Microscopy is a crucial tool across various scientific domains. Due to light diffraction, the images acquired from optical microscopes are usually blurry and corrupted by noise. For an accurate quantitative analysis, the measured images need to be deconvolved to achieve higher resolution. Deconvolution processes are computationally expensive, due to the large data size. This leads to out-of-memory issues and extended computation time.

To address these problems, this project aims to develop a novel convolution scheme. It utilizes the special structure of the Point Spread Function, which in conventional microscopy techniques has most of its energy concentrated in the center. And implement multi-resolution signal processing methods. This approach enhances computational efficiency while retaining computational accuracy.

In autonomous driving, environmental perception, crucial for navigation and decision-making, depends on integrating data from multiple sensors like cameras and LiDAR. Camera-LiDAR fusion combines detailed imagery with precise depth, improving environmental awareness. Effective data fusion requires accurate extrinsic calibration to align camera and LiDAR data under one coordinate system. We aim to calibrate the camera and LiDAR extrinsic automatically and without specific targets. Targetless, non-automated calibration methods are time-consuming and labor-intensive. Existing advanced methods have proven that automatic calibration methods based on edge features are effective, and most focus on the extraction and matching of single features. The proposed method matches 2D edges from LiDAR's multi-attribute density map with image-derived intensity gradient and semantic edges, facilitating 2D-2D edge registration. We innovate by incorporating semantic feature and addressing random initial setting through the PnP problem of centroid pairs, enhancing the convergence of the objective function. We introduce a weighted multi-frame averaging technique, considering frame correlation and semantic importance, for smoother calibration. Tested on the KITTI dataset, it surpasses four current methods in single-frame tests and shows more robustness in multi-frame tests than MulFEAT.
Our algorithm leverages semantic information for extrinsic calibration, striking a balance between network complexity and robustness. Future enhancements may include using machine learning to convert sparse matrices to dense formats for improved optimization efficiency.

The brain stands as the most powerful processor in the known universe. It generates a continuous stream of electrical and chemical signals that underpin every thought, sensation, and action. Our past efforts in decoding these signals have made it possible to diagnose and treat many neurological disorders, helped us gain a deeper understanding of cognitive processes and consciousness, and paved the way for brain-computer interfaces. To take another step forward in our long but rewarding journey of discovering the brain's complex organization, we rely on advances in imaging technologies and signal processing.

Functional ultrasound is a neuroimaging technique that has emerged in the last decade, and has gained remarkable attention since then. The popularity of this technique stems from its portability, high resolution and affordability. Functional ultrasound can detect subtle fluctuations in local blood dynamics, which serve as delayed indicators of the underlying changes in neuronal activity. The goal of this thesis is to develop novel signal models and processing algorithms that can reveal the spatial and temporal characteristics of hemodynamic activity induced by external stimuli using functional ultrasound.

Existing techniques that explore how the brain reacts in response to stimuli model the design variables using a linear time-invariant system with binarized input representations, marking when a stimulus is on or off. However, experimental evidence suggests that the brain reacts in a more intricate manner. While some regions exhibit consistent responses to repeated stimuli, others can show substantial variation even when exposed to the same stimulus seconds apart. In our in-vivo experiments, we particularly focus on key regions within the mouse visual processing pathway, which are analogous to those in the human brain. We track how visual information flows across these areas, and propose methods that can incorporate the spatiotemporal variability of brain responses when identifying evoked activity. Using these methods, we show that functional ultrasound can capture the dynamic nature of brain responses with high spatial and temporal resolution, and provide us with further insights into the functional organization of the brain. Future directions of this dissertation include multimodal processing of the functional ultrasound signal together with neuronal activity, aiming to enhance our understanding of neurovascular coupling.@en

Rank detection is crucial in array processing applications, as many algorithms rely on accurately estimating the rank of the data matrix to ensure optimal performance. Under Gaussian white noise, rank can be detected through eigenvalue analysis. However, in arbitrary noise, prewhitening the data matrix with the noise covariance matrix is necessary, and rank detection is achieved by examining the generalized eigenvalues. Existing methods often assume the noise covariance structure or require a large number of noise samples. This thesis focuses on addressing the rank detection problem in scenarios with limited noise samples and arbitrary noise environments.

