NRU :: Browsing by Author "Eyobu, Odongo Steven"

Browsing by Author "Eyobu, Odongo Steven"

Now showing 1 - 14 of 14

A broadcast scheme for vehicle-to-pedestrian safety message dissemination
(International Journal of Distributed Sensor Networks, 2017) Eyobu, Odongo Steven; Joo, Jhihoon; Seog Han, Dong
Ensuring cooperative awareness by periodic message beaconing in vehicular environments is necessary to address pedes- trian safety. However, high periodic basic safety message broadcasting in dense vehicular environments makes accessing the communication channel very competitive. Furthermore, high-frequency periodic broadcasting causes fast device energy dissipation which is a key issue for small computing devices used in wireless sensor and mobile communications. Therefore, in order to achieve reliable message dissemination for vehicle-to-pedestrian safety, energy loss minimization mechanisms for pedestrian mobile devices should be developed. This article proposes controlling the number of broad- casts by eliminating periodic safety message broadcasts from pedestrian nodes; these nodes only receive broadcasts from vehicles and then conditionally communicate with the vehicles when safety alerts are raised. When the pedestrian nodes do not receive messages from any vehicle for a specified period, pedestrian nodes broadcast a high-priority message advertising their position. Furthermore, for the pedestrian, adaptive message emission rates and transmission duration are proposed based on defined vehicle-to-pedestrian separation distances. This approach reduces the pedestrian device energy consumption and end-to-end delay and improves the packet delivery ratio compared to the vehicular broadcast approach for safety messages defined in the IEEE 802.11 standard.
A Comparative Review of Hand-Eye Calibration Techniques for Vision Guided Robots
(IEEE Access, 2021) Enebuse, Ikenna; Foo, Mathias; Salam Ksm Kader Ibrahim, Babul; Ahmed, Hafiz; Supmak, Fhon; Eyobu, Odongo Steven
Hand-eye calibration enables proper perception of the environment in which a vision guided robot operates. Additionally, it enables the mapping of the scene in the robots frame. Proper hand-eye calibration is crucial when sub-millimetre perceptual accuracy is needed. For example, in robot assisted surgery, a poorly calibrated robot would cause damage to surrounding vital tissues and organs, endangering the life of a patient. A lot of research has gone into ways of accurately calibrating the hand-eye system of a robot with different levels of success, challenges, resource requirements and complexities. As such, academics and industrial practitioners are faced with the challenge of choosing which algorithm meets the implementation requirements based on the identi ed constraints. This review aims to give a general overview of the strengths and weaknesses of different hand-eye calibration algorithms available to academics and industrial practitioners to make an informed design decision, as well as incite possible areas of research based on the identi ed challenges. We also discuss different calibration targets, which is an important part of the calibration process that is often overlooked in the design process.
A Multi-Model Fusion-Based Indoor Positioning System Using Smartphone Inertial Measurement Unit Sensor Data
(IEEE, 2020) Adong, Priscilla; Eyobu, Odongo Steven; Oyana, Tonny J.; Han, Dong Seog
We propose novel multi-model fusion-based step detection and step length estimation approaches that use the Kalman filter. The proposed step detection approach combines results from three conventional step detection algorithms, namely, findpeaks, localmax, and advanced zero crossing to obtain a single and more accurate step count estimate. The proposed step length estimation approach combines results from two popular step length estimation algorithms namely Weinberg’s and Kim’s methods. In our experiment, we consider five different smartphone placements, that is, when the smartphone is handheld, handheld with an arm swing, placed in the backpack, placed in a trousers’ back pocket and placed in a handbag. The system relies on inertia measurement unit sensors embedded in smartphones to generate accelerometer, gyroscope and magnetometer values from the human subject’s motion. Results from our experiments show that our proposed fusion based step detection and step length estimation approaches outperform the convectional step detection and step length estimation algorithms, respectively. Our Kalman fusion approach achieves a better step detection, step length estimation for all the five smart phone placements hence providing a better positioning accuracy. The performance of the proposed multimodel fusion-based positioning system was measured using the root mean square error (RMSE) of the displacement errors and step count errors exhibited by all the the step length and step count algorithms. The results show that the proposed Kalman fusion approach for step count estimation and step length estimation provides the least RMSE for all the smartphone placements. The proposed approach provides an average RMSE of 0.26 m in terms of the final position estimate for all the smartphone placements.
