Distributed and Collaborative Processing in Wireless Sensor Networks
Wireless sensor networks, formed by numerous tiny devices capable of sensing, computing, and wireless communication, are emerging as a revolutionary technology with applications in diverse areas. The unique features of wireless sensor networks, and in particular the power scarcity of sensor nodes, have brought new challenges and problems to the field of distributed and collaborative information processing. In this dissertation, we address some important problems within this broad topic, including data gathering, distributed detection, and distributed consensus, with the emphasis on efficient use of stringent system resources to achieve certain application-specific objectives.
We start in Chapter 2 with an investigation of schemes for collecting sensor data at a sink node, which are basic building blocks for all sensor network applications in hierarchical networks. A central problem is to explore the inherent tradeoff between two inconsistent performance measures: throughput and energy efficiency. We consider a cross-layer framework, where spatial diversity is exploited through multiuser detection techniques at the physical layer to achieve dramatically increased throughput, and deterministic or randomized medium access methods are employed to avoid excessive interference. Our results on the optimal performance of different medium access control schemes coupled with different linear multiuser detectors provide useful insights into cross-layer design in WSN.
Most existing works on distributed detection have studied optimal local mapping rules and fusion rules without considering communication constraints and power-efficiency. We improve such
designs by treating communication jointly with decision fusion. In Chapter 3, distributed detection over a multi-access channel (MAC) with correlated sensor observations is considered. MAC fusion is motivated by its much smaller bandwidth requirement compared with fusion over a parallel-access channel (PAC) which is widely assumed in literature. Correlated observations arise naturally from the dense deployment of sensors nodes. We consider two exemplary problems: detection of a deterministic signal in correlated Gaussian noise and detection of a first-order autoregressive signal in independent Gaussian noise. It is shown that in addition to vastly improved bandwidth efficiency, MAC fusion with optimal local mapping rules yields better detection performance as measured by error exponents compared with PAC fusion under the same transmission power constraint. Subsequently, in Chapter 4, we investigate distributed detection via multi-hop transmissions, which significantly reduces the energy consumption over direct transmission in networks where nodes have widely different distances towards the fusion center. Several multihop fusion rules, including Multihop Forwarding (MF), Histogram Fusion (HF) and Log-likelihood Ratio Fusion (LF) are investigated. We demonstrate how transmission structures are designed along with fusion rules to achieve optimal tradeoffs between detection performance and energy efficiency.
Distributed consensus where nodes aim to reach agreement on their average value through iterative local information exchange is studied in Chapter 5 and 6. Our purpose is to mitigate the slow convergence and high communication cost of known distributed consensus algorithms in literature. In Chapter 5, we incorporate clustering techniques to distributed consensus, which provides the dual benefits of avoiding redundant transmissions and improving the network connectivity. Cluster-based distributed averaging algorithms in forms of fixed iteration and random gossiping are proposed. Both are shown to provide faster convergence and reduced communication and computation complexity than their non-cluster-based variants. While the above schemes use reversible Markov chains, in Chapter 6 we explore distributed consensus based on nonreversible Markov chains, which are known to converge to stationarity faster than reversible ones by suppressing the diffusive behavior. We propose a class of Location-Aided Distributed Averaging (LADA) algorithms, where nodes' location information is used to construct nonreversible chains with fast-mixing properties. Our analysis and numerical results demonstrate that LADA algorithms significantly outperform known algorithms based on reversible chains.
Advisor:Huaiyu Dai; Brian Hughes; Alexandra Duel-Hallen; Hamid Krim; Hien Tran
School:North Carolina State University
School Location:USA - North Carolina
Source Type:Master's Thesis
Date of Publication:08/21/2007