Browsing by Author "Katumba, Andrew"
Now showing 1 - 20 of 35
Results Per Page
Sort Options
Item 50 Gb/s DMT and 120 Mb/s LTE signal transmission over 5 km of optical fiber using a silicon photonics transceiver(In Integrated Photonics Research, Silicon and Nanophotonics, 2018) Rahim, Abdul; Abbasi, Amin; Shahin, Mahmoud; Sequeira André, Nuno; Richter, André; Kerrebrouck, Joris Van; Van Gasse, Kasper; Katumba, Andrew; Moeneclaey, Bart; Yin, Xin; Morthier, Geert; Baets, Roel; Roelkens, GuntherNext-generation passive optical networks will require the use of low-cost, high-performance transceivers to cope with the increasing bandwidth demands for emerging applications such as fixed-mobile convergence for 5G. Silicon photonics is widely acknowledged as a technology that can provide manufacturing of low-cost photonic integrated circuits by using existing CMOS fabrication infrastructure. Intensity modulation/direct detection solutions can reach 100 Gb/s per wavelength, but require high-speed electronics and photonics, which adversely affects the cost. An alternative approach is to use advanced multi-carrier modulation schemes, such as Discrete Multi-Tone (DMT), a real-valued Orthogonal Frequency Division Multiplexing (OFDM) scheme. This technique uses Digital Signal Processing (DSP) to relax electrical and optical bandwidth requirements on the transmitter and receiver side. It promises high spectral efficiency and granularity, higher tolerance to fiber impairments and channel adaptation through flexible multi-level / multi-carrier coding [1]. DMT transmission at 100 Gb/s and even 4x100 Gb/s using modest bandwidth (~ 20 GHz) electronic and optical components has already been demonstrated [2-4]. Despite requiring computationally more expensive DSP compared to single carrier baseband schemes (e.g., OOK, PAM), DMT’s added advantage is that it allows transmission of a mobile data signal within its bandwidth using the same optical transceiver [5]. In this work we demonstrate the combined transmission of a Long Term Evolution (LTE) 4G mobile communication signal (at 3.48 GHz carrier frequency) and a 50 Gb/s DMT signal using a directly modulated InP-on-Silicon Distributed Feedback (DFB) laser. Direct modulation is poised to provide low power consumption and a reduced number of optical components in the transceiver. On the receiver side, a silicon-waveguide-coupled germanium photodiode (GeSi-PD) with a co-designed trans-impedance amplifier (TIA) is used and its performance is compared with a commercial III-V photodiode and TIA.Item 69 Gb/s DMT direct modulation of a Heterogeneously Integrated InP-on-Si DFB Laser(In Optical Fiber Communication Conference, 2017) Rahim, Abdul; Abbasi, Amin; Sequeira André, Nuno; Katumba, Andrew; Louchet, Hadrien; Van Gasse, Kasper; Baets, Roel; Morthier, Geert; Roelkens, GuntherEmerging applications such as high definition video streaming and cloud computing are the main drivers for the user-driven increase in the Internet traffic for the past few years. This has led to an increase in the processing capacity of the data centers demanding high-speed intra-datacenter communication links [1]. To address the expected growth of such short reach high speed links, the use of Wavelength Division Multiplexing (WDM) [2] and advanced modulation formats such as Quadrature Amplitude Modulation (QAM), multiband Carrierless Amplitude Phase Modulation (multi-CAP), Pulse Amplitude Modulation (PAM), and Discrete Multi-Tone (DMT) modulation [3,4] have been reported. Among these approaches, DMT has gained a lot of attention recently due to its ability to deliver 100G transmission using as low as ~20GHz optical devices [5]. Important considerations for such short reach communication links are low cost, small form factor and low power consumption. Silicon photonics is an emerging technology expected to deliver these attributes. Recently, data rates of 400 Gb/s by multiplexing 4 channels [5] and 0.88 Tb/s by multiplexing 10 channels [6] have been reported using silicon photonics. The power consumption of the optical frontend and footprint can be further reduced by implementing Directly Modulated Lasers (DMLs) on a heterogeneously integrated InP-on-Si platform [7]. Further more such lasers have been shown recently to have state-of-the-art modulation bandwidth performance [8]. In this paper we demonstrate single channel 69 Gb/s DMT modulation using a directly modulated heterogeneously integrated InP-on-Si DFB laser.Item All-optical NRZ wavelength conversion based on a single hybrid III-V/Si SOA and optical filtering(Optics Express, 