We consider a multiterminal sourcecoding problem in which a random source signal is estimated from encoded versions of multiple noisy observations. Each encoded version, however, is compressed so as to minimize a loc...
详细信息
ISBN:
(纸本)9781509018062
We consider a multiterminal sourcecoding problem in which a random source signal is estimated from encoded versions of multiple noisy observations. Each encoded version, however, is compressed so as to minimize a local distortion measure, defined only with respect to the distribution of the corresponding noisy observation. The original source is then estimated from these compressed noisy observations. We denote the minimal distortion under this coding scheme as the compress-and-estimate distortion-rate function (CE-DRF). We derive a single-letter expression for the CE-DRF in the case of an i.i.d source. We evaluate this expression for the case of a Gaussian source observed through multiple parallel AWGN channels and quadratic distortion and in the case of a non-uniform binary i.i.d source observed through multiple binary symmetric channels under Hamming distortion. For the case of a Gaussian source, we compare the performance for centralized encoding versus that of distributed encoding. In the centralized encoding scenario, when the code rates are sufficiently small, there is no loss of performance compared to the indirect sourcecoding distortionrate function, whereas distributed encoding achieves distortion strictly larger then the optimal multiterminal sourcecoding scheme. For the case of a binary source, we show that even with a single observation, the CE-DRF is strictly larger than that of indirect sourcecoding.
Wireless sensor networks (WSNs) consisting of battery-powered sensors are increasingly deployed for a myriad of Internet of Things applications, e. g., environmental, industrial, and healthcare monitoring. Since wirel...
详细信息
Wireless sensor networks (WSNs) consisting of battery-powered sensors are increasingly deployed for a myriad of Internet of Things applications, e. g., environmental, industrial, and healthcare monitoring. Since wireless access is typically the main contributor to battery usage, minimizing communications is crucial to prolong network lifetime and improve user experience. The objective of this thesis is to develop and analyze energy-efficient distributed compressed data acquisition techniques for WSNs. The thesis proposes four approaches to conserve sensors energy by minimizing the amount of information each sensor has to transmit to meet given application requirements. The first part addresses a cross-layer design to minimize the sensors sum transmit power via joint optimization of resource allocation and multi-path routing. A distributed consensus optimization based algorithm is proposed to solve the problem. The algorithm is shown to have superior convergence compared to several baselines. The remaining parts deal with compressed sensing (CS) of sparse/compressible sources. The second part focuses on the distributed CS acquisition of spatially and temporally correlated sensor data streams. A CS algorithm based on sliding window and recursive decoding is developed. The method is shown to achieve higher reconstruction accuracy with fewer transmissions and less decoding delay and complexity compared to several baselines, and to progressively refine past estimates. The last two approaches incorporate the quantization of CS measurements and focus on lossy sourcecoding. The third part addresses the distributed quantized CS (QCS) acquisition of correlated sparse sources. A distortion-rate optimized variable-rate QCS method is proposed. The method is shown to achieve higher distortion-rate performance than the baselines and to enable a trade-off between compression performance and encoding complexity via the pre-quantization of measurements. The fourth part investigates
暂无评论