Evariste Twahirwa
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Evariste Twahirwa has a total of 40 co-authors. Here are the top co-authors by number of shared publications:
- Kambombo Mtonga
African Center of Excellence in Internet of Things (ACEIoT), College of Science and Technology, University of Rwanda, Kigali P.O. Box 3900, Rwanda
3 shared publications Raja Datta
Indian Institute of Technology Kharagpur
3 shared publications- Raja Datta
Department of Electronics and Electrical Communication Engineering, IIT Kharagpur, Kharagpur, India
3 shared publications Willie Kasakula
African Center of Excellence in IoT CST, University of Rwanda,Kigali,Rwanda
3 shared publications- Damien Hanyurwimfura
African Center of Excellence in Internet of Things, University of Rwanda, Kigali, Rwanda
2 shared publications - Emmanuel Ndashimye
School of ECE, Carnegie Mellon University Africa, Kigali, Rwanda
2 shared publications
Latest Publications
- Open Access
- 23 October 2026
- Published by MDPI AG in Preprints.org
TRUST-V: A Prior-Guided Spatio-Temporal Graph Learning Framework for Cross-Layer Misbehavior Detection in Vehicular Safety Communications
Vehicular safety communications rely on authenticated basic safety messages (BSMs); however, cryptographic authentication cannot prevent compromised insiders from disseminating correctly signed yet malicious information. Existing misbehavior detection systems often rely on individual-message or isolated kinematic analysis, limiting their ability to identify plausible falsifications, identity-based attacks, and coordinated multi-vehicle attacks. In this paper, we propose TRUST-V, a prior-guided spatio-temporal graph learning framework for cross-layer V2X misbehavior detection. TRUST-V first applies physical plausibility, freshness, and transmission-rate checks to remove readily identifiable violations and derive vehicle-level behavioral evidence (Stage 1). It then constructs dynamic interaction graphs in which spatial, mobility, and communication consistency priors explicitly guide graph attention while temporal dependencies are captured through recurrent modeling (Stage 2). Independent multi-label heads enable unified detection of attack classes related to false BSMs, denial-of-service, Sybil-based identity churning, coordinated multi-vehicle misbehavior, replay, identity cloning, and the overlapping behaviors of these attacks. We further develop a high-resolution Network Simulator 3 (NS-3)-generated dataset comprising 830,706 BSM records spanning benign, individual, stealthy, identity-based, compound, and coordinated attack scenarios. Evaluations demonstrate that Stage 1 removes 39.73% of the raw BSM workload while the Stage 2 graph-based analysis identifies most relational behaviors. Experimental results show an overall micro-F1 score of 93.16%, with F1-scores of 93.20%, 97.44%, 93.19%, 88.30%, 98.90%, and 96.94% for False-BSM, DoS, Sybil, Coordinated, Replay, and Identity-Clone attacks, respectively. On a Raspberry Pi 5, the complete online pipeline requires approximately 8.63 ms per 100-ms observation window under the evaluated workload, demonstrating the feasibility of combining physical knowledge with spatio-temporal relational learning for low-latency V2X misbehavior detection.
