arnold jojo kattampally

Radar- und Ethernet-Expertise: Entwicklung robuster Funktionen für die Erfassung von Radar-Paketen via UDP über Ethernet, Sicherstellung der Integrität der Metadaten und Vollständigkeit des Zyklus. Signalverarbeitung und eingebettete Systeme: Aktualisierung eines Radar-Wiedergabewerkzeugs für RFFT-Pakete und OpenCL-basierte Signalverarbeitung. Einführung eines CFAR-Schwellenwerts für präzise Signalanalyse.

- Advanced the Technology Readiness Level (TRL) of Radar-based Human Recognition using Neural Networks from TRL 2 (idea) to TRL 4. - Redesigned the Kalman tracking mechanism for enhanced accuracy and robustness in multiple object tracking. - Implemented voxelization as a preprocessing technique to facilitate feature extraction for Convolutional Neural Networks - Improved recognition accuracy by training a Conv3D + LSTM model on voxelized data, enabling effective human recognition.

- Implemented radar data processing pipeline for radar simulation, resulting in a remarkable reduction in execution time from 1 minute to under 10 seconds. - Implemented signal processing algorithms, including CFAR detector, direction of arrival estimation, Fast Fourier Transform (FFT), and demodulation, in C++ for precise and reliable detection outcomes.

- Independently developed a fully functional web page utilizing Bootstrap, HTML, CSS, and JavaScript, showcasing proficiency in front-end development. (https://heatbeat.de/) - Enhanced and expanded APIs utilized as crucial data points in the heatbeat simulation by leveraging Django framework. - Achieved significant improvements in page speed performance by optimizing mobile version to 90% and desktop version to 100%

Developed Python automation scripts that efficiently monitor server status using Zabbix, automatically detecting and provisioning new servers in the event of downtime, resulting in a significant reduction of downtime to less than 5 seconds.

Successfully scaled the foreign exchange app, designed for facilitating monetary transactions between UAE and India, from supporting 1000 users to effortlessly managing a user base exceeding 10,000 users, guaranteeing performance, scalability, and uninterrupted service