In this study multiple design approaches have been tried to accurately determine the respiratory rate every 30 seconds of a patient in a hospital bed using six piezoelectric pressure sensors located sand­wiched between the mattress and the bed. After four design iterations using a variety of different methods, a final design using a Least Mean Squares (LMS) filter and peak detection in the frequency domain was implemented. According to the test results, this method determines an accurate respira­tory rate 85% of the time in a controlled, experimental environment with the subject lying completely still. Unfortunately this accuracy is below the set requirements. However, it is believed that multiple improvements can still be made on the algorithm to increase the accuracy.

Pressure ulcers are wounds that arise when a person sits or lays in the same posture for an extended period of time. Starting from the skin and possibly continuing to the bone, tissue slowly decays when the pressure keeps being applied to the same spot. Prevention is the best method of treating the ulcers and it is done by regularly changing posture. Normally a person does this automatically but for a portion of hospital patients this option is not there. The goal of this project was to design and create a prototype which can track the body position of a person on a bed. This way an alarm can be triggered when a person stays in the same body position for too long. To complete this goal a list of requirements was created to guide the design process. The most import design requirement was the ability to track the body position non-intrusively, which means a person should not perceive the tracker in any way. Furthermore, minimizing cost and interference were also important design considerations. Using piezoelectric and piezoresistive sensors beneath a mattress, the pressure a person applies to it can be measured and used to determine the body position. The sensor encasing is thin enough to go unnoticed through the mattress when lied upon. Therefore, the non-intrusive tracking is successfully implemented. By keeping the circuitry simple and looking for the most cost-effective components the overall costs for the entire electrical system are kept to a reasonable minimum. For the piezoelectric circuit a parallel zener diode was all that was necessary and the piezoresistive circuit consists of an inverting amplifier and low-pass filter. The cost per system comes down to $€146.39$, which can still decrease when a more tailor-made microcontroller is used for reading out the analog-to-digital converters and when costs decrease due to buying in bulk. Minimizing the interference for the system is implemented to some degree for the piezoresistive by adding a low-pass filter before the ADC input. The lack of physical shielding results in a poor interference reduction. This will have to be addressed in the future. The required cooperation within the project resulted in many prototypes being developed to ensure the algorithm could use the most recent version of the system at any time. The collaboration with a mechanical engineering student resulted in an integrated prototype of both mechanical and electrical design. The additional targets set for the project consisted of detecting respiration rate and heart rate, creating a PCB and creating a pilot worthy prototype. Of these targets the PCB creation is the only target which is nearly completed. Soldering the PCB can however not be done before completing this report due to delivery times.

In this paper we will research the possibility, reliability and accuracy of calculating the heart rate of a patient at rest using Ballistocardiography in clinical settings. The vital signs of the patient will be extracted by using piezoelectric sensor which is embedded in a Bedsense device developed by Momomedical. The device developed by Momomedical is placed under the mattress and has no direct contact with the patient. The piezoelectric sensor generates a signal that can be separated into motion, respiration and ballistocardiogram (BCG) waveforms, from which movement, respiration rate (RR) and heart rate (HR) can be obtained respectively. To improve the accuracy of the measurement several signal processing methods will be applied to reduce noise and outliers or sharpen the waveform before calculating the heart rate using peak detection. The reliability is determined by comparing the percentage difference with a reference heart rate. The accuracy was found to be 55.74% of the time within 5% and 91.8% within 10% when lying on your back. This was found to be 84.43% of the time within 5% when lying on your right side and for the left side this was 80.33% of the time within 5%.

In this thesis five different Direction Of Arrival algorithms will be developed for use with Microflown's Acoustic Vector Sensors, which will determine the direction an acoustic signal originates from. These algorithms will run on a main-station that will remotely receive data from an array of AVSs. The performance of these algorithms is examined using simulations and measurements obtained by a test setup. Finally, it is discussed under what circumstances which algorithm delivers the best performance.