- 23 October 2026
- Published by MDPI AG in Preprints.org
Open AccessTRUST-V: A Prior-Guided Spatio-Temporal Graph Learning Framework for Cross-Layer Misbehavior Detection in Vehicular Safety Communications
Vehicular safety communications rely on authenticated basic safety messages (BSMs); however, cryptographic authentication cannot prevent compromised insiders from disseminating correctly signed yet malicious information. Existing misbehavior detection systems often rely on individual-message or isolated kinematic analysis, limiting their ability to identify plausible falsifications, identity-based attacks, and coordinated multi-vehicle attacks. In this paper, we propose TRUST-V, a prior-guided spatio-temporal graph learning framework for cross-layer V2X misbehavior detection. TRUST-V first applies physical plausibility, freshness, and transmission-rate checks to remove readily identifiable violations and derive vehicle-level behavioral evidence (Stage 1). It then constructs dynamic interaction graphs in which spatial, mobility, and communication consistency priors explicitly guide graph attention while temporal dependencies are captured through recurrent modeling (Stage 2). Independent multi-label heads enable unified detection of attack classes related to false BSMs, denial-of-service, Sybil-based identity churning, coordinated multi-vehicle misbehavior, replay, identity cloning, and the overlapping behaviors of these attacks. We further develop a high-resolution Network Simulator 3 (NS-3)-generated dataset comprising 830,706 BSM records spanning benign, individual, stealthy, identity-based, compound, and coordinated attack scenarios. Evaluations demonstrate that Stage 1 removes 39.73% of the raw BSM workload while the Stage 2 graph-based analysis identifies most relational behaviors. Experimental results show an overall micro-F1 score of 93.16%, with F1-scores of 93.20%, 97.44%, 93.19%, 88.30%, 98.90%, and 96.94% for False-BSM, DoS, Sybil, Coordinated, Replay, and Identity-Clone attacks, respectively. On a Raspberry Pi 5, the complete online pipeline requires approximately 8.63 ms per 100-ms observation window under the evaluated workload, demonstrating the feasibility of combining physical knowledge with spatio-temporal relational learning for low-latency V2X misbehavior detection.
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Communications Society
- Vol. 7, 13939-13958
BeamGuard: Risk-Aware Multimodal Beam Forecasting and Adaptive Virtual Beamwidth Control for 6G mmWave V2I Links
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Communications Society
- Vol. 7, 13939-13958
Open AccessBeamGuard: Risk-Aware Multimodal Beam Forecasting and Adaptive Virtual Beamwidth Control for 6G mmWave V2I Links
- Open Access
- 1 January 2026
- Published by Wiley in Advances in Agriculture
- Vol. 2026 (1)
Leveraging IoT‐Based Smart Irrigation Systems to Enhance Rice Productivity: A Case Study of Smallholder Farmers in the Muvumba Valley Irrigation Scheme, Eastern Rwanda
This study presents the results of a needs assessment done with small‐scale rice farmers in the Muvumba Valley irrigation scheme in Eastern Rwanda. The goal of the assessment was to learn about how farmers currently farm and what they need to help create a system that can help with rice production by telling them when to water and fertilize their crops. The research employed a mixed methodology. Data were collected from 325 of 2084 smallholder rice farmers in the Muvumba Valley. We made a structured questionnaire and used Kobo Toolbox to give it out. The survey asked farmers about their farming methods, how they monitor field conditions, and whether they would be open to using technology‐based solutions in agriculture. We used SPSS to analyze the quantitative data and create descriptive statistics. We also used content analysis to understand farmers’ experiences and feelings better. The assessment examined various aspects of rice farming, including farmers’ knowledge of soil types and rice varieties, common farming methods, access to farming supplies, watering and fertilizing practices, and how weather changes affect rice production. All participants indicated that rainfall patterns have changed in recent years, making it harder to get water and affecting how they water their plants. The findings revealed that while rice farmers in the Muvumba Valley face several production challenges, there is also a strong interest in technologies that can help improve irrigation and fertilization management, where results show that 99.40% of the farmers indicated willingness to learn and use smart farming systems.
- 1 January 2026
- Published by Wiley in Advances in Agriculture
- Vol. 2026 (1)
Open AccessLeveraging IoT‐Based Smart Irrigation Systems to Enhance Rice Productivity: A Case Study of Smallholder Farmers in the Muvumba Valley Irrigation Scheme, Eastern Rwanda
This study presents the results of a needs assessment done with small‐scale rice farmers in the Muvumba Valley irrigation scheme in Eastern Rwanda. The goal of the assessment was to learn about how farmers currently farm and what they need to help create a system that can help with rice production by telling them when to water and fertilize their crops. The research employed a mixed methodology. Data were collected from 325 of 2084 smallholder rice farmers in the Muvumba Valley. We made a structured questionnaire and used Kobo Toolbox to give it out. The survey asked farmers about their farming methods, how they monitor field conditions, and whether they would be open to using technology‐based solutions in agriculture. We used SPSS to analyze the quantitative data and create descriptive statistics. We also used content analysis to understand farmers’ experiences and feelings better. The assessment examined various aspects of rice farming, including farmers’ knowledge of soil types and rice varieties, common farming methods, access to farming supplies, watering and fertilizing practices, and how weather changes affect rice production. All participants indicated that rainfall patterns have changed in recent years, making it harder to get water and affecting how they water their plants. The findings revealed that while rice farmers in the Muvumba Valley face several production challenges, there is also a strong interest in technologies that can help improve irrigation and fertilization management, where results show that 99.40% of the farmers indicated willingness to learn and use smart farming systems.