2016) Wu, Yingchen; Huang, Qiangsheng; Keyvaninia, Shahram; Katumba, Andrew; Zhang, Jing; Xie, Weiqiang; Morthier, Geert; He, Jian-Jun; Roelkens, GuntherThe rapid growth of the internet and internet-related services call for large-capacity and highspeed data networking. All-optical networking allows for data reconfiguration directly in the optical layer rather than in the electronic layer. Since no optical-to-electrical-to-optical (OEO) data conversions are involved, power consumption and bandwidth of such communication networks can be substantially improved. An important function that needs to be implemented for all-optical networks is all-optical wavelength conversion (AOWC). For example, in wavelength division multiplexed (WDM) networks AOWC can remove wavelength blocking in optical cross connects [1]. AOWC based on semiconductor optical amplifiers (SOA) has a number of advantages, including high-speed operation and easy integration with other optoelectronic and passive waveguide components. Many SOA-based AOWC techniques on the InP platform have been reported, achieving data rates up to 320 Gb/s [2–6], based on cross-gain modulation (XGM), cross-phase modulation (XPM) or four wave mixing (FWM). 2R/3R regeneration and the wavelength conversion of advanced modulation format signals have also been realized using SOA-based Mach-Zehnder interferometric structures [7, 8]. Most experiments are carried out using a Return-to-Zero (RZ) data format, while the Non-Return-to-Zero (NRZ) data format is still the dominant one in commercial optical networks. AOWC for NRZ signals has been reported by means of an SOA-MZI based push-pull structure [9], bidirectional driving schemes [10] or a differentially-biased schemeItem All-Optical Reservoir Computing on a Photonic Chip Using Silicon-Based Ring Resonators(IEEE Journal of Selected Topics in Quantum Electronics, 2018) Denis-Le Coarer, Florian; Sciamanna, Marc; Katumba, Andrew; Freiberger, Matthias; Dambre, Joni; Bienstman, Peter; Rontani, DamienWe present in our work numerical results on the performance of a 4 × 4 swirl-topology photonic reservoir integrated on a silicon chip. Nonlinearmicroring resonators are used as nodes. We analyze the performance of such a reservoir on a classical nonlinear Boolean task (the delayed XOR task) for: various designs of the reservoir in terms of lengths of the waveguides between consecutive nodes, and various injection parameters (injected power and optical detuning). From this analysis, we find that this kind of reservoir can perform–for a large variety of parameters–the delayed XOR task at 20 Gb/s with bit error rates lower than 10−3 and an averaged injection power lower than 2.5 mW.Item Behavioral modeling of integrated phase-change photonic devices for neuromorphic computing applications(APL Materials, 2019) Carrillo, Santiago G.-C.; Gemo, Emanuele; Li, Xuan; Youngblood, Nathan; Katumba, Andrew; Bienstman, Peter; Pernice, Wolfram; Bhaskaran, Harish; Wright, C. DavidThe combination of phase-change materials and integrated photonics has led to the development of new forms of all-optical devices, including photonic memories, arithmetic and logic processors, and synaptic and neuronal mimics. Such devices can be readily fabricated into photonic integrated circuits, so potentially delivering large-scale all-optical arithmetic-logic units and neuromorphic processing chips. To facilitate in the design and optimization of such large-scale systems, and to aid in the understanding of device and system performance, fast yet accurate computer models are needed. Here, we describe the development of a behavioral modeling tool that meets such requirements, being capable of essentially instantaneous modeling of the write, erase, and readout performance of various integrated phase-change photonic devices, including those for synaptic and neuronal mimics.Item A behavioural model for integrated phase-change photonics devices(European Phase Change and Ovonics Symposium, 2017) Carrillo, Santiago G-C; Gemo, Emanuele; Youngblood, Nathan; Li, Xuan; Katumba, Andrew; Bienstman, Peter; Pernice, Wolfram; Bhaskaran, Harish; Wright, C. DavidThe use of phase-change materials in integrated photonics applications has enabled the development of new types of all-optical devices, including multilevel photonic memories, arithmetic and logic processors and synaptic and neuron mimics. In order to design, optimise and understand the performance of large-scale systems, fast and accurate material and device models are needed. Here we present a behavioural model for phase-change photonic devices that can simulate the write, erase and readout operations in time spans compatible with