The founders of Momo Medical envisioned a health care product that would help nurses worldwide with pressure ulcer prevention. Pressure ulcers are a chronic wound that affects the skin of patients who do not regularly change bed posture. As it currently stands, nurses lack the manpower and time to change the postureof patients in a timely manor. Thus, a goal was set to create a product that would reduce this strain on nurses.The goal was to design a non-intrusive product that keeps track of the posture of a patient on a hospital bed, and send an alarm if this has not changed in a set amount of time. A constraint with this was that the product has to lay directly underneath the mattress of the patient. Furthermore, the product has to make use of as few sensors as possible. With these constrains, the project group was divided in three subgroups: the sensor, the algorithm, and the testing subgroup.This report will focus on the design, implementation, and evaluation of the algorithm. The algorithm has to be designed to process the sensor data and return, with 90% accuracy, the posture of the patient. It also has to be able to determine with 99% accuracy whether or not there is a patient on the bed. An optional objective was to design an algorithm to determine the heart and respiration rate of the patient, with 80% accuracy, but this was of lower priority. It was soon decided to give priority to posture detection and not further investigate detection of the heart and respiration rate.In the final product, the information about the patient’s posture will be displayed in a graphical user interface (GUI). This GUI has to be user-friendly and intuitive to use by the nurses. During the prototyping phase, however, a different GUI was used, one that displayed various debugging information. This prototype GUI was necessary to test the various algorithms.Several different mathematical models have been investigated, implemented, and tested. These models range from taking the variance of the sensor data, to deploying a Fourier transform on a polynomial curve based on the sensor data. With preliminary test data, an accuracy of 93% was achieved using a neural network. With the final testing data however, only the neural network reached the required accuracy of 90%, though only just, at 90.8%. With future iterations of the algorithm, more research and data collection is advised, as neural networks work better with larger data sets.

In the current Big Data era, large amounts of data are collected from complex systems, such as sensor networks and social networks. The emerging field of graph signal processing (GSP) leverages a network structure (graph) to process signals on an irregular domain. This thesis studies the forecasting of multi-dimensional graph processes, i.e., where each entity in the network carries a multivariate time series. Recent research has proposed to use product graphs to model the dependencies between different variables in multi-dimensional graph processes and employ them in graph-based vector autoregressive models to predict future values. A problem with these product graph-based models is that they can be too restrictive. In this work, it is proposed to combine product graph-based models with multiple one-dimensional models to implement more estimation flexibility. To further increase the degrees of freedom, the use of multiple-input-multiple-output graph filters is also proposed. The proposed models are implemented and tested on synthetic and real-world data sets, which shows an improved forecasting performance compared to state-of-the-art alternatives.

In recent years, the large increase in connected devices and the data that is collected by these devices has caused a heightened interest in distributed processing. Many practical distributed networks are of heterogeneous nature. Because of this, algorithms operating within these networks need to be simple, robust against network dynamics and resource efficient. Additionally, if privacy preservation methods are properly implemented, they add to the power of distributed processing by making it possible to leverage the data of many different users, without infringing the privacy of the individuals involved. In this study we focus on the primal-dual method of multipliers (PDMM), which is a promising dis- tributed optimisation algorithm that seems to be suitable for distributed optimisation in heterogeneous networks. Most theoretical work that can be found in existing literature focuses on synchronous ver- sions of PDMM. However, in heterogeneous networks, asynchronous algorithms are favourable over synchronous algorithms. So far, simulation results have indicated that asynchronous PDMM converges and can even converge in the presence of transmission losses. In this work we analyse the properties of stochastic PDMM, which is a general framework that can model variations of PDMM such as asynchronous PDMM and PDMM with transmission losses. We build upon previous empirical results of PDMM and formulate theoretical proofs to substantiate these results. After defining stochastic PDMM and proving its convergence, we compare a number of PDMM variations that have been mentioned throughout the literature. Lastly, we derive a lower bound for the variance of the auxiliary variable in the context of stochastic PDMM, assuming uniform updating probabilities. This lower bound indicates that subspace based privacy preservation is applicable to certain instances of stochastic PDMM, like asynchronous PDMM. The main result of this work is a theoretical proof that shows that stochastic PDMM converges almost surely if the updating probabilities of each auxiliary variable are nonzero. Two important conclusions that follow from this proof are the almost sure convergence of asynchronous PDMM and unicast PDMM with transmission losses. Another useful result is the fact that subspace based privacy preservation is effective when using asynchronous PDMM.