- Open Access
IoT and Machine Learning for Smart Bird Monitoring and Repellence: Techniques, Challenges, and Opportunities
The activities of birds present increasing challenges in agriculture, aviation, and environmental conservation. This has led to economic losses, safety risks, and ecological imbalances. Attempts have been made to address the problem, with traditional deterrent methods proving to be labour-intensive, environmentally unfriendly, and ineffective over time. Advances in artificial intelligence (AI) and the Internet of Things (IoT) present opportunities for enabling automated real-time bird detection and repellence. This study reviews recent developments (2020–2025) in AI-driven bird detection and repellence systems, emphasising the integration of image, audio, and multi-sensor data in IoT and edge-based environments. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses framework was used, with 267 studies initially identified and screened from key scientific databases. A total of 154 studies met the inclusion criteria and were analysed. The findings show the increasing use of convolutional neural networks (CNNs), YOLO variants, and MobileNet in visual detection, and the growing use of lightweight audio-based models such as BirdNET, MFCC-based CNNs, and TinyML frameworks for microcontroller deployment. Multi-sensor fusion is proposed to improve detection accuracy in diverse environments. Repellence strategies include sound-based deterrents, visual deterrents, predator-mimicking visuals, and adaptive AI-integrated systems. Deployment success depends on edge compatibility, power efficiency, and dataset quality. The limitations of current studies include species-specific detection challenges, data scarcity, environmental changes, and energy constraints. Future research should focus on tiny and lightweight AI models, standardised multi-modal datasets, and intelligent, behaviour-aware deterrence mechanisms suitable for precision agriculture and ecological monitoring.
Open AccessIoT and Machine Learning for Smart Bird Monitoring and Repellence: Techniques, Challenges, and Opportunities
The activities of birds present increasing challenges in agriculture, aviation, and environmental conservation. This has led to economic losses, safety risks, and ecological imbalances. Attempts have been made to address the problem, with traditional deterrent methods proving to be labour-intensive, environmentally unfriendly, and ineffective over time. Advances in artificial intelligence (AI) and the Internet of Things (IoT) present opportunities for enabling automated real-time bird detection and repellence. This study reviews recent developments (2020–2025) in AI-driven bird detection and repellence systems, emphasising the integration of image, audio, and multi-sensor data in IoT and edge-based environments. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses framework was used, with 267 studies initially identified and screened from key scientific databases. A total of 154 studies met the inclusion criteria and were analysed. The findings show the increasing use of convolutional neural networks (CNNs), YOLO variants, and MobileNet in visual detection, and the growing use of lightweight audio-based models such as BirdNET, MFCC-based CNNs, and TinyML frameworks for microcontroller deployment. Multi-sensor fusion is proposed to improve detection accuracy in diverse environments. Repellence strategies include sound-based deterrents, visual deterrents, predator-mimicking visuals, and adaptive AI-integrated systems. Deployment success depends on edge compatibility, power efficiency, and dataset quality. The limitations of current studies include species-specific detection challenges, data scarcity, environmental changes, and energy constraints. Future research should focus on tiny and lightweight AI models, standardised multi-modal datasets, and intelligent, behaviour-aware deterrence mechanisms suitable for precision agriculture and ecological monitoring.
