Reserch On Algorithm Of MAC Protocol In Wireless Sensor Networks

Wireless sensor networks is a emerging research field integrates a variety ofdisciplines . Sensor nodes use batteries for power supply, so they are power limited. Thenthe ultimate goal in the research of wireless sensor networks is to prolong the networklifetime. Media access control protocol is responsible for channel resource allocation andsharing communication medium. It is in the underlying part of network protocols stackand plays an important role in the network performance. The MAC protocol in wirelesssensor networks are classified and compared in the paper. With regard to the S-MACback-off algorithm and related mechanism, a new improved algorithm is proposed.First of all, a simple Markov chain model is quoted to analysis the back-offmechanism in S-MAC protocol. Based on the two parameters, back-off hexadecimalnumber and the probability of conflict, a new back-off algorithm is proposed. Thesimulation results show that the new algorithm increases the success probability ofsending data packets, thus the throughput is improved and the energy consumption isreduced.Secondly, the duty ratio is fixed in the S-MAC protocol. When the network trafficbecomes busy, it can cause great delay because of not sending the queued packetspromptly. When the network traffic becomes busy, the long listen-sleep cycle is dividedinto several short cycles, thus the duty cycle is increased and there are more packetstransmitted during one long listen-sleep cycle. The adaptive listening scheme is alsoadopted, one packet can be transmitted further to two hops in one basic cycle, so thelatency is reduced. Simulation results show that the improved algorithm can obtain lowerlatency and reduce energy consumption.Finally, in order to extend the network lifetime, a back-off algorithm based on loadbalance is proposed. The new algorithm improved the competitive back-off mechanism ofS-MAC protocol. The new algorithm takes the node’s current energy into consideration toselect the deferring time adaptively. The nodes that have more current energy can accessthe channel with priority. The nodes with less current energy only need to send the data sensed by themselves which avoids the task of packet delivery. Simulation results showthat the network load is balanced and the network lifetime is prolonged.