In this work, we deal with the problem of reconstructing a complete bandlimited graph signal from partially sampled noisy measurements. For a known graph structure, some efficient centralized algorithms are proposed to partition the nodes of the graph into disjoint subsets such that sampling the graph signal from any of these subsets leads to a sufficiently accurate reconstruction. Furthermore, we consider the situation when the graph has a massive size, where processing the data centrally is impractical anymore. To overcome this issue, a distributed framework is proposed that allows us to implement the centralized algorithms in a parallelized fashion. Finally, we provide numerical simulation results on synthetic and real-world data to show that our proposals outperform state-of-the-art node partitioning techniques.

This thesis consists of two parts in both data science and signal processing over graphs. In the first part of this thesis, we aim to solve the problem of graph construction in big data scenario, which is critical for practical tasks, like collaborative filtering in recommender systems, spectral embedding or clustering in learning algorithms. We achieve to accelerate the data-driven graph construction algorithms by relying on an approximation technique for large matrix multiplication, diamond sampling. We show its potential in real problems by extensive experiments. In the second part, we improve the performance of the graph signal reconstructions by exploiting the local properties of graph signals. We propose a node-adaptive regularization with an improved degree of freedom, so a more general signal smoothness assumption is allowed. Different regularization weights design methods are proposed to achieve its best performance. By comparing it with Tikhonov regularization, we observe its superiority in graph signal reconstruction and interpolation, also in graph signal sampling.

Time-varying network data are essential in several real-world applications, such as temperature forecasting and earthquake classification. Spatial and temporal dependencies characterize these data and, therefore, conventional machine learning tools often fail to learn these joint correlations from data. On the one hand, hybrid models to learn from time-varying network data combine several specialized models able to capture these dependencies separately, thus ignoring their joint spatio-temporal interactions. On the other hand, state-of-the-art approaches for jointly learning time-varying network data do not exploit the useful prior provided by a graph-time product graph. This prior structural knowledge can aid learning and help the models improve their performance. For this reason, we propose a novel neural network architecture to learn from time-varying network data using product graphs and graph convolutions, thus exploiting this prior during learning. In particular, our architecture (i) learns the most suitable graph-time product graph to represent the time-varying network data and model the graph-time interactions; (ii) performs graph convolutions over this product graph to learn the graph-time interactions from data; (iii) employs graph-time pooling to reduce the dimensionality over layers. To the best of our knowledge, no research has yet attempted to capture spatial and temporal dependencies using graph convolutions over product graphs. Results on synthetic and real-world data show that the proposed method is useful in learning from time-varying network data for both regression and classification tasks. For classification, we cure a real-world dataset for earthquake classification and compare the proposed approach with state-of-the-art models. Results indicate that graph-aware models outperform graph-unaware models on this task. For regression, we evaluate the proposed method on two real-world temperature datasets and compare our architecture with graph-aware and graph-unaware models. Results on the smaller dataset show that linear models outperform neural-network models. On the larger dataset, we find that prior knowledge about graph-time interactions seems to be less beneficial in case of abundance of data. For the synthetic experiments, we find that (i) learning the structure of the graph-time product graph from data improves the performance compared to adopting a fixed type of product graph; (ii) learning sparser graph-time product graphs further improves the performance; (iii) the proposed graph-time pooling technique contributes to the model's generalization capabilities.

Current subsea LiDAR implementations are inherently depth limited, and make LiDAR applications in the deep-sea costly. To this end, the SLiDAR project aims to develop a pressure tolerant LiDAR system for use at any ocean depth. This thesis elaborates the design and implementation of the laser pulse transmission of the LiDAR and the circuit which will supply the bias voltage for the Avalanche photodiode (APD) from the receiving stage. Although testing of the transmission stage showed the laser can be pulsed, there can be more optimizations done in the future as better laser system can be designed to achieve higher optical output power and an even smaller pulse width. Furthermore, future optimizations should be considered for the APD bias circuit. Unfortunately, the LiDAR system as a whole was not tested in practise, due to time limitations. Hence, this thesis aims to provide a basis for future development, testing and verification of both the LiDAR system and its laser pulse transmission stage.

