. The data was analysed by using qualitative methods. The data analysis revealed a number of themes related to challenges in MFI-MB interactions. A number of engineering solutions to address the identified challenges are suggested.

ABSTRACT. The performance of an enterprise network is affected not only by its protocol specification, its communication channel, design capacity and architecture of the firewall but also by its implementation and traffic management. Firewall is a perimeter security solution that is useful for addressing network traffic. It introduces a single point through which all traffic passes and as a result it creates performance bottleneck on enterprise network by increasing latency, reducing bandwidth and throughput. The challenges of firewall architecture on enterprise network performance and the proposed solution to enhance it are presented in this paper. Literature review and simulated experiment are employed to study the practice in firewall configuration and management, firewall security and network performance. The stateful firewall architecture is studied and redesigned into three layers or modules: The application identification and control module, content awareness and filtering module, and enforcement module integrated with traffic optimization to accommodate the applications performance and security requirement of the enterprise network. The application identification features is tested using OPNET for efficient application identification. Its performance is then compared with a sample firewall system based on scenarios to meet the temporal quality of service requirement under DDoS envision. We have achieved firewall performance improvements of 94.4% on CPU utilization, 98.9% on throughput, and 69.20% on queue delay against the base scenario. We also achieved performance improvement of 49.23% on servers task processing and 54.35% on DB query response against the base scenario.