| Wireless Sensor Networks (WSNs) have promising and valuable applications. More and more researchers from military, industry and academic have conducted research in this area.MAC protocols are basic protocols in WSNs and they have more impacts on the performance of networking system. In this paper, we propose a WSNs application, which is deployed at crossroad for the safety of pedestrian. We analyze the application-aware, event-driven, and multimodality features of this application and design and implement a MAC protocol, called App-MAC. App-MAC supports prioritized event delivery, maintains fairness between events and sensor nodes, and has high channel utilization and efficient energy consumption. App-MAC leverages the advantages of contention-based and reservation-based MAC protocols to coordinate the channel access, and propose channel contention and reservation algorithms to adaptively allocate channel slots according to application requirements and current events status. To evaluate App-MAC, we have conducted simulations through TOSSIM simulator and empirical studies using Berkeley TelosB motes with target recognition events, and compared with three state-of-the-art MAC protocols, i.e., S-MAC, TDMA, and TRAMA, in terms of five proposed performance metrics, namely average event delivery latency, event fairness index, sensor fairness index, channel utilization efficiency, and energy consumption efficiency. We found that App-MAC outperforms other protocols tremendously, including decreasing the average event delivery latency from 58% to 84% (simulation) and 4% to 75% (empirical study), improving the channel utilization efficiency from 122% to 520% (simulation) and 13% to 58% (empirical study), while at the same time improving the energy consumption efficiency from 55% to 79% (simulation) and 46% to 59% (empirical study).Routing protocols in WSNs are application-oriented, and none of the protocols is suitable for all applications. Therefore, designing a routing protocols, which are application-aware, energy-efficient, and reliable is a hot research topic. Recent study in wireless sensor networks (WSN) has found that the link quality varies significantly with spatial and temporal factors and approximately 5% to 15% of all links are asymmetric links. These asymmetric links have heavy impacts on the performance of routing protocols. Furthermore, there are routing holes in WSNs, which also impair the reliability of routing protocols. We design energy-efficient and reliable routing protocols in order to solve the asymmetric links and routing holes problems.In this thesis, we design and implement the link layer service to discover neighborhood and estimate the timeliness link quality between neighbors using active probing, passive overhearing and WMEWMA. The link layer relay service presented in this thesis mitigates the effect of asymmetric links so as to discover more neighbors. We evaluate the algorithm using both static analysis and simulation through the TOSSIM simulator. Results show that the algorithm is effective and more than 40% of nodes identify more outbound neighbors and the percentage of increased outbound neighbors is between 14% and 100%. We develop a distributed algorithm to build the most reliable routing path for every node using the link layer services. From statistic analysis and simulation using TOSSIM, we found that the algorithm can prevent building a broken routing path and can build a better routing path, and more than 17% nodes have built more reliable routing path and the percentage of the improved reliability is about 2% to 51%. Motivated by realistic sensor network scenarios that have fading environment and opportunistic transmission, we propose a novel routing protocol, OMR, which integrates routing and MAC protocols to improve the packet delivery ratio, reduce the packet delivery latency, and decrease the energy consumption in multi-hop wireless sensor networks. OMR chooses a packet's route at each hop after the transmission to that hop and the choice is based on the fact that intermediate nodes actually receive the packet and can forward the packet to any nodes at the next hop with high probability. The nodes coordinate to guarantee hop-based reliability while avoiding redundant retransmission by blindly flooding. The timeout and retransmission mechanisms by intermediate nodes and sink in OMR guarantee the end-to-end reliability. Piggyback acknowledgement shares the packet and acknowledgement information among neighbors to mitigate the effects of lossy and asymmetric links. The results of intensive simulation using TOSSIM show that OMR performs well compared with traditional routing protocols.In this thesis, we describe the routing holes problem met by greedy geographical routing in wireless sensor network. An algorithm is proposed to identify the edge of the routing hole using right-hand rule. Another algorithm is also proposed to classify the edge nodes of the irregular routing hole and leverage horizontal scan to approximately calculate its area in a planar topology. The area of the routing hole can inform the design of routing protocols and the redeployment of the sensor nodes. To meet the application reqirements in WSNs and handle the routing hole problem, we focus on designing a general scalable, load balanced, fault-tolerant, and energy-efficient routing protocol, i.e., WEAR. The novelty of WEAR is to take into consideration four factors that affect the routing policy, namely the distance to the destination, the energy level of the sensor, the global location information, and the local holes information. All these factors are integrated into the notion of weight in the WEAR protocol. Furthermore, to efficiently handle routing holes, we propose a scalable, hole-size-oblivious holes identification and maintenance protocol. To evaluate the efficiency and effectiveness of the proposed protocol, we define eight general performance metrics for evaluating routing protocols in WSN. In terms of these metrics, our comprehensive simulation results show that, compared with GPSR, WEAR extends the lifetime of the sensor network about 15% longer, increases the reply successful delivery rate more than 7%. Compared with GEAR, WEAR also performs better in terms of almost all the performance metrics.
|
PDF Full Text Request