Firstly, we investigate the largest generalized eigenvalue threshold for the prewhitened data sample covariance matrix according to the random matrix theory. We develop a rank detection algorithm based on the threshold via a sequential test, and provide the performance analysis. A series of simulations demonstrate its superiority over conventional methods such as Minimum Description Length (MDL) and Akaike's Information Criterion (AIC).

Secondly, since the Short-time Fourier Transform (STFT) is commonly used for non-stationary signal analysis, we extend our rank detection method to the STFT domain. The correlations introduced by the STFT have a significant impact on the distribution of the noise. Therefore, we develop a technique to remove correlations among time-frequency bins based on exact expressions of these correlations. After successfully eliminating these correlations, our proposed rank detection method achieves enhanced reliability and performance in the STFT domain.

Lastly, we evaluate the effectiveness of our rank detection method in speech enhancement applications. Simulations confirm that utilizing the estimated rank improves speech quality compared to using the known number of sources.

This project develops a channel estimation technique for millimeter wave (mmWave) communication systems. Our method exploits the sparse structure in mmWave channels for low training overhead and accounts for the phase errors in the channel measurements due to phase noise at the oscillator. Specifically, in IEEE 802.11ad/ay-based mmWave systems, the phase errors within a beam refinement protocol packet are almost the same, while the errors across different packets are substantially different.

We show that standard compressed sensing algorithms that treat phase noise as a constant fail when channel measurements are acquired over multiple beam refinement protocol packets. Most of the methods that have addressed this problem treat phase noise as purely random, missing the inherent structure within the measurement packets. We present a novel algorithm called partially coherent matching pursuit for sparse channel estimation under practical phase noise perturbations. The proposed approach leverages this partially coherent structure in the phase errors across multiple packets. Our algorithm iteratively detects the support of sparse signal and employs alternating minimization to jointly estimate the signal and the phase errors.

We numerically show that our algorithm can reconstruct the channel accurately at a lower complexity than the benchmarks, and derive a preliminary support detection bound as a performance guarantee.

This report entails one of two subsystems in a joint project to provide a web-based platform for smartwatch data acquisition, for applications in healthcare. In this work, we design and implement algorithms for human activity recognition using various machine learning approaches and test them on data acquired online as well as using our own developed platform. Together with the web-based platform, this provides a solid base for more research using data gathered from smartwatches. The human activity recognition is implemented first using a classical machine learning approach
with feature extraction and a random forest classifier. Next, both convolutional neural network and a recurrent neural network are implemented using Tensorflow [1]. We further perform several tests to investigate: (i) the optimal segment size with respect to classification accuracy, (ii) the effect of filtering and preprocessing on the classification results, and (iii) the best classifier for activity detection.

This report details the design and implementation of a subsystem providing a web-based platform for smartwatch data acquisition. A smartwatch application is developed to read out the accelerometer, gyroscope and heart rate sensor on the smartwatch and transmit the sensor data to the platform. A platform is developed to receive the data and store it in a database. Recordings can be downloaded from the platform to use in research of human activity. The platform presents the type of activity the smartwatch user is doing using machine learning models, by integrating the other subsystem. Multiple tests have been performed to analyse and improve the performance of the system. The battery life of the smartwatch has been tested using various settings in the smartwatch application to determine the most power-efficient settings.

We propose the Stationary spectrum Plus Low-rank Iterative TransmiTtance EstimatoR (SPLITTER) for removing wideband atmospheric noise from observations of high-redshift galaxies. This algorithm has specifically been developed for the DEep Spectroscopic HIgh-redshift MApper (DESHIMA) 2.0, a spectrometer that is designed to observe the waveband from 220 GHz to 440 GHz in 347 spectral channels. This octave bandwidth poses a challenge, due to the spectrotemporal changes in the atmosphere column between the instrument and the target source. Removing the time-varying nonlinear interference and distortion caused by the atmosphere is a difficult task, as the atmospheric emission is much stronger than a typical galaxy signal.