Graphs are used to model irregular data structures and serve as models to represent/capture the interrelationships between data. The data in graphs are also referred as graph signals. Graph signal processing (GSP) can then be applied which basically extends classical signal processing to solve problems. Anomaly detection is an example of such a problem. Two hypothetical situations are given, and a detector has to be designed to distinguish between these. Under the null hypothesis, graph structures are considered to be untouched. Under the alternative hypothesis, (unknown) topological changes might have occurred. Now by incorporating a priori knowledge about the graphs, the decision making process should improve. In most works, a priori knowledge of the graphs under the null and alternative hypothesis was incorpo- rated. This means that detectors were designed which were able to anticipate on possible topological changes. In this thesis, the problem is considered where only a priori knowledge of the graph under the null hypothesis is exploited. This means that detectors are not able to anticipate on potential changes and this where blind detection comes into play. Blind detection is important because it considers a more realistic scenario. In this work, the blind topology change detector (BTCD) and the constrained blind topology change detector (CTCD) are derived which exploit different properties of the data re- lated to the known graph structure. For the BTCD, the bandlimitedness of graph signals was exploited and for the CTCD, the graph signal smoothness. The main question in this work, was to find out what the potentials are with the blind detection principle for graph change detection. Different test scenarios are used to evaluate the detectors on both synthetic and real data. For the BTCD, the obtained results compare well when information about the alternative graph is available. For this detector, the potential of blind detection was highly visible. For bandlimited graph signals, the BTCD as good as detectors using full information. For the CTCD, comparable results (with detectors using full information) are attained for just a few test scenarios. For small changes, the graph signal smoothness seems to be less powerful as to the graph signal bandlimitedness. This study showed that graph change detection is still possible without having full information. Some graph signal properties are more powerful w.r.t. others.

In the context of the Bachelor Graduation Project at the Delft University of Technology Department of Electrical Engineering, we have been tasked with a project to design prototype of a bi-static radar. This thesis describes the waveform design of a bi-static radar system and its implementation with Software Design Radio, SDR for short. The radar system can be mounted on delivery drones and used to detect other drones and obstacles to avoid collisions. The thesis focuses mainly on the radar waveform design. In addition, the thesis provides analysis of the radar system performance and system implementation with commercial off the shelf SDR. A proof of concept has been realised by means of simulations using the open source software GNU Radio Companion.

Current subsea LiDAR implementations are inherently depth limited, and make LiDAR applications in the deep-sea costly. To this end, the SLiDAR project aims to develop a pressure tolerant LiDAR system for use at any ocean depth. This thesis elaborates the high-level system design of the LiDAR system, as well as the design and implementation of the laser pulse reception stage and the onboard central control unit. Due to the short time frame of the project and the high work load, the LiDAR system as a whole and its subsystems are not tested in practise. Hence, this thesis aims to provide a basis for future development, testing and verification of both the LiDAR system, its laser reception stage, and its central control unit.

Though several sensors are available for underwater scanning and ranging, they all have their limits. SONAR sensors are limited in resolution, and scanning mechanisms using a form of light for carrying the data suffer from high attenuation in turbid waters. The goal of this project was to design a LiDAR system that is capable of overcoming these limitations. To tackle this problem, three groups were formed that focused on different parts of the LiDAR. However, first a complete system overview has been created by the entire project team. First a literature research has been performed. After that, system wide design decisions were discussed and made and also requirements were set. From there on, every subgroup would focus on their own part of the system. In this thesis a more detailed description of the beam steering module shall be given. Again some literature research has been performed. After that, the design decisions were made, based on the requirements. Finally a design was implemented and tested. For the beam steering module to be successful, it should provide accurate control over the angle at which the laser beam is sent out. Though improvements can be made, the system does comply to the minimum accuracy requirement.