An Accurate Indoor User Position Estimator For Multiple Anchor UWB Localization
(IEEE, 2020) Poulose, Alwin; Emeršic, Žiga; Eyobu, Odongo Steven; Seog Han, Dong
UWB-based positioning systems have been proven to provide a significant high level of accuracy hence offering a huge potential for a variety of indoor applications. However, the major challenges related to UWB localization are multipath effects, excess delay, clock drift, signal interferences and system computational time to estimate the user position. To compensate for these challenges, the UWB system uses multiple anchors in the experiment area and this gives accurate position results with minimum localization errors. However, the use of multiple anchors in the UWB system means processing large amounts of data in the system controller for localization, which leads to high computational time to estimate the current user position. To reduce the complexity of the UWB systems, we propose a position estimator for multiple anchor indoor localization, which uses the extended Kalman filter (EKF). The proposed UWB-EKF estimator was mathematically analysed and the simulation results were compared with classical localization algorithms considering the mean localization errors. In the simulation, three classical localization algorithms: linearized least square estimation (LLSE), weighted centroid estimation (WCE) and maximum likelihood estimation (MLE) were used for performance comparison. Thorough extensive simulation done in this study achieves results which demonstrate the effectiveness of the proposed UWB-EKF estimator for multiple anchor UWB indoor localization.
Analysis of Machine Learning Algorithms for Prediction of Short-Term Rainfall Amounts Using Uganda’s Lake Victoria Basin Weather Dataset
(IEEE, 2024-05) Gahwera, Tumusiime Andrew; Eyobu, Odongo Steven; Isaac, Mugume
As a result of climate change, the difficulty in the prediction of short-term rainfall amounts has become a necessary area of research. The existing numerical weather prediction models have limitations in precipitation forecasting especially due to high computation requirements and are prone to errors. Precipitation amount prediction is challenging as it requires knowledge on a variety of environmental phenomena, such as temperature, humidity, wind direction, and more over a long period of time. In this study, we first of all present our Lake Victoria Basin weather dataset and then use it to conduct a rigorous analysis of machine learning algorithms to do short term rainfall prediction. The rigorous analysis includes algorithm optimizations to improve prediction performance. In particular, we intend to validate our weather dataset using various machine learning regression models which include Random Forest regression, Support vector regression, Neural Network regression, Least Absolute Shrinkage and Selection Operator regression, Gradient boosting regression, and Extreme Gradient boosting regression. The performance of the models was assessed using Mean Absolute Error (MAE), and Root Mean Square Error (RMSE) performance metrics. The findings demonstrate that, in comparison to other algorithms, Extreme Gradient Boost Regression had the lowest MAE values of 0.006, 0.018, 0.005 for Lake Victoria basin weather data in Uganda, Kenya, and Tanzania respectively.
ANN-based Stride Detection Us ing Smartphones for Pedestrian Dead Reckoning
(IEEE, 2018) Kim, Youngwoo; Eyobu, Odongo Steven; Seog Han, Dong
Position awareness is a very important issue for internet of thing (IoT) applications using smartphones. Pedestrian dead reckoning (PDR) is one of the methods used to estimate a user’s indoor position. The accuracy of a stride detection is very important to guarantee the estimation accuracy of the user location. This paper proposes an algorithm to detect the stride using acceleration spectrogram feature by utilizing the accelerometer in a smartphone. An artificial neural network (ANN) technology is applied to detect the stride. The proposed algorithm has an accuracy of 97.7% for stride detection.
Augmented CWT Features for Deep Learning-Based Indoor Localization Using WiFi RSSI Data
(Applied Sciences, 2021) Ssekidde, Paul; Eyobu, Odongo Steven; Seog Han, Dong; Oyana, Tonny J.