The goal of this thesis is to develop a method that can estimate both narrow spectral lines and the broad continuum emission with a higher sensitivity than the currently used method of directly subtracting noisy on- and off-source spectra. We develop a logarithmic data model for separating atmospheric noise from the galaxy signal in position switching-observations. Because the atmospheric transmittance appears as a multiplicative term in both the atmospheric interference and the signal modulation, the logarithmic model allows for an additive decomposition of the data. The atmospheric transmittance behaves as a low-rank component in this model.

Using the model, we develop an optimization algorithm (SPLITTER) to perform the separation of the signal and the low-rank atmospheric transmittance. Several implementations are discussed. The final algorithm uses a Singular Value Decomposition (SVD) to estimate the atmosphere component and the Alternating Directions Method of Multipliers (ADMM) for estimating the source signal. Instead of subtracting the noisy estimate of the source from the data directly, a denoised model is used in this step, such that we can trade some spectral resolution for a higher sensitivity.
SPLITTER is tested on simulated data using the Time-dependent End-to-end Model for Post-process Optimization of the DESHIMA spectrometer (TiEMPO), a dedicated software package for simulating DESHIMA observations. We show that SPLITTER is able to estimate the spectrum with a higher sensitivity than the conventional method. The improvement factor in our weighted root mean squared error is up to ~1.7 for the full spectrum and up to ~1.3 for the spectral lines only compared to the conventional method. The larger improvement for the full spectrum is achieved by trading spectral resolution for a higher sensitivity in the smooth continuum. With these results, we have an indication that a statistically driven method for DESHIMA observations can provide better estimates than the current method with the same amount of observing time.

More work is needed to create a robust version of the algorithm, because although the sensitivity benefit of SPLITTER is larger in the continuum regions, there are also situations where the continuum is overestimated. The conditions for this to occur are not yet clear. A more robust version could make SPLITTER a reliable new method that can replace current data reduction methods for wideband atmospheric noise removal. In this way, it can be used to make background-limited direct detection spectrometers on both existing and future telescopes observe more efficiently.

The objective of this thesis was to develop a rodent tracker using the FlashTrack implants, and allow for tracking data to be used for behavioural research. The task was split into three main problems: detection, tracking and error detection. The detection was solved with basic image processing. The first step was background removal, using median filtering. Later, blob detection after some processing was done. Detections were differentiated into cases where mice are together in contact events and where they are alone. Bounding boxes were generated for the contours of lone mice, while the distance transform was employed to detect the joint ones. To track the moving targets, Kalman filtering was used on the bounding boxes of the detector. This approach was based on the Simple Online Realtime Tracking framework, adapting it to the particularities of mice. Other approaches were tried, but SORT was the chosen one. The infrared subcutaneous implants of FlashTrack allow for identity verification through code detections. To process the codes of each track, a Gaussian Mixture Model is trained to be the classifier of the detections. A track handling module was built to monitor the estimated tracks and code detections, and verify correct assignments. Detected erroneous tracks were discarded. Synthetic data was used to evaluate the tool. Artificial datasets were developed in Blender. Common metrics for evaluation of multiple object trackers were gathered and discussed, as well as a comparison with one of the state of the art animal trackers.

Currently, most of FMCW radar systems for target detection and localization are based on the radar system with multiple receiving antennas, but little based on the SISO system. In this project, we will show a unique signal processing pipeline based on the 8 GHz SISO FMCW radar system. An advanced algorithm of multi-target detection and tracking will be designed to monitor the range, angle, and Doppler velocity information of targets.

