• AUTONOMOUS NAVIGATION IN DYNAMIC ENVIRONMENTS

      Buxton, Jonas; Thomure, Logan; Downs, Roger; Bosanko, Garrett; Kosbar, Kurt; Missouri University of Science and Technology (International Foundation for Telemetering, 2018-11)
      Robotic systems that operate indoors are often unable to rely on GPS, and dynamic environments prove difficult to navigate for robotic systems that rely on SLAM (Simultaneous Location and Mapping). Autonomous navigation without the use of GPS or SLAM techniques require a system to rely on more fundamental hardware and software concepts. The challenge is made even greater when the system is intended to fly, interact with moving targets, and avoid moving obstacles. This is the design criteria that our autonomous multirotor is adhering to for the International Aerial Robotics Competition. This paper will describe the purpose behind each of our multirotor's sensors, such as LIDAR (Light Detection and Ranging) systems and Optical Flow sensors, that enable it to accurately interact with its environment without SLAM techniques, as well as the multirotor's onboard software that powers its autonomous capabilities.
    • COMMUNICATION SYSTEMS FOR CUBESAT MISSIONS

      Case, Anna; Kosbar, Kurt; Missouri University of Science and Technology, Department of Electrical and Computer Engineering (International Foundation for Telemetering, 2018-11)
      Several design iterations of communication systems at the Missouri S&T Satellite Research Team reveal that software defined radios (SDR) are viable for low cost, fully functional, and reliable communication systems. Recent licensing policy changes have impacted a number of CubeSat missions, prompting the necessity of bandwidth efficient communication. In searching for solutions to minimize spectral congestion, these systems need to minimize power consumption and maximize data throughput. The flexibility that SDRs provide allows for dynamic link control in orbit. Once completed, the code used to implement this system will be open-sourced for future missions use.
    • DECOUPLING HARDWARE AND SOFTWARE CONCERNS IN AIRCRAFT TELEMETRY SDR SYSTEMS

      Price, Nathan; Kosbar, Kurt; Missouri University of Science & Technology, Dept. of Electrical & Computer Engineering (International Foundation for Telemetering, 2018-11)
      Prior work has shown that software defined radio has the ability to reduce the size, weight, power and cost of telemetry and avionics. We propose a virtualized transceiver architecture that supports multiple concurrent software defined radio (SDR) applications running on shared SDR hardware. This paper applies the concept of virtual transceivers to SDR for telemetry and avionics. The proposed design allows for transceivers to be shared between different SDR applications by taking advantage of time separation and frequency adjacency. This paper addresses the system layout, hardware selection, and software organization. Improvements include a scalable and considerations for both redundancy and upgradability.
    • FAST CLASSIFICATION OF LEAF IMAGES FOR AGRICULTURAL REMOTE SENSING APPLICATIONS

      Gajjar, Viraj; Lai, Ze-Hao; Kosbar, Kurt; Missouri University of Science and Technology (International Foundation for Telemetering, 2018-11)
      This paper introduces a method of classifying leaves using machine learning. Considerable emphasis has been put on leaf classification for use in remote sensing applications such as plant phenotyping and precision agriculture. Convolutional neural networks (CNN) have been extensively used in computer vision for image classification. However, CNN can be computationally expensive. This paper describes a method that achieves a comparable accuracy, with a lower computational burden, using a support vector machine (SVM) classifier. This method uses image processing algorithms to extract features from Hough transform and Hough Lines. These features are then integrated with those extracted from binary images, and “eigenleaves” extracted from grayscale, gradient, and different color-space images of leaves as data preprocessing for classification. The classifier is implemented on two publicly available datasets: Flavia and Swedish; and is able to achieve state-of-the-art accuracies using a SVM classifier.
    • IMPACT OF PARAMETER SELECTION IN SOFT-DECISION FEEDBACK TURBO EQUALIZATION

      Nassr, Husam; Kosbar, Kurt; Missouri University of Science & Technology, Dept. of Electrical & Computer Engineering (International Foundation for Telemetering, 2018-11)
      In wireless communication systems, turbo equalization has been used to mitigate the intersymbol interference caused by dispersive channels. Despite its computational complexity, turbo equalization achieves high performance compared to systems that implement the equalization and coding processes separately. The large performance gain achieved through turbo equalization comes from exchanging soft information between the equalizer and decoder in an iterative manner. However, the computational complexity of turbo equalization can be a significant challenge for systems with limited hardware capabilities. This paper examines the performance gain versus computational complexity trade-off for a soft-decision feedback turbo equalizer (SDFTE).We show how to select parameters that achieve a desired performance specification, while minimizing implementation overhead. Sample results are presented from a simulation of a system using a Proakis channel exhibiting severe ISI using QPSK, 8PSK and 16QAM modulation schemes.
    • WIRELESS SOIL SENSOR PODS FOR LONG-TERM DATA COLLECTION

      Lipina, Jacob; Van Horn, Andrew; Schad, Judah; Kosbar, Kurt; Missouri University of Science and Technology (International Foundation for Telemetering, 2018-11)
      This paper discusses the applications of a wireless telemetry module used to collect remote sensor data used in a teleoperated electric vehicle that competed in the 2018 Mars University Rover Challenge (URC). Remote wireless soil sensor pods, 100 cc in volume, equipped with a 32-bit microcontroller and embedded IEEE 802.11 b/g/n Wi-Fi were distributed at key locations to relay soil moisture and temperature values over a local repeater to a remote base station. Combined with a low power deep sleep mode (1.84 mW), two 2500 mAh lithium-ion polymer batteries, and voltage regulation electronics, such a device could periodically relay telemetry data for many years without recharge. The small size presents the opportunity for large scale production and distribution across exoplanetary surfaces for monitoring soil characteristics over time.