Abstract
Wireless sensor networks (WSNs) have been widely used for surveillance in harsh environments. Each sensing node has to distribute its sensing data to all other nodes for ubiquitous data collection. The amount of battery energy consumed to transmit a packet to other nodes directly affects sensors’ lifetime and thus is considered as a critical issue in WSNs. In this thesis, we propose a scheduling network coding protocol to broadcast data more efficiently than previous network coding schemes. In our protocol, first, each node collects its neighbors’ information in two hops, and then it determines a local optimal encoding packet transmitting policy. According to this policy, each node decides whether to send out the encoding packet or do nothing to reduce the redundant transmissions when receiving a novel packet at the certain time. Our protocol can be applied to any broadcasting scenario, such as single-source broadcasting, multi-source broadcasting or all-to-all broadcasting. The simulation results show that our protocol not only reduces the number of broadcastings significantly compared to previous works, but also promises that each node in the network can recover all sensing data with a high probability.