With the ever-expanding need for accuracy in the world of navigation, Global Navigation Satellite Systems(GNSS) such as GPS and Galileo have become the primary option around the world. As such, the potential damage that can be caused by malicious tampering with the receivers continues to grow. One such threat, known as Spoofing, comprises of transmission of altered GNSS signals to a target receiver(s). This process leads to false positional and/or time data on the receiver's end. Spoofing has been a topic of discussion for roughly two decades. With many theoretical approaches to its detection, this work focuses on the Angle of Arrival technique via the construction of a synthetic array from a single moving element, aided by positional information provided by Inertial Measurement Unit (IMU). This method relies on the periodic nature of the L1 signal, and focuses primarily on the case of GPS as an example. Using simulation and sample data, the possibility and limitations of constructing virtual antenna arrays is explored. It is shown that despite being viable for low number of sources during the simulation, the complexity of signal propagation within the real world implementation of GNSS system renders this technique inoperable in its first iteration.

In the last two decades, a lot of attention has been focused on con- tactless radar-based vital signs monitoring (heartbeat and respiration rate) as an emerging and complementary value to our medical care. It is very challenging in real indoor environments to perform concurrent localization and reliable vital signs monitoring of multiple subjects within practical distance ranges. In fact, the multipath propagation results in the reflected signal dispersed in time, which not only causes false ToF (Time of Flight) estimation but also leads to inter-subject interference, jeopardizing the vital signs extraction and the localiza- tion. Here we show a methodology based on radar techniques to auto- matically locate multiple subjects in indoor environments while keep monitoring their vital signs. This approach, based on the paramet- ric models both of the propagation channel and of the radar signals, is able to cancel the undesired contributions from static clutters and multipath components, by which it is possible to accurately locate the subjects and extract their heart rates and respiration rates.

LoRa (Long Range) is a low-power, long-range and low-cost wireless communication system that can facilitate a wide variety of infrastructures for the Internet of Things (IoT). Current algorithms to locate LoRa tags have a resolution of 100 m in practice, and a question is if that can be improved without changing the tags or adding too much to the gateways (basestations). Conventional delay estimation ranging algorithms extract useful information from the channel frequency response and use this information to estimate delays. In this thesis, three localization techniques are presented: the matched filter, FBCM-MUSIC and TLS-ESPRIT algorithms. Then a multiband architecture is proposed and integrated into the matched filter. These algorithms are implemented in the LoRa system model. The simulations indicate that FBCM-MUSIC and TLS-ESPRIT have better performance than the matched filter in NLOS channels. The results also show that TLS-ESPRIT is more effective and robust compared to MUSIC. The proposed multiband architecture can improve the resolution of TOA estimation and decreases the 90th percentile error by around 40%.

Radio astronomy image formation can be treated as a linear inverse problem. However, due to physical limitations, this inverse problem is ill-posed. To overcome the ill-posedness, side information should be involved. Based on the sparsity assumption of the sky image, we involve l1-regularization. We formulate the image formation problem into a l1-regularized weighted least square (WLS) problem and associate each variable with one regularization parameter. We use Bayesian learning to learn the regularization parameters from data by maximizing the posterior density. With the iterative update of the regularization parameters, the solution is updated until convergence of the regularization parameters. We involve a stopping rule based on the noise level to improve the computational efficiency and control the sparsity of the solution. We compare the performance of this Bayesian learning method with other existing imaging methods by simulations. Finally, we propose some future research directions in improving the performance of this Bayesian learning method.