- Open Access
- 16 July 2025
- Published by MDPI AG in Preprints.org
IoT and Machine Learning for Smart Bird Monitoring and Repellence: Techniques, Challenges, and Opportunities
The activities of birds present increasing challenges in agriculture, aviation, and environmental conservation. This has led to economic losses, safety risks, and ecological imbalances. Attempts have been made to address the problem, with traditional deterrent methods proving to be labour-intensive, environmentally unfriendly, and ineffective over time. Advances in Artificial Intelligence (AI) and the Internet of Things (IoT) present opportunities for enabling automated real-time bird detection and repellence. This study reviews recent developments (2020–2025) in AI-driven bird detection and repellence systems, emphasising the integration of image, audio, and multi-sensor data in IoT and edge-based environments. The Preferred Reporting Items for Systematic reviews and Meta-Analyses framework was used, with 267 studies initially identified and screened from key scientific databases. A total of 154 studies met the inclusion criteria and were analysed. The findings show the increasing use of convolutional neural networks (CNNs), YOLO variants, and MobileNet in visual detection, and the growing use of lightweight audio-based models such as BirdNET, MFCC-based CNNs, and TinyML frameworks for microcontroller deployment. Multi-sensor fusion is proposed to improve detection accuracy in diverse environments. Repellence strategies include sound-based deterrents, visual deterrents, predator-mimicking visuals, and adaptive AI-integrated systems. Deployment success depends on edge compatibility, power efficiency, and dataset quality. The limitations of current studies, include species-specific detection challenges, data scarcity, environmental changes, and energy constraints. Future research should focus on tiny and lightweight AI models, standardised multi-modal datasets, and intelligent, behaviour-aware deterrence mechanisms suitable for precision agriculture and ecological monitoring.
- 16 July 2025
- Published by MDPI AG in Preprints.org
Open AccessIoT and Machine Learning for Smart Bird Monitoring and Repellence: Techniques, Challenges, and Opportunities
The activities of birds present increasing challenges in agriculture, aviation, and environmental conservation. This has led to economic losses, safety risks, and ecological imbalances. Attempts have been made to address the problem, with traditional deterrent methods proving to be labour-intensive, environmentally unfriendly, and ineffective over time. Advances in Artificial Intelligence (AI) and the Internet of Things (IoT) present opportunities for enabling automated real-time bird detection and repellence. This study reviews recent developments (2020–2025) in AI-driven bird detection and repellence systems, emphasising the integration of image, audio, and multi-sensor data in IoT and edge-based environments. The Preferred Reporting Items for Systematic reviews and Meta-Analyses framework was used, with 267 studies initially identified and screened from key scientific databases. A total of 154 studies met the inclusion criteria and were analysed. The findings show the increasing use of convolutional neural networks (CNNs), YOLO variants, and MobileNet in visual detection, and the growing use of lightweight audio-based models such as BirdNET, MFCC-based CNNs, and TinyML frameworks for microcontroller deployment. Multi-sensor fusion is proposed to improve detection accuracy in diverse environments. Repellence strategies include sound-based deterrents, visual deterrents, predator-mimicking visuals, and adaptive AI-integrated systems. Deployment success depends on edge compatibility, power efficiency, and dataset quality. The limitations of current studies, include species-specific detection challenges, data scarcity, environmental changes, and energy constraints. Future research should focus on tiny and lightweight AI models, standardised multi-modal datasets, and intelligent, behaviour-aware deterrence mechanisms suitable for precision agriculture and ecological monitoring.
