Research On High Energy-Efficient Performance Of MAC Protocols For Wireless Sensor Networks

Recently, wireless sensor network (WSN) has been paid much attention and it is very attractive. Its appearance has a deep influence, not only changes modern science and technology, but also alters people's mind. Compared with the traditional network technologies, WSN is a combination of modern wireless communication technology, micro-sensor technology and network technology. It provides numerous applications, i.e., national defense, environmental monitoring, home automation, and transportation. WSN consists of a large number of simple, inexpensive sensor nodes which could do some uncomplicated compute and wireless communication. However the energy of such sensor node is limited, and it cannot keep working for a long time. Once the energy of the sensor node is exhaustion, the structure of the WSN will brake up leading to a damage of the network topology. Therefore, how to improve the energy efficiency of sensor nodes is a vital issue for the design of WSN protocols.As known, one of the main services and functions of the medium access control (MAC) protocol is allocating limited wireless communication resources to the sensor nodes meