Research And Design Of Energy Efficient Stochastic Routing In Wireless Sensor Networks

Aiming at the problems of the unreliability of wireless links and the limited energy of sensor nodes in wireless sensor networks,the energy efficient stochastic routing protocols are studied and designed in this thesis.In order to reduce the impact of external factors such as the environment on the wireless link quality,opportunistic routing is used to improve the reliability of data transmission.In addition,the energy consumption of the network is reduced by the energy saving technologies such as power control and sleep scheduling.However,utilizing these energy saving technologies and the random data forwarding mechanism synthetically causes some difficulties in the design of routing protocols.The adjustment of the transmitting power not only changes the energy consumption of sensor nodes,but also has great influence on the quality of the wireless links and the selection of the candidate nodes in opportunistic routing.Meanwhile,there is less communication and extra waiting delay between any two nodes with asynchronous sleep scheduling,which makes the routing decision in the data forwarding process much critical.To solve these problems,this thesis mainly uses the the theories of Markov chain and Markov decision process(MDP)to model the processes of random data forwarding and routing decision.Then an opportunistic routing based on power control and two random forwarding mechanisms with asynchronous sleep scheduling are designed respectively,and they can improve the transmission efficiency of packets and reduce the energy consumption effectively.The main contents of this thesis are organized as follows:(1)The first chapter introduces the characteristics of wireless sensor networks and several energy saving technologies,analyzes the design problems faced by the energy efficient stochastic routing protocols and the research status,and then presents the main research contents of this thesis.(2)The second chapter studies the design problem of opportunistic routing based on power control.Markov chain is used to model the process of data forwarding in opportunistic routing,and then we study the influence of the transmitting power on the routing metrics and network performance.A distributed algorithm of candidate node selection is used to determine the optimal transmitting power and the corresponding candidate forwarder set for each node,and an opportunistic routing protocol based on power control is designed.Finally,the experimental simulations show that the designed protocol can reduce theenergy consumption of sensor nodes and prolong the network lifetime effectively.(3)The third chapter analyzes the nodes' wake-up process and data forwarding process in detail,and designs two kinds of random forwarding mechanisms with asynchronous sleep scheduling.Based on the idea of opportunistic routing,the first packet forwarding strategy uses the expected end-to-end delay to optimize the candidate forwarder set,then selects the first neighboring node that wakes up among multiple candidate nodes to forward data packets.By contrast,the second kind of forwarding strategy uses the MDP theory to model the process of single hop data forwarding,where the sender decides whether transmit to the current node directly or wait for other nodes to wake up with a certain probability.Finally,an optimal stochastic forwarding strategy is obtained by solving the finite horizon MDP problem,and simulation results show that these two kinds of random forwarding mechanisms can effectively reduce network delay and network energy consumption in wireless sensor networks with asynchronous sleep scheduling,especially the latter strategy can guarantee optimal performance in the low duty cycle networks.(4)In the fourth chapter,we summarize the main research results,and propose some future research problems.