| Recent advances in micro-electro-mechanical systems (MEMS) technology, wireless communications, and digital electronics have enabled the development of low-cost, low-power, multifunctional sensor nodes that are small in size and communicate untethered in short distances. These tiny sensor nodes, which consist of sensing, data processing, and communicating components, leverage the idea of Wireless Sensor Networks (WSN) based on collaborative effort of a large number of nodes.A Wireless Sensor Network does not require any pre-existing infrastructure or configuration but is formed spontaneously by sensor nodes that wish to communicate. Each node in the network acts as a router and forwards packets on behalf of other nodes, allowing nodes that are not within wireless range of each other to communicate over "multi-hop" paths. In contrast to single-source single-destination unicast schemes, group communication schemes become more popular in WSN applications than ever, while multicasting aiming at identifying one-to-many transmission paths is the most possible way to it.My thesis in this dissertation is focused on light-weight, distributed and on-demand fashion multicast routing techniques for WSN.According to the features and application demands, the thesis proposes LDMRP, a distributed, light-weight multicast protocol for Wireless Sensor Networks. Based on the shortest-path algorithm, LDMRP utilizes the shared sub-paths between the source and each receiver to construct a multicast group so as to reduce the redundant messaging and increase the bandwidth efficiency as well as the energy efficiency. The simulation results in NS-2 prove that LDMRP performs better than those on-demand multicast protocols designed for Ad Hoc network which are also workable in WSN.Inspired by the achievements of LDMRP, the thesis presents an optimized distributed Steiner-like multicast routing protocol, DstMRP. Given a set of receiver nodes, the transmitting node first groups them into clusters according to their locations and then splits the WSN into several sections regarding the clusters, during calculating the next hop for each cluster, the intermediate node will employ a constrained-flooding schema regarding whether it belongs to one of the sections or not. Thus, the control packet exploring the multicast routing paths will carry moreinformation of network topology before it arrives at the receivers, which is called after delayed-effect. The extensive simulation of DstMRP shows that it compares better against LDMRP which utilizes a simple flooding schema during the multicast path exploration.As the mobile object tracking scenarios are concerned, LDMRP and DstMRP will construct multicast trees frequently which leads to high control overhead and low energy efficiency. The thesis offers a convey-tree mechanism to prolong the lifetime of each multicast tree and reduce the total numbers of multicast trees during the tracking task. The convey-tree is built up in a limited area near the object and a convey-header is elected to whom the nodes of the tree send their generations. Also the header will initialize a LDMRP or DstMRP multicast tree rooted at itself and re-transmit all the packets from the convey tree to receivers. The thesis also provides an optimized reconstruction rule for convey-trees intended to the minimal energy consuming.In the thesis, an efficient detection of broken links and localized reparation mechanism is introduced, which utilizes the characteristic of wireless transmission adequately. With very little control overhead, it enhances the reliability and robustness of LDMRP and DstMRP as well as the performance.
|
PDF Full Text Request