عنوان

فیلترینگ تطبیقی تنک توزیع‌شده در شبکه‌های سنسوری

Number

‭۱۸۹۰۱پ‬

.Language of Text, Soundtrack etc

per

Title Proper

فیلترینگ تطبیقی تنک توزیع‌شده در شبکه‌های سنسوری

First Statement of Responsibility

/حمید شیری

Name of Publisher, Distributor, etc.

: مهندسی برق و کامپیوتر

Date of Publication, Distribution, etc.

، ‮‭۱۳۹۶‬

Name of Manufacturer

، افشار

Text of Note

زبان چکیده: استاد مشاور

Text of Note

چاپی

Dissertation or thesis details and type of degree

دکتری

Discipline of degree

مهندسی برق- مخابرات

Date of degree

‮‭۱۳۹۶/۱۱/۱۷‬

Body granting the degree

تبریز

Text of Note

دانشور، سبلان

Text of Note

.Recent advances in micro-electromechanical systems technology (MEMS) have made it possible to design low-cost intelligent sensors which can perform multiple functions such as measurement, computing, and communication. Such intelligent devices are networked using wireless links, which are known as wireless sensor networks (WSNs). In recent years, distributed adaptive estimation has been considered as a proper method for estimation problems in WSNs. It is shown that the distributed estimation methods have good performance in terms of the error value, convergence rate, and resistance to node and link failures in WSNs. In this dissertation, we examine the problem of distributed estimation of a sparse signal where a group of nodes collaborates with each other in order to estimate a sparse parameter vector. The problem of distributed estimation has many challenges. In this thesis, two states of resource constraint in the networks are considered. The first state is when the network is not rich in terms of energy and memory resources, and the second state is when the network has fewer restrictions in terms of energy and memory.In the first case, the distributed adaptive estimation problem based on the diffusion strategy is considered, and then the sparsity of the signal is used to improve the estimation error and reducing communication costs. The main focus of most distributed estimation works is on the convergence rate and the estimation error, while communications between the nodes require a lot of transmissions (and thereby energy). Therefore, the goal is to provide a solution based on the diffusion strategy and to exploit the sparsity of the signal to reduce communication cost. Here, a new version of the sparse-aware diffusion LMS algorithm is proposed to consider both the communication and error costs. Then the computational complexity and communication cost for each node in the network are obtained. The efficiency of the proposed method is compared with the well-known existing method through simulations in terms of computation and communication cost, and flexibility to signal changes

شیری، حمید

طینتی، محمد علی، استاد راهنما

دانشور، سبلان، استاد مشاور

نمایه‌سازی قبلی