system level performance evaluation.Item A Deep Learning-based Detector for Brown Spot Disease in Passion Fruit Plant Leaves(arXiv preprint arXiv, 2020) Katumba, Andrew; Bomera, Moses; Mwikirize, Cosmas; Namulondo, Gorret; Ajeroy, Mary Gorret; Ramathaniy, Idd; Nakayima, Olivia; Nakabonge, Grace; Okello, Dorothy; Serugunda, JonathanPests and diseases pose a key challenge to passion fruit farmers across Uganda and East Africa in general. They lead to loss of investment as yields reduce and losses increases. As the majority of the farmers including passion fruit farmers, in the country are smallholder farmers from low-income households, they do not have sufficient information and means to combat these challenges. While, passion fruits have the potential to improve the well-being of these farmers given their short maturity period and high market value [1], without the required knowledge about the health of their crops, farmers can not intervene promptly to turn the situation around. For this work, we partnered with the Uganda National Crop Research Institute (NaCRRI) to develop a dataset of expertly labeled passion fruit plant leaves and fruits, both diseased and healthy. We made use of their extension service to collect images from five districts in Uganda to create the dataset. Using the dataset, we are applying state-of-the-art techniques in machine learning, specifically deep learning at scale for object detection and classification for accurate plant health status prediction. While deep learning techniques have been applied to various disease diagnosis contexts with varying degrees of success([2], [3], [4], [5], [6]), there has not been any significant effort, to the best of our knowledge, to create a dataset or apply machine learning techniques to passion fruits despite their obvious financial benefits. With this work, we hope to fill this gap by generating and making publically available an image dataset focusing on passion fruit plant diseases and pest damage and training the first generation of machine learning-based models for passion fruit plant disease identification using this dataset. The initial focus is on the locally prevalent woodiness (viral) and brown spot (fungal) diseases.Item Development of Online Laboratories for Modulation and Combinational Logic Circuit Analysis Using NI ELVIS IITM Platform(New Generations, 2010) Mwikirize, Cosmas; Tumusiime Asiimwe, Arthur; Musasizi, Lea; Namuswa, Victoria; Nakasozi, Mary Dawn; Mugga, Charles; Katumba, Andrew; Tickodri - Togboa, Sandy Stevens; Butime, Julius; Musasizi, Paul IsaacThis paper describes the work carried out by the Makerere University iLabs Project Team, hereafter referred to as iLabs@MAK. The procedures followed to develop Online Laboratories using the National Instruments Educational Laboratory Virtual Instrumentation Suite (NI ELVIS II™) platform is discussed. The laboratories were developed based on the Massachusetts Institute of Technology (MIT) iLabs Shared Architecture (ISA), a model that provides highly reliable generic services independent of the experiment domains. Modifications were made to the ELVIS version 1.0 code to introduce desired functionalities. Experiments were selected from three fundamental courses offered in the Department of Electrical Engineering, Faculty of Technology, Makerere University. Starting with the rationale for development of iLabs in specific courses, the paper presents the methods employed and the results obtained from the various experiments. Experiences and perceptions from over 300 students who performed the experiments were captured as a core aspect of the Research.Item Edge AI Face Recognition for Public Transport Fare Payment(TechRxiv, 2022) Rusoke, Blaise Marvin; Musinguzi, Denis; Miyingo, Simon Peter; Katumba, AndrewFace Recognition technology is chiefly concerned with accurately re-identifying individuals through the use of mathematical face representations. It presents a window of opportunity for the introduction of a fast, automated, seamless and easy to deploy form of biometric technology. In this research we design a fast, easy to use, and privacy oriented contactless payment system for public transportation that chiefly makes use of face recognition and internet of things technologies. We demonstrate a one-shot face recognition model and also prepare and test it for real-time inference on the edge. Our system makes use of a Siamese Model built on top of the Inception-Resnet V1 architecture with accuracy, precision and recall values of 93.81%, 90.91% and 97.35% on our validation set. The model was deployed on a Raspberry Pi 4 Model B with an Intel Neural