Indoor positioning using Bluetooth addressed a great concern. The properties of narrowband usage, low energy consumption and universality on devices attract numerous customers and promote researchers to investigate potential of Bluetooth in indoor positioning field. It has already supported Angle-of-Arrival (AoA) and Angle-of-Departure (AoD) in angle domain for indoor localization. In range domain, using Received Signal Strength (RSS) is practicable but it cannot provide enough resolution. It is needed to develop a technique that can improve resolution for range finding feature. The challenges of localization in indoor environments are mainly from the multipath propagation of signals. In this thesis, a subspace-based super-resolution algorithm for time delay dispersion estimate is developed. Range finding is realized by computing time-of-arrival (ToA) parameter of signal arriving at direct line-of-sight (DLoS). An essential procedure before the developed algorithm is applied is that to correctly separate subspace into signal space and noise space. Techniques for subspace separation are investigated in this thesis. In this thesis, we want to explore the potential of indoor localization using Bluetooth narrowband radios. To start with, a data model according to the property of the conducted measurement data is developed. The conducted measurement data is radio channel measurements based on channel sounding technique. Then the data model is developed as channel impulse response model and multipath signals are indicated by different time delays. Since an accurate covariance matrix of measurement data is required for super-resolution algorithm, smoothing techniques is employed. The smoothing techniques considered are forward smoothing technique and forward-backward smoothing technique. For the purpose of obtaining an accurate subspace separation, two techniques are investigated in this thesis, namely MDL criteria algorithm and the threshold method. in order to investigate the performance and reliability of those two techniques, experiments are taken out using different parameter values. Comparison is made between the results of these two techniques. Afterwards, subspace-based super-resolution algorithm is taken into consideration. In this thesis, the super-resolution algorithm implemented is MUSIC algorithm. The functionality of MUSIC algorithm on narrowband radios measurements is tested and evaluated firstly by simulation experiments, which demonstrates the practicability of applying MUSIC algorithm on narrowband radios measurements. Then experiments are extended to the measurement data that conducted from real indoor environments, for the purpose of indoor localization realization using narrowband radios.

In the next generation of Bluetooth standard, the Bluetooth SIG wants to incorporate multiple antenna systems into the Bluetooth Low Energy specification to enable direction-finding features. The features are aimed to improve the accuracy of off-the-shelf Asset Tracking Profile (ATP) and Indoor Positioning Service (IPS) including two modes – Angle-of-Arrival (AoA) mode and Angle-of-Departure (AoD) mode. In this thesis, we only focus on the AoA mode.

The new standard raises several challenges. First, the direction finding algorithm shall be derived sincethe standard gives only the framework. The algorithm shall cope with dense multipath effects in indoorenvironments and identify the angle of Line-of-Sight (LOS) component. Second, the new standard specifiesthe usage of an RF switch such that a single receiver can access multiple antennas. This mechanism reducesthe device cost and complexity but poses difficulties to the array processing. There are inevitably informationloss during antenna switching. It also raises requirements of channel stationarity and efficient compensationof CFO. Third, towards the system implementation, practical considerations that deviate the ideal datamodelshall be taken into account. These considerations include the effect of mutual coupling (MC), and the phase imbalance of the RF switch. During this project, these effects have been studied to obtain insight on theinfluence on algorithmperformance and compensation techniques.

In this thesis, we formulated the data model for a single receiver using a uniformlinear multiple antennasystem with an RF switch. The importance of CFO compensation, channel stationarity, and the color of noiseare addressed. A maximum likelihood (ML) based CFO estimation algorithm is proposed. Furthermore, we modeled the effect of mutual coupling and imbalance of switch. Next, we analyzed why the delay estimation is not feasible within the context of Bluetooth LE. We proposed two Line-of-Sight direction identification (LOS-Id) algorithms based on the power signature in the data covariance matrix, which are referred to as MUSIC LOS-Id and CLEAN-MUSIC LOS-Id. Further performance improvements are achieved by making use of the frequency hopping feature of Bluetooth. By aggregatingmore than one packets at different frequencies, the performance can be improved substantially. This technique is called the multi-tone technique, or packet aggregation (PA).

For evaluating the effectiveness of the proposed methods and models, a Bluetooth LE simulator is built. The performance verification is divided into two phases that differentiate themselves by the channel model. In the first phase, a simulated channel model, which is obtained by applying the ray tracer in an empty rectangular room, is used. The mutual coupling effect is simulated using the Antenna Toolbox in Matlab. The switch characteristics are verified by measurements using a Vector Network Analyzer (VNA). In the second phase, the real channel is measured with the VNA. Three campaigns of measurements are carried out with a 1x4, 1x8, and 2x4 antenna array respectively. Performance is evaluated by applying both channel models. The simulations reveal that the multipath effect is the dominant influencing factor of the performance in our indoor scenario, while the mutual coupling and the switch imbalance have little influence. The results also show that both proposed LOS-Id algorithms yield satisfying accuracy. However, we paid less attention to the CLEAN-MUSIC algorithm because of its complexity even though it indeed performs better than MUSIC LOS-Id in our simulated scenario. Finally, the usage of multi-tone technique improves the LOS-Id performance substantially. With an 8-element ULA and aggregating 8 tones, the MUSIC LOS-Id algorithm can achieve 10 degrees of RMSE for 90% of transmission positions with measured channels, and 3 degrees of RMSE for 50% of transmission positions.

