Wireless sensor network: energy efficiency, security, and, fault tolerance

Wireless sensor network (WSN) is an emerging networking paradigm that promises a wide range of potential applications in both civilian and military areas. WSN runs different kinds of applications in a variety of physical environments, which offers many challenges. The main design constraints include energy efficiency, fault tolerence, and security. In this thesis, we investigate the research problems involved in three types of sensor networks including the Under Water Sensor Network (UWSN), the Wireless Terrestrial Sensor Network (WTSN), and the Wireless Multimedia Sensor Network (WMSN). -- We first formulate the node clustering problem into a cluster-centric cost-based optimization problem with an objective to improve the energy efficiency and prolong the network lifetime in UWSN. A novel distributed clustering protocol called minimum-cost clustering protocol (MCCP) is proposed, which selects a set of non-overlapping clusters from all potential clusters based on the cost metric assigned to each potential cluster and attempts to minimize the overall cost of the selected clusters. To provide a robust clustered architecture against cluster-head failures in UWSNs, a dependable clustering protocol is proposed in which two mechanisms are employed: fault prevention clustering and cluster head replication. Fault prevention clustering attempts to select those healthy nodes as cluster heads to prevent cluster head failures, and cluster head replication attempts to select a primary cluster head and a backup cluster head for each cluster member so that the constructed cluster hierarchy can tolerate cluster-head failures. -- The successful working of any fault recovery schemes heavily depends on a proper and efficient fault detection mechanism. Therefore, we propose a cooperative fault detection mechanism, which can accurately and quickly detect the failure of a cluster head through the independent fault status detection from each cluster member and a distributed agreement process for final decision. It runs concurrently with normal network operation at each cluster member and makes use of the data periodically sent by a cluster head as the heartbeats for fault detection. An agreement can be efficiently achieved within two TDM frames in each detection process. To address the energy efficiency problem in WTSN, Slepian-Wolf coding is employed to remove the redundancy caused by the data correlation. We first consider the clustered Slepian-Wolf coding problem, which aims at selecting a set of disjoint potential clusters to cover the whole network such that the global compression gain of Slepian-Wolf coding is maximized, and then propose a distributed optimal-compression clustering protocol to solve the problem. Based on the resulting cluster hierarchy constructed, we study the optimal intra-cluster rate allocation problem to minimize the intra-cluster communication cost and further combine with the explicit entropy coding to minimize the inter-cluster communication cost. -- Furthermore, based on inherent characteristic of Slepian-Wolf coding, we propose a combined data aggregation and encryption scheme, called spatially selective encryption, for efficient and secure data transmission in WTSNs. Using this mechanism, as long as the data of the cluster head (a.k.a. the visual key) is properly protected, the data from all cluster members are secure. This novel approach can significantly reduce the energy consumption for data encryption. An energy-efficient key establishment protocol is also proposed to securely and efficiently establish the key used for encrypting the visual key. -- Finally, we propose a clustered on-demand multi-channel MAC protocol (COM-MAC) to support energy-efficient, high-throughput, and reliable data transmission in WMSNs. A scheduled multi-channel medium access is used within each cluster so that cluster members can operate in a contention-free manner in both time and frequency domains to avoid collision, idle listening and overhearing. A traffic-adaptive and QoS-aware scheduling algorithm is executed to maximize the network throughput. A spectrum-aware ARQ is further incorporated to better exploit the unused spectrum for a balance between reliability and retransmission