Compute Stick 2. Inference was performed through the inference engine API of the OpenVINO toolkit on the Neural Compute Stick plugged into the Raspberry Pi. The system is composed of three other subsystems, i.e. the edge device, cloud database and user interface subsystems which work together to ensure that payment is complete in under 2 seconds.Item Face Recognition as a Method of Authentication in a Web-Based System(arXiv preprint arXiv, 2021) Mugalu, Ben Wycliff; Wamala, Rodrick Calvin; Serugunda, Jonathan; Katumba, AndrewOnline information systems currently heavily rely on the username and password traditional method for protecting information and controlling access. With the advancement in biometric technology and popularity of fields like AI and Machine Learning, biometric security is becoming increasingly popular because of the usability advantage. This paper reports how machine learning based face recognition can be integrated into a web-based system as a method of authentication to reap the benefits of improved usability. This paper includes a comparison of combinations of detection and classification algorithms with FaceNet for face recognition. The results show that a combination of MTCNN for detection, Facenet for generating embeddings, and LinearSVC for classification out performs other combinations with a 95% accuracy. The resulting classifier is integrated into the web-based system and used for authenticating users.Item Implementing Smartphone-Based Telemedicine for Cervical Cancer Screening in Uganda: Qualitative Study of Stakeholders’ Perceptions(Journal of Medical Internet Research, 2023-10) Kabukye, Johnblack K; Namugga, Jane; Mpamani, Collins Jackson; Katumba, Andrew; Nakatumba-Nabende Joyce; Nabuuma, Hanifa; Musoke, Stephen Senkomago; Nankya, Esther; Soomre, Edna; Nakisige, Carolyn; Orem, JacksonBackground In Uganda, cervical cancer (CaCx) is the commonest cancer, accounting for 35.7% of all cancer cases in women. The rates of human papillomavirus vaccination and CaCx screening remain low. Digital health tools and interventions have the potential to improve different aspects of CaCx screening and control in Uganda. Objective This study aimed to describe stakeholders’ perceptions of the telemedicine system we developed to improve CaCx screening in Uganda. Methods We developed and implemented a smartphone-based telemedicine system for capturing and sharing cervical images and other clinical data, as well as an artificial intelligence model for automatic analysis of images. We conducted focus group discussions with health workers at the screening clinics (n=27) and women undergoing screening (n=15) to explore their perceptions of the system. The focus group discussions were supplemented with field observations and an evaluation survey of the health workers on system usability and the overall project. Results In general, both patients and health workers had positive opinions about the system. Highlighted benefits included better cervical visualization, the ability to obtain a second opinion, improved communication between nurses and patients (to explain screening findings), improved clinical data management, performance monitoring and feedback, and modernization of screening service. However, there were also some negative perceptions. For example, some health workers felt the system is time-consuming, especially when it had just been introduced, while some patients were apprehensive about cervical image capture and sharing. Finally, commonplace challenges in digital health (eg, lack of interoperability and problems with sustainability) and challenges in cancer screening in general (eg, arduous referrals, inadequate monitoring and quality control) also resurfaced. Conclusions This study demonstrates the feasibility and value of digital health tools in CaCx screening in Uganda, particularly with regard to improving patient experience and the quality of screening services. It also provides examples of potential limitations that must be addressed for successful implementation. CrossRefItem Improving Time Series Recognition and Prediction with Networks and Ensembles of Passive Photonic Reservoirs(IEEE Journal of Selected Topics in Quantum Electronics, 2019) Freiberger, Matthias; Sackesyn, Stijn; Ma, Chonghuai; Katumba, Andrew; Bienstman, Peter; Dambre, JoniAs the performance increase of traditional Von- Neumann computing attenuates, new approaches to computing need to be found. A promising approach for low-power computing at high bitrates is integrated photonic reservoir computing. In the past though, the feasible reservoir size and computational power