Localization is one of the current challenges in indoor navigation research. The conventional global positioning system (GPS) is affected by weak signal strengths due to high levels of signal interference and fading in indoor environments. Therefore, new positioning solutions tailored for indoor environments need to be developed. In this paper, we propose a deep learning approach for indoor localization. However, the performance of a deep learning system depends on the quality of the feature representation. This paper introduces two novel feature set extractions based on the continuous wavelet transforms (CWT) of the received signal strength indicators’ (RSSI) data. The two novel CWT feature sets were augmented with additive white Gaussian noise. The first feature set is CWT image-based, and the second is composed of the CWT PSD numerical data that were dimensionally equalized using principal component analysis (PCA). These proposed image and numerical data feature sets were both evaluated using CNN and ANN models with the goal of identifying the room that the human subject was in and estimating the precise location of the human subject in that particular room. Extensive experiments were conducted to generate the proposed augmented CWT feature set and numerical CWT PSD feature set using two analyzing functions, namely, Morlet and Morse. For validation purposes, the performance of the two proposed feature sets were compared with each other and other existing feature set formulations. The accuracy, precision and recall results show that the proposed feature sets performed better than the conventional feature sets used to validate the study. Similarly, the mean localization error generated by the proposed feature set predictions was less than those of the conventional feature sets used in indoor localization. More particularly, the proposed augmented CWT-image feature set outperformed the augmented CWT-PSD numerical feature set. The results also show that the Morse-based feature sets trained with CNN produced the best indoor positioning results compared to all Morlet and ANN-based feature set formulations.
CMD: A Multichannel Coordination Scheme for Emergency Message Dissemination in IEEE 1609.4
(Mobile Information Systems, 2018) Eyobu, Odongo Steven; Joo, Jhihoon; Seog Han, Dong
(e IEEE 1609.4 legacy standard for multichannel communications in vehicular ad hoc networks (VANETs), specifies that the control channel (CCH) is dedicated to broadcast safety messages, while the service channels (SCHs) are dedicated to transmit infotainment service content. However, the SCHs can be used as an alternative to transmit high priority safety messages in the event that they are invoked during the service channel interval (SCHI). (is implies that there is a need to transmit safety messages across multiple available utilized channels to ensure that all vehicles receive the safety message. Transmission across multiple SCHs using the legacy IEEE 1609.4 requires multiple channel switching and therefore introduces further end-to-end delays. Given that safety messaging is a life critical application, it is important that optimal end-to-end delay performance is derived in multichannel VANET scenarios to ensure reliable safety message dissemination. To tackle this challenge, three primary con- tributions are in this article: first, a cooperative multichannel coordinator (CMD) selection approach based on the least average separation distance (LAD) to the vehicles that expect to tune to other SCHs and operates during the control channel interval (CCHI) is proposed. Second, a model to determine the optimal time intervals in which CMD operates during the CCHI is proposed. (ird, a contention back-off mechanism for safety message transmission during the SCHI is proposed. Computer simulations and mathematical analysis show that CMD performs better than the legacy IEEE 1609.4 and a selected state-of-the-art multichannel message dissemination scheme in terms of end-to-end delay and packet reception ratio.
DNAP: Dynamic Nuchwezi Architecture Platform - A New Software Extension and Construction Technology
(IEEE, 2020) Lutalo, Joseph Willrich; Eyobu, Odongo Steven; Kanagwa, Benjamin
The need to improve or build new software systems to solve new and old business challenges is a persistent challenge in the software consumer and development industry, yet costly. To minimize these costs, the construction method should be designed with the following qualities in mind; software portability, extensibility, and simplicity. To achieve these qualities, this paper proposes the Dynamic Nuchwezi Architecture Platform (DNAP), which is a new software construction and extension technology. DNAP offers a visual programming paradigm with a capability of generating production-ready business automation software for both mobile and web. It also offers a simple mechanism for the extension of existing softwares using embeddable components. To evaluate and justify DNAP, eight Software Operating Environment (SOE) metrics have been developed and together with the SOE model, are used to contrast DNAP against four alternative software construction technologies namely; Android Platform, .NET Framework, Java SE Platform and Python. The performance evaluation results show that DNAP offers an average of 33% reduction in software construction complexity and an 11% enhancement in language efficiency when compared to alternative technologies.
Feature Representation and Data Augmentation for Human Activity Classification Based on Wearable IMU Sensor Data Using a Deep LSTM Neural Network
(Sensors, 2018) Eyobu, Odongo Steven; Seog Han, Dong
Wearable inertial measurement unit (IMU) sensors are powerful enablers for acquisition of motion data. Specifically, in human activity recognition (HAR), IMU sensor data collected from human motion are categorically combined to formulate datasets that can be used for learning human activities. However, successful learning of human activities from motion data involves the design and use of proper feature representations of IMU sensor data and suitable classifiers. Furthermore, the scarcity of labelled data is an impeding factor in the process of understanding the performance capabilities of data-driven learning models. To tackle these challenges, two primary contributions are in this article: first; by using raw IMU sensor data, a spectrogram-based feature extraction approach is proposed. Second, an ensemble of data augmentations in feature space is proposed to take care of the data scarcity problem. Performance tests were conducted on a deep long term short term memory (LSTM) neural network architecture to explore the influence of feature representations and the augmentations on activity recognition accuracy. The proposed feature extraction approach combined with the data augmentation ensemble produces state-of-the-art accuracy results in HAR. A performance evaluation of each augmentation approach is performed to show the influence on classification accuracy. Finally, in addition to using our own dataset, the proposed data augmentation technique is evaluated against the University of California, Irvine (UCI) public online HAR dataset and yields state-of-the-art accuracy results at various learning rates.
