Stephen Cullen Email & Phone Number

Software engineer and recent graduate of MSc in Applied Cyber Security in Technological University Dublin.Postgraduate final year project: An investigation into the robustness of the security of voice activated devices against over the air attacks.This research project examines the current state of GAN’s in relation to audio and implements three available attacks on the open source Deepspeech speech to text system and tests their robustness in an over the air setting with the aim of being used against Amazon’s Alexa and Googles voice activated systems. Some success was achieved attacking Deepspeech over the air as components of the generated adversarial examples were maintained after introducing filters, impulse responses and white noise into the generation process. While this did not translate to Amazon’s and Googles systems, it laid some groundwork to expand.Undergraduate final year project: Simple Automated Metering Infrastructure.This project involved the research, comparison and implementation of different types of interfaces that could be used to retrieve electrical values from an Electric Smart Meter (ESM) by way of Modbus TCP. Python scripts were developed to decode packets both at source with a Raspberry Pi and remotely with any python enabled client. The python scripts would use TCP to connect with the ESM, send a request and decode the response packet to retrieve the real time electrical parameter values such as Voltage, kWh and Current. The python script would then use the internet as an infrastructure to upload these values to a cloud service by way of a REST API or MQTT depending on what type of interface was being used.Electronic engineering final year project: Electrocardiograph heart rate monitor with Arduino. This was a challenging and interesting project as it combined elements of analog electronics with software design while bridging the gap between them with digital communications. The project involved the use of medical electrodes attached to the chest which would detect the voltage changes caused by the heartbeat. This voltage was fed into an instrumentation amplifier to amplify the signal and then a low pass RC filter and high pass CR filter were used to remove noise from the signal, which was fed into the Arduino's analog input. Code was then written for the Arduino to plot the waveform on a monitor and calculate the heart rate. Overall the project was a good way to bring various subjects from the course together such as electronics, software and communications and implement them in both a physical and practical sense.