Research On Energy Efficient S-MAC Protocol For WSN

Wireless Sensor Nodes are energy-constrained and always deployed in special andvast areas. Therefore, saving power is one vital consideration when researchers designany Wireless Sensor Network applications. The MAC protocol decides nodes how toaccess the wireless channel and directly control the communication module whichconsumes the most energy. So designing an energy efficient MAC protocol is onesubject of Wireless Sensor Networks.This thesis describes MAC protocol for Wireless Sensor Network, includingdesign goals and key technologies. S-MAC is a classic contention based MAC protocolfor Wireless Sensor Network, so this thesis focuses on it deeply. By using S-MACprotocol, nodes of WSN can only access the wireless channel in listening period of aschedule. This fixed way may cause data delay and high conflict probability. To solvethis problem, this thesis split the original schedule period of S-MAC into multiple smallschedule periods which are called micro duties. Wireless Sensor Nodes set the amountof micro duties they have and in which they will used to work by calculating theirresidual energy. In this way, nodes will be distributed into different micro duties anduse their own micro duties to compete channel for communication. Those nodes whichwork in the same micro duty constitute the micro clusters. This decentralized way willreduce the wireless channel competitiveness as well as collision probability amongnodes. Furthermore, nodes can communicate with each other in different micro duties,which is similar to TDMA, will also reduce delay.The traffic of Wireless Sensor Network is dynamic, but the duty cycle of S-MACis fixed, which cannot adapt to the dynamic traffic very well. To address this issue, thisthesis proposes an adaptive duty cycle based on micro duty. Nodes can change theirduty cycle of micro duty by calculating the length of their buffer queue and double theduty cycle of micro duty if the buffer queue length is above a threshold. In this way, theadaptive duty cycle will effectively reduce data transmission delay. Different from otheradaptive duty cycle MAC protocol in which nodes still work in a centralized way, thedecentralized way of micro duty will reduce collision probability as they are distributedin different micro duties. This thesis also proposes two adaptive backoff algorithms: F-BEB algorithm and CA-BEB algorithm. Nodes use these two algorithms based on dutycycle and the traffic in different situation.By introducing micro duty, adaptive duty cycle of micro duty and adaptive backoff algorithm into S-MAC, this thesis proposes an improved S-MAC protocolwhich is called MDA-SMAC. This thesis also simulates this two MAC protocol byusing NS2in different topologies. As MDA-SMAC uses micro duty and adaptivemechanism, the simulation results show that it outweighs S-MAC in differentperformance index such as average energy consumption, average delay and effectivethroughput, especially when the traffic flow is high.