Hearing aids as a form of audio preprocessing is increasingly common in everyday life. The goal of this thesis is to implement a blind approach to the cocktail party problem and challenge some of the regular assumptions made in literature. We approach the problemas wideband FD-BSS. From this field of research, the common assumption of contineous activity is dropped. Instead a number of users detection is implemented as a preprocessing step and ensure the appropriate number of demixing vectors for each time frequency bin. The validity of the standard mixing model used for STFT’s is challenged by looking at the response of a linear array. Source separation is achieved by demixing vectors based on the GSVD, derived in a model-based approach. While most ermutation solvers offer an a posteriori solution for all users, we looked at finding local solutions for a single user. Combining this with the user identification called the alignment step, we conclude that the permutation problem can be reduced to selecting a demixing vector for each discrete time-frequency instance.
The correlation coefficient proves to be a sufficient metric to couple reconstructions to the original data as it selects most of the active time-frequency bins. In the far-field case, our approach performs in a comparable
but not superior manner. We did find that our method is much more robust against inaccuracies introduced when narrowband channels are assumed but not actually available. This is strongly exemplified by our experiment of a changing DFT-size.
The Frobinius norm was suggested as a measure of distance between the estimate STFT and the orignial signals time frequency domain description but it resulted in counter intuitive results which didn’t correspond with other metrics used in this thesis. It is expected that there are effects induced by changing the size of the STFT which are not accounted for.
Our demixing vectors achieve comparable inteligibility, measured by STOI, as the compared techniques and it is more robust against smaller sample sizes than the theoretically SINR optimal MVDR.

Signal Analysis techniques are routinely used in Neuroscience to interpret raw signals harvested from the Nervous System. From a simple Fourier analysis to more complicated methods such as multiresolution wavelet analysis, such techniques must be used for signal manipulation in order to reach informed conclusions on the measurements taking place.

In neuroscientific research, it is common practice to scan a neural recording manually to label the areas of the signal that are relevant to the research at hand. This can, obviously, be very time-consuming for the researchers. What is more, this method can prove imperfect, seeing as two different researchers can disagree on the labeling of the data.

In every experiment, signal epochs are isolated because they stand out from the rest of the recording due to a special characteristic, which is, in the previous case, visible with the unaided eye. In signal processing terms, this means that the signal displays specific spectrotemporal characteristics during these epochs. Thus, these characteristics can be isolated, quantified and studied independently, while a detection algorithm can be developed so that the detection and labeling of the significant signal epochs can be carried out automatically.

In this project, the spontaneous activity of the neurons in the Prefrontal Cortex was analyzed in relation to neural synchrony, using time-varying ARModels. It was concluded that the signal epochs of neural synchrony display common characteristics besides being visually similar. This allowed the isolation of the synchrony epochs based on model parameters.

However, the training of the models is very computationally intensive, so a detection algorithm was developed, based on a matched filter which made use of one of the isolated epochs as a template. The detection scheme was then validated using a recording harvested during electrical stimulation of the deep brain, evaluating the quality of the scheme was evaluated.

Finally, the detection scheme was applied to stimulation recordings to study the electrical behavior of the Prefrontal Cortex during electrical stimulation of deep brain structures.

