Opportunistic Routing In Mobile Wireless Sensor Networks

As one of the important technologies in the 21st century, Wireless Sensor Networks (WSNs) have attracted a great of concerns and are widely used. Among them, the wildlife monitoring application should be the hottest one. Wildlife monitoring application network belongs to the Mobile Wireless Sensor Network (MWSN), and its routing technology faces with a lot of challenges.As compared with traditional wireless routing algorithms, Opportunistic Routing (OR) takes full advantage of the broadcasting nature of wireless channel, as well as the time-varying characteristic of wireless channel quality. Thus, OR has improved network performance greatly. As a new post-determined routing, Opportunistic Routing has drawn a wide range of concerns and become one of the hottest researches in MANET.However, Opportunistic Routing is still facing many urgent problems, such as forwarding-candidate selection, spatial reuse, and so on. To solve these problems, this thesis mainly focuses on the designing of Opportunistic Routing for Mobile Wireless Sensor Networks.To solve the forwarding-candidate selection problem, the thesis has proposed an opportunistic routing based on RSSI (Received Signal Strength Indicator), named OR-RSSI, for mobile wireless sensor networks. In the proposed protocol, Opportunistic Probability (OP) based on the RSSI of the Sink's Beacon packets and Mobility Vector (MV) is established and the best node with the highest OP available at that instant is used to store and relay packets at each hop after packets are broadcasted.The idea of OR-RSSI comes from behavior of honey bees. Different from existing Opportunistic Routing algorithms, OR-RSSI does not need the support of expensive devices or consume a great of energy for interchanging routing packets. What is more, OR-RSSI introduces the mobility vector mv to reflect nodes'mobility.We have implemented OR-RSSI in NS2 and done a number of simulation experiments. Simulation results show that OR-RSSI is more feasible for sparse MWSN and performs better on Successful Delivery Ratio, Goodput and Energy Overhead as compared with traditional TinyAODV.To solve the problem of spatial reuse and redundant copy packets, researchers have proposed Opportunistic Routing based Network Coding (NC), named OR-NC. However, these algorithms all adopt the way of Full Network Coding, which makes the destination node wait to decode the original packets until K linear independent coded packets have arrived. Therefore, this situation results in an increase of the packet average delay and makes the destination node uncomfortable to deal with the packets.So in the thesis, we have proposed an Opportunistic Routing based on Partial Network Coding, named OR-PNC, trying to lower the packet average delay. In OR-PNC, forwarding candidate nodes randomly choose received packets to encode new packet with random length and the destination node can decode original packets randomly. We have implemented OR-PNC in Eclipse 3.0 and done practical experiments with two wireless nodes, which are equiped with 802.11g wireless network cards. Experiments results show that: with different values of parameter K, the packet average delay of OR-PNC is commonly smaller than that of OR-NC. When K is 6, the average delay can decrease by 26.489%. However, with increasing of value of K, the improvement of OR-PNC will slow down, or even worse than OR-NC.In summary, we have proposed two opportunistic routing, OR-RSSI and OR-NC, for mobile wireless sensor networks. Moreover, some in-dept experiments, including simulation experiments and practical experiments, have been done for Opportunistic Routing. Experiment results are valuable.