of integrated photonic reservoirs have been limited by hardware constraints. An alternative solution to building larger reservoirs is the combination of several small reservoirs to match or exceed the performance of a single bigger one. This work summarizes our efforts to increase the available computational power by combining multiple reservoirs into a single computing architecture. We investigate several possible combination techniques and evaluate their performance using the classic XOR and header recognition tasks as well as the well-known Santa Fe chaotic laser prediction task. Our findings suggest that a new paradigm of feeding a reservoir’s output into the readout structure of the next one shows consistently good results for various tasks as well as for both electrical and optical readouts and coupling schemes.Item Integrated photonic delay-lasers for reservoir computing(SPIE, 2020) Sande, Guy Van der; Harkhoe, Krishan; Katumba, Andrew; Bienstman, Peter; Verschaffelt, GuyCurrently, multiple photonic reservoir computing systems show great promise for providing a practical yet pow- erful hardware substrate for neuromorphic computing. Among those, delay-based systems o er a simple techno- logical route to implement photonic neuromorphic computation. Its operation boils down to a time-multiplexing with the delay length limiting the processing speed. As most optical setups end up to be bulky employing long ber loops or free-space optics, the processing speeds are ranging from kSa/s to tens of MSa/s. Therefore, we focus on external cavities which are far shorter than what has been realized before in such experiments. We present experimental results of reservoir computing based on a semiconductor laser, operating in a single mode regime around 1550nm, with a 10.8cm delay line. Both are integrated on an active/passive InP photonic chip built on the Jeppix platform. Using 23 virtual nodes spaced 50 ps apart in the integrated delay section, we increase the processing speed to 0.87GSa/s. The computational performance is benchmarked on a forecasting task applied to chaotic time samples. Competitive performance is observed for injection currents above thresh- old, with higher pumps having lower prediction errors. The feedback strength can be controlled by electrically pumping integrated ampli ers within the delay section. Nevertheless, we nd good performance even when these ampli ers are unpumped. To proof the relevance and necessity of the external cavity on the computational ca- pacity, we have analysed linear and nonlinear memory tasks. We also propose several post-processing methods, which increase the performance without a penalty to speed.Item Low-Loss Photonic Reservoir Computing with Multimode Photonic Integrated Circuits(Scientific reports, 2018) Katumba, Andrew; Heyvaert, Jelle; Schneider, Bendix; Uvin, Sarah; Dambre, Joni; Bienstman, PeterWe present a numerical study of a passive integrated photonics reservoir computing platform based on multimodal Y-junctions. We propose a novel design of this junction where the level of adiabaticity is carefully tailored to capture the radiation loss in higher-order modes, while at the same time providing additional mode mixing that increases the richness of the reservoir dynamics. With this design, we report an overall average combination efficiency of 61% compared to the standard 50% for the singlemode case. We demonstrate that with this design, much more power is able to reach the distant nodes of the reservoir, leading to increased scaling prospects. We use the example of a header recognition task to confirm that such a reservoir can be used for bit-level processing tasks. The design itself is CMOScompatible and can be fabricated through the known standard fabrication procedures.Item Machine Learning-Aided Optical Performance Monitoring Techniques: A Review(Frontiers in Communications and Networks, 2022) Tizikara, Dativa K.; Serugunda, Jonathan; Katumba, AndrewFuture communication systems are faced with increased demand for high capacity, dynamic bandwidth, reliability and heterogeneous traffic. To meet these requirements, networks have become more complex and thus require new design methods and monitoring techniques, as they evolve towards becoming autonomous. Machine learning has come to the forefront in recent years as a promising technology to aid in this evolution. Optical fiber communications can already provide the high capacity required for most applications, however, there is a need for increased scalability and adaptability to changing user demands and link conditions. Accurate performance monitoring is an integral