- 27 October 2023
- Published by Springer Nature
- p. 272-285
An IoT Based Embedded Gateway for Smart Health-Care (IoTEGH) Management
- 27 October 2023
- Published by Springer Nature
- p. 272-285
An IoT Based Embedded Gateway for Smart Health-Care (IoTEGH) Management
- 28 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 716-721
IoT and AI Technological Empowered Covid-19 Control Approaches
The Covid-19 virus has continued to mutate with some variants proving more aggressive and more transmissible. Clearly, this calls for continued efforts to strengthen public health systems preparedness and readiness to respond to both current and future pandemics. Technological innovations have played pivotal roles in the fight against the Covid-19 pandemic. Unfortunately, in Sub-saharan Africa there has been lack of rapid adoption and deployment of indigenous technological innovations. This work presents three internet of things and artificial intelligence powered technological innovations that can be deployed to strengthen the public health system readiness to fight the Covid-19 pandemic and other similar pandemics. Specifically, these include, (i) a social distance violation checker, which can be deployed to protect front desk employees; (ii) a portable vital signs screening and monitoring kit, suitable for use in both crowded environments and in private homes; and (iii) an Artificial Intelligence integrated smart camera system for checking unfitted mask wearing. These innovations can help to enforce public adherence to social distance guides, mask wearing and also timely detection and isolation of potential cases, all of which are important strategies in the fight against the Covid-19 pandemic. These innovations have been tested and their performance is comparable to existing commercial solutions.
- 28 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 716-721
IoT and AI Technological Empowered Covid-19 Control Approaches
The Covid-19 virus has continued to mutate with some variants proving more aggressive and more transmissible. Clearly, this calls for continued efforts to strengthen public health systems preparedness and readiness to respond to both current and future pandemics. Technological innovations have played pivotal roles in the fight against the Covid-19 pandemic. Unfortunately, in Sub-saharan Africa there has been lack of rapid adoption and deployment of indigenous technological innovations. This work presents three internet of things and artificial intelligence powered technological innovations that can be deployed to strengthen the public health system readiness to fight the Covid-19 pandemic and other similar pandemics. Specifically, these include, (i) a social distance violation checker, which can be deployed to protect front desk employees; (ii) a portable vital signs screening and monitoring kit, suitable for use in both crowded environments and in private homes; and (iii) an Artificial Intelligence integrated smart camera system for checking unfitted mask wearing. These innovations can help to enforce public adherence to social distance guides, mask wearing and also timely detection and isolation of potential cases, all of which are important strategies in the fight against the Covid-19 pandemic. These innovations have been tested and their performance is comparable to existing commercial solutions.
- 16 July 2022
- Published by University of Dar es Salaam - DIGITAL COMMONS JOURNALS in Tanzania Journal of Engineering and Technology
- Vol. 41 (1) , 28-37
Real-Time Data Analytics for Monitoring Electricity Consumption Using IoT Technology
Rising electricity bills as a result of climate variability and new home electrical and electronic appliances are becoming a major source of concern for most end users. Consumers are typically unaware of their household electricity consumption patterns and the costs associated with them, making proper planning and budgeting difficult. Monitoring and controlling energy consumption on appliances can reduce energy costs for end-users. The Internet of Things (IoT) has the potential to provide remote monitoring and control of devices via automated monitoring and control. In this study, we propose an IoT-enabled system for monitoring and controlling energy consumption in homes to save money and make electricity more affordable to low-income people. Autonomous sensor nodes attached to power outlets collect and route power consumption data to the GSM-enabled gateway. Data aggregation is performed by the gateway for data received from all end-nodes within its coverage area. The gateway sends the data to the cloud, where real-time data analytics is performed. We also developed an algorithm that enables users to not only understand their usage patterns but also remotely control their appliances at any time. Mobile phones are used to remotely turn on and off household appliances via a machine-to-machine interface, as well as to provide real-time visualization of household energy consumption patterns.