Atrial fibrillation (AF) is one of the more common clinical arrhythmias with a high morbidity and mortality. Despite this, the electrophysiological and pathological mechanisms associated with AF largely remain a mystery, encouraging the use of ever more sophisticated techniques to extract vital information for diagnostic and therapeutic purposes. Contamination by signals of ventricular origin is considered the main artifact present in high-resolution epicardial electrograms (EGMs) that hinders the accurate and efficient analysis of AF EGM datasets. Furthermore, the complexity and dynamism of AF signals calls for robust data analysis tools that can effectively reduce or remove ventricular activity (VA) while preserving the texture and morphology of atrial activity (AA). Multiway component analysis, specifically block term decomposition (BTD), proves useful for the decontamination of epicardial EGMs as demonstrated in this project by enabling the automatic estimation of VA on an electrode-by-electrode basis, which is thereafter temporally and/or power spectrally subtracted thus retaining AA at a relatively high accuracy. The performance of BTD compared to average beat subtraction (ABS) and the more restrictive canonical polyadic decomposition (CPD) is visually verified and numerically confirmed based on a set of key performance indices. Additionally, the technique is entirely data-driven i.e., does not depend on any statistical properties, but if/when available, can contribute to enhanced performance via the imposition of appropriate constraints in the tensor decomposition.

In today's society, we are flooded with massive volumes of data in the order of a billion gigabytes on a daily basis from pervasive sensors. It is becoming increasingly challenging to locally store and transport the acquired data to a central location for signal/data processing (i.e., for inference). To alleviate these problems, it is evident that there is an urgent need to significantly reduce the sensing cost (i.e., the number of expensive sensors) as well as the related memory and bandwidth requirements by developing unconventional sensing mechanisms to extract as much information as possible yet collecting fewer data. The first aim of this thesis is to develop theory and algorithms for data reduction. We develop a data reduction tool called sparse sensing, which consists of a deterministic and structured sensing function (guided by a sparse vector) that is optimally designed to achieve a desired inference performance with the reduced number of data samples. The first part of this thesis is dedicated to the development of sparse sensing mechanisms and convex programs to efficiently design sparse sensing functions. We design sparse sensing functions under the assumption that the data is not yet available and the model information is perfectly known. Sparse sensing offers a number of advantages over compressed sensing (a state-of-the-art data reduction method for sparse signal recovery). One of the major differences is that in sparse sensing the underlying signals need not be sparse. This allows us to consider general signal processing tasks (not just sparse signal recovery) under the proposed sparse sensing framework. Specifically, we focus on fundamental statistical inference tasks, like estimation, filtering, and detection. In essence, we present topics that transform classical (e.g., random or uniform) sensing methods to low-cost data acquisition mechanisms tailored for specific inference tasks. The developed framework can be applied to sensor selection, sensor placement, or sensor scheduling, for example. In the second part of this thesis, we focus on some applications related to distributed sampling using sensor networks. Sensor networks can be used as a spatial sampling device, that is, to faithfully represent distributed signals (e.g., a spatially varying phenomenon such as a temperature field). On top of that, the distributed signals can exist in space and time, where the temporal sampling is achieved using analog-to-digital converters, for example. Each sensor has an independent sample clock, and its stability essentially determines the alignment of the temporal sampling grid across the sensors. Due to imperfections in the oscillator, the sample clocks drift from each other, resulting in the misalignment of the temporal sampling grids. To overcome this issue, we devise a mechanism to distribute the sample clock wirelessly. Specifically, we perform wireless clock synchronization based on the time-of-flight measurements of broadcast messages. In addition, clock synchronization also plays a central role in other time-based sensor network applications such as localization. Localization is increasingly gaining popularity in many applications, especially for monitoring environments beyond human reach, e.g., using robots or drones with several sensor units mounted on it. Consequently we now have to localize more than one sensor or even localize the whole sensing platform. Therefore, we extend the classical localization paradigm to localize a (rigid) sensing platform by exploiting the knowledge of the sensor placement on the platform. In particular, we develop algorithms for rigid body localization, i.e., for estimating the position and orientation of the rigid platform using distance measurements. Given the central role of sensing and sensor networks, the results presented in this thesis impacts a wide range of applications.@en