The versatility to learn from a handful of samples is the hallmark of human intelligence. Few-shot learning is an endeavour to transcend this capability down to machines. Inspired by the promise and power of probabilistic deep learning, we propose a novel variational inference network for few-shot classification (coined as TRIDENT) to decouple the representation of an image into semantic and label latent variables, and simultaneously infer them in an intertwined fashion. To induce task-awareness, as part of the inference mechanics of TRIDENT, we exploit information across both query and support images of a few-shot task using a novel built-in attention-based transductive feature extraction module (we call AttFEX). Our extensive experimental results corroborate the efficacy of TRIDENT and demonstrate that, using the simplest of backbones, it sets a new state-of-the-art in the most commonly adopted datasets miniImageNet and tieredImageNet (offering up to 4% and 5% improvements, respectively), as well as for the recent challenging cross-domain miniImagenet --> CUB scenario offering a significant margin (up to 20% improvement) beyond the best existing cross-domain baselines.

Problem Definition This bachelor final project was an assignment given by the start up Momo Medical named: "Improving health care with technology". This project was to use their existing technology and ideas to create a system to prevent pressure ulcers. Pressure ulcer are the medical condition that happens when a patient is left too long in the same position in bed. In order to combat this the nurses turn the patient onto a new posture every 3 to 4 hours. To improve upon this Momo Medical decided to built a sensor pad that has to be placed under the mattress in order to track the patients bed posture and to alarm the nurses when a patient has been on a side too long. The project group was divided into three parts: the sensor group who would look at the sensor pad, the algorithm group who would use the data from the sensor pad to determine the bed posture and the test group that had to built a system that could verify the algorithm. This report will talk about the test group. Based upon the data that Momo Medical acquired it was also possible to acquire the heart rate and the respiratory rate from the sensor data. A way to acquire this data was also looked at. The design requirement that were set were 95% accuracy for bed posture and 90% for respiratory rate and heart rate. Design Description In order achieve these requirements a few different types of reference signals were looked at that could be used in order to detect the bed posture, respiratory rate and heart rate. In the end an accelerometer and an commercial grade ECG chest strap were chosen. The accelerometer could measure the patients orientation and thus his bed posture. When the accelerometer is attached to the chest it is also able to measure the chest compressing as a result of breathing. The ECG chest strap was chosen as this gave a good platform upon which the accelerometer and the micro controller could be attached. Evaluation A prototype was built upon our design and it was tested in different situations. In the end the accuracy for the bed posture detection was 99%, the accuracy of the respiratory rate was 91% and of the heart rate it was 95%. With these accuracies it meant that the design requirements were met. The disadvantage was that the system influences the sensor pad while the patient is wearing the prototype. Another disadvantage was that due to the low sampling speed it was not possible to attach the heart rate receiver to the testing group's prototype, but had to be attached to the sensor group's prototype. These two disadvantages could be improved in the future by making the prototype smaller and faster.

In this era of data deluge, we are overwhelmed with massive volumes of extremely complex datasets. Data generated today is complex because it lacks a clear geometric structure, comes in great volumes, and it often contains information from multiple domains. In this thesis, we address these issues and propose two theoretical frameworks to handle such multidomain dataset. To begin with, we extend the recently developed geometric deep learning framework to multidomain graph signals, e.g., time-varying signals, defining a new type of convolutional layer that will allow us to deal with graph signals defined on top of several domains, e.g., electroencephalograms or traffic networks. After discussing its properties and motivating its use, we show how this operation can be efficiently implemented to run on a GPU and demonstrate its generalization abilities on a synthetic dataset. Next, we consider the problem of designing sparse sampling strategies for multidomain signals, which can be represented using tensors. To keep the framework general, we do not restrict ourselves to multidomain signals defined on irregular domains. Nonetheless, this particularization to multidomain graph signals is also presented. To do so, we leverage the multidomain structure of tensor signals and propose to acquire samples using a Kronecker-structured sensing function, thereby circumventing the curse of dimensionality. For designing such sensing functions, we develop several low-complexity greedy algorithms based on submodular optimization methods that compute near-optimal sampling sets. To validate the developed theory, we present several numerical examples, ranging from multi-antenna communications to graph signal processing.