- 16 July 2022
- Published by University of Dar es Salaam - DIGITAL COMMONS JOURNALS in Tanzania Journal of Engineering and Technology
- Vol. 41 (1) , 28-37
Real-Time Data Analytics for Monitoring Electricity Consumption Using IoT Technology
Rising electricity bills as a result of climate variability and new home electrical and electronic appliances are becoming a major source of concern for most end users. Consumers are typically unaware of their household electricity consumption patterns and the costs associated with them, making proper planning and budgeting difficult. Monitoring and controlling energy consumption on appliances can reduce energy costs for end-users. The Internet of Things (IoT) has the potential to provide remote monitoring and control of devices via automated monitoring and control. In this study, we propose an IoT-enabled system for monitoring and controlling energy consumption in homes to save money and make electricity more affordable to low-income people. Autonomous sensor nodes attached to power outlets collect and route power consumption data to the GSM-enabled gateway. Data aggregation is performed by the gateway for data received from all end-nodes within its coverage area. The gateway sends the data to the cloud, where real-time data analytics is performed. We also developed an algorithm that enables users to not only understand their usage patterns but also remotely control their appliances at any time. Mobile phones are used to remotely turn on and off household appliances via a machine-to-machine interface, as well as to provide real-time visualization of household energy consumption patterns.
- 29 May 2022
- Published by Springer Nature
- p. 15-25
Context Sensing from Deployed IoT Sensors for Improved Public Transport Systems
- 29 May 2022
- Published by Springer Nature
- p. 15-25
Context Sensing from Deployed IoT Sensors for Improved Public Transport Systems
- Open Access
- 12 March 2022
- Published by Wiley in Journal of Advanced Transportation
- Vol. 2022, 1-12
Travel Matrix Enabled Delta-Based Roadside Units Deployment for Vehicular Ad Hoc Networks: A Case of Kigali City
Placements of Road Side Units (RSUs) are an important issue of vehicular networks in urban areas. The merged cost of procurement, installation, and maintenance of intelligent RSUs is high, and therefore, cost-effective deployment strategies are necessary. In this article, we propose a scheme that optimally deploys intelligent roadside units using a travel matrix scheme based on the classical delta strategy where urban vehicles are involved in RSUs communication in the course of their travel times. Four (4) vehicular communication modes are studied, namely, (a) travel matrix based on delta RSUs deployment communication, (b) road intersection-based RSUs deployment communication, (c) road segmentation, and (d) free vehicle-to-vehicle communication. A baseline algorithm is suggested to determine the optimal locations of RSUs in terms of their geographical positions. A travel matrix technique is proposed to trace vehicles’ routeways and travel times in some points of interest (POI). Our intention is to seek an approach that reduces the required number of RSUs and ensures greater network performance effectiveness in terms of packets delivery ratio, throughput, message delay, and jitter; from our study, travel matrix delta-based placement of RSUs becomes the best in our case study scenario. The simulation results indicate that the travel matrix deployment is a suitable deployment scheme in the case study area since it can reduce the number of RSUs while enhancing the vehicular communication abilities under different vehicle density scenarios.
- 12 March 2022
- Published by Wiley in Journal of Advanced Transportation
- Vol. 2022, 1-12
Open AccessTravel Matrix Enabled Delta-Based Roadside Units Deployment for Vehicular Ad Hoc Networks: A Case of Kigali City
Placements of Road Side Units (RSUs) are an important issue of vehicular networks in urban areas. The merged cost of procurement, installation, and maintenance of intelligent RSUs is high, and therefore, cost-effective deployment strategies are necessary. In this article, we propose a scheme that optimally deploys intelligent roadside units using a travel matrix scheme based on the classical delta strategy where urban vehicles are involved in RSUs communication in the course of their travel times. Four (4) vehicular communication modes are studied, namely, (a) travel matrix based on delta RSUs deployment communication, (b) road intersection-based RSUs deployment communication, (c) road segmentation, and (d) free vehicle-to-vehicle communication. A baseline algorithm is suggested to determine the optimal locations of RSUs in terms of their geographical positions. A travel matrix technique is proposed to trace vehicles’ routeways and travel times in some points of interest (POI). Our intention is to seek an approach that reduces the required number of RSUs and ensures greater network performance effectiveness in terms of packets delivery ratio, throughput, message delay, and jitter; from our study, travel matrix delta-based placement of RSUs becomes the best in our case study scenario. The simulation results indicate that the travel matrix deployment is a suitable deployment scheme in the case study area since it can reduce the number of RSUs while enhancing the vehicular communication abilities under different vehicle density scenarios.