Feature Selection Based on Variance Distribution of Power Spectral Density for Driving Behavior Recognition
(IEEE, 2019) Nassuna, Hellen; Eyobu, Odongo Steven; Kim, Jae-Hoon; Lee, Dongik
Abnormal driving detection and recognition is a crucial area of research towards achieving safety in intelligent transportation systems (ITS). In this study, we propose a feature extraction approach and use the extracted features to train a deep learning model that is used for abnormal driving behavior recognition. The proposed approach derives the features based on variances calculated from each frequency bin containing the power spectrum data that is generated using the short time fourier transform. A subset of features is further selected based on variance similarity of the power spectral data. Similarity is realized by finding intersecting variance data of different variance samples obtained from defined data segments of a given driving behavior class. The driving behaviors considered are weaving, sudden braking and normal driving. Experiments were performed using an artificial neural network to test the efficiency of the proposed feature extraction approach. Results show that an accuracy of 91.0% can be achieved with accelerometer data. The accuracy is further improved to 96.1% by combining accelerometer with gyroscope data.
Feature Selection for Abnormal Driving Behavior Recognition Based on Variance Distribution of Power Spectral Density
(IEMEK Journal of Embedded Systems and Applications, 2020) Nassuna, Hellen; Kim, Jaehoon; Eyobu, Odongo Steven; Lee, Dongik
The detection and recognition of abnormal driving becomes crucial for achieving safety in Intelligent Transportation Systems (ITS). This paper presents a feature extraction method based on spectral data to train a neural network model for driving behavior recognition. The proposed method uses a two stage signal processing approach to derive time-saving and efficient feature vectors. For the first stage, the feature vector set is obtained by calculating variances from each frequency bin containing the power spectrum data. The feature set is further reduced in the second stage where an intersection method is used to select more significant features that are finally applied for training a neural network model. A stream of live signals are fed to the trained model which recognizes the abnormal driving behaviors. The driving behaviors considered in this study are weaving, sudden braking and normal driving. The effectiveness of the proposed method is demonstrated by comparing with existing methods, which are Particle Swarm Optimization (PSO) and Convolution Neural Network (CNN). The experiments show that the proposed approach achieves satisfactory results with less computational complexity.
Localization Error Analysis of Indoor Positioning System Based on UWB Measurements
(IEEE, 2019) Poulose, Alwin; Eyobu, Odongo Steven; Kim, Myeongjin; Seog Han, Dong
Ultra wide band (UWB) systems use time information instead of the popular received signal strength indication (RSSI). UWB is known for its high position accuracy in localization. RSSI-based localization is easily affected by signal attenuation and has a poor localization accuracy as compared to the time of arrival (TOA) technique. In this paper, different localization algorithms for the UWB system were analytically reviewed. The performance of the localization algorithms is discussed in terms of root mean square and cumulative distribution function of localization errors. The experiment results demonstrate the effectiveness of different localization algorithms for UWB indoor positioning. The fingerprint estimation algorithm shows better performance compared to linearized least square estimation and weighted centroid estimation algorithms. The experimental results show that the linearized least square algorithm has poor performance for UWB indoor localization.
Measurement Based V2V Path Loss Analysis in Urban NLOS Scenarios
(IEEE, 2016) Joo, Jhihoon; Eyobu, Odongo Steven; Seog Han, Dong
The importance of an accurate path loss model of vehicular environments is critical for the vehicular communication system design. However, estimating the path loss in vehicular environments is difficult due to high dynamics and low antenna heights. In this paper, we propose a line-of-sight (LOS) path loss model in vehicle-to-vehicle (V2V) scenarios and provide deep analysis of shadow fading in urban non-LOS (NLOS) scenarios by the deductive method with the proposed LOS model. The results can be utilized as reference material for further analysis of V2V path loss measurements.

Browsing by Author "Eyobu, Odongo Steven"

Results Per Page

Sort Options