Research On Topology Control And Hierarchical Routing In Wireless Sensor Networks

In recent years, wireless sensor networks(WSNs) which have taken much attention at home and abroad play important part of the Internet of things. Due to resource con-straints to sensor nodes, such as energy, ability of computing and communication etc., the further development and practical applications of WSNs are seriously hindered. As the critical issues of network layer in WSNs, topology control and routing is utilized to build reduced network topology and data transmission path. With effective topology construc-tion mechanism, reliable topology maintenance strategies and efficient distributed dynamic hierarchical routing protocols, network connectivity and coverage can be guaranteed, while it helps to decrease redundancy of the nodes and flow of information. simultaneously, it reduces frequency degree and interference of communication, and promotes the network load balancing so that the energy consumption rate can be cut down and network lifetime will be prolonged.In this dissertation, it focuses on topology control and hierarchical routing in WSNs. As starting with a single method of topology construction, a topology maintenance strategy is integrated into complete topology control process to show how to construct and maintain the reduced network topology. simultaneously, dynamic distributed hierarchical routing is researched to build efficient data transfer path for both homogeneous and heterogeneous WSNs. Furthermore, topology control and hierarchical routing is applied in intelligent transportation systems in order to alleviate the current serious urban traffic conditions based theory research. The main work and results of this thesis are introduced in detail as follows:1. A question of finding optimal virtual backbone through fully connected network is abstracted and converted into minimum connected dominating set problem(MCDS). A new mathematical model based mixed integer programming(NMIP-MCDS) is proposed in order to solve MCDS which is called NP hard problem. Furthermore, with the help of an- alyzing MCDS solution, the product of token distribution number and energy consumption from nodes is built as the goal of optimization function. Simultaneously, optimal MCDS is constructed via manner of token distribution and network energy load balancing.2. A topology construction algorithm based spanning tree (RTCST) and reliable topol-ogy maintenance policy, which make up of complete topology control mechanism, are pro-posed for complicated network topology and reducing network energy consumption. As starting with the Sink node, RTCST algorithm lets network topology transpire surrounding and ultimately forms a spanning tree with the core of Sink through circularly executing three sub-procedures, which are composed of neighbor detection, children selection and self-redemption. Combined with static and dynamic techniques, factors of time and energy consumption are considered as the triggering criteria simultaneously. Furthermore, topol-ogy will be switched to others in static or dynamically called topology construction when current network topology is no longer optimal or some nodes have failed.3. An adaptive distributed clustering routing protocol (ADCR) is proposed for homo-geneous WSNs. ADCR determines the current optimal cluster number adaptively through dispersion of node distribution and curvature change based on Nth-order nearest-neighbor theory when position of sensor node or network topology is changed. Furthermore, with the residual energy of sensor nodes, the optimal proportion of cluster head in network is cal-culated to elect the cluster head set according to optimal cluster number. Simultaneously, Hausdorff distance is introduced to adjust the initial rapid formation clusters.4. An efficient and dynamic distributed clustering scheme(EDDCS) is proposed for routing problem in heterogeneous WSNs. In order to guide the process of cluster head election for a given multi-level heterogeneous network, EDDCS determines the probability of node to be a cluster head through average network residual energy estimation in next round by average energy consumption forecast in ideal state and reference value of his-torical energy consumption simultaneously. Furthermore, analogous universal gravitation is introduced to make non-cluster head node joining cluster in terms of gravitation during cluster formation process.Simulation results have proved the efficiency on research all of above. Compared with congeneric protocols or algorithm, they save more energy and efficiently improve load balancing so that network lifetime can be prolonged.According to theoretical research, topology control and hierarchical routing is tried to apply in intelligent transportation. Therefore, a feasible architecture for urban traffic mon-itoring and intelligent guidance based on environment in which three completely different types of network protocols work together is designed. Following as the discussion of soft-ware and hardware platform, it still has incorporated other new techniques, such as wireless multimedia sensor networks, ultra wide band communication, collaborative in-network data processing, etc.. With a special hybrid of topology control and routing mechanism, some part of prototype systems have been implemented in this thesis finally.