part of this transformation. In this paper, we review optical performance monitoring techniques where machine learning algorithms have been applied. Moreover, since many performance monitoring approaches in the optical domain depend on knowledge of the signal type, we also review work for modulation format recognition and bitrate identification. We additionally briefly introduce a neuromorphic approach as an emerging technique that has only recently been applied to this domain.Item The Makerere Radio Speech Corpus: A Luganda Radio Corpus for Automatic Speech Recognition(arXiv, 2022) Mukiibi, Jonathan; Katumba, Andrew; Nakatumba-Nabende, Joyce; Hussein, Ali; Meyer, JoshBuilding a usable radio monitoring automatic speech recognition (ASR) system is a challenging task for under-resourced languages and yet this is paramount in societies where radio is the main medium of public communication and discussions. Initial efforts by the United Nations in Uganda have proved how understanding the perceptions of rural people who are excluded from social media is important in national planning. However, these efforts are being challenged by the absence of transcribed speech datasets. In this paper, The Makerere Artificial Intelligence research lab releases a Luganda radio speech corpus of 155 hours. To our knowledge, this is the first publicly available radio dataset in sub-Saharan Africa. The paper describes the development of the voice corpus and presents baseline Luganda ASR performance results using Coqui STT toolkit, an open source speech recognition toolkit.Item Mid-infrared Vernier racetrack resonator tunable filter implemented on a germanium on SOI waveguide platform [Invited](Optical Materials Express, 2018) Radosavljevic, Sanja; Beneitez, Nuria Teigell; Katumba, Andrew; Muneeb, Muhammad; Vanslembrouck, Michael; Kuyken, Bart; Roelkens, And GuntherCurrently, most widely tunable lasers rely on an external diffraction grating to tune the laser wavelength. In this paper we present the realization of a chip-scale Vernier tunable racetrack resonator filter on the Ge-on-SOI waveguide platform that allows for wide tuning (108 nm free spectral range) in the 5 m wavelength range without any moving parts. The fabricated racetrack resonators have a loaded Q-factor of 20000, resulting in a side-peak suppression of more than 20 dB, which is more than sufficient for wavelength selection in an external cavity laser.Item A Multiple-Input Strategy to Efficient Integrated Photonic Reservoir Computing(Cognitive Computation,, 2017) Katumba, Andrew; Freiberger, Matthias; Bienstman, Peter; Dambre, JoniPhotonic reservoir computing has evolved into a viable contender for the next generation of analog computing platforms as industry looks beyond standard transistor-based computing architectures. Integrated photonics reservoir computing, particularly on the Silicon-on-Insulator platform, presents a CMOS-compatible, wide-bandwidth, parallel platform for implementation of optical reservoirs. A number of demonstrations of the applicability of this platform for processing optical telecommunications signals have been made in the recent past. In this work, we take it a stage further by performing an architectural search for designs that yield the best performance while maintaining power efficiency.Item Neuromorphic computing based on silicon photonics(IEEE Journal of Selected Topics in Quantum Electronics, 2018) Katumba, Andrew; Freiberger, Matthias; Laporte, Floris; Lugnan, Alessio; Sackesyn, Stijn; Ma, Chonghuai; Dambre, Joni; Bienstman, PeterWe present our latest progress using new neuromorphic paradigms for optical information processing in silicon photonics. We show how passive reservoir computing chips can be used to perform a variety of tasks (bit level tasks, nonlinear dispersion compensation, ...) at high speeds and low power consumption. In addition, we present a spatial analog of reservoir computing based on pillar scatterers and a cavity, that can be used to speed up classification of biological cells.Item Neuromorphic information processing using silicon photonics(IEEE Journal of Selected Topics in Quantum Electronics, 2018) Bienstman, Peter; Dambre, Joni; Katumba, Andrew; Freiberger, Matthias; Laporte, FlorisWe present our latest results on silicon photonics neuromorphic information processing based a.o. on techniques like reservoir computing. First, we dicuss how passive reservoir computing can be used to perform non-linear signal equalisation in telecom links. Then, we introduce a training method that can deal with limited weight resolution for a hardware implementation of a photonic readout.