< Back to previous page

Project

Distributed Spatial Filtering in Wireless Sensor Networks

Wireless sensor networks (WSNs) paved the way for many new technological and scientific advances by accessing information previously unavailable. For example, continuously being able to monitor the brain is crucial for many treatments in the biomedical and healthcare fields, especially for patients suffering from neurological disorders. Using sensor networks, electrical activity of the brain can be measured by sensors placed on the scalp of the patients and processed to extract valuable information, for example, to detect epileptic seizures, or control limbs. Alternatively, consider microphone networks that continuously monitor sounds in their environment, which can be used to distinguish between speakers in hearing aids, or to enhance speech in telephony. Traditional techniques that can be applied to achieve these goals include what we refer to as spatial filtering, or in other words, combining the data collected from all sensors, e.g., resulting from the electrical activity of the brain. Chronic monitoring is however only possible if the wearable devices are small enough, either to be comfortably worn in daily life activities or to be deployed in various locations in space, which means that they cannot have unlimited energy resources due to their small battery size. In its current state, this procedure requires the devices to continuously transmit the data these sensors measure to a central device, such as a smartphone or computer, which is unfortunately not sustainable for a long period of time for the sensors, as it requires large amounts of energy resources. A promising solution is to completely remove this central machine from consideration, such that the sensors achieve their goals by only collaborating with each other. In this project, we will design adaptive multi-channel spatial filtering algorithms for various applications, amenable to low-power distributed architectures with constrained energy resources as envisaged in the WSN concept.

Date:2 Sep 2019 →  15 Nov 2023
Keywords:Signal Processing, Biomedical Signal Processing, Distributed Signal Processing
Disciplines:Signal processing, Ubiquitous computing, Biomedical signal processing, Analogue and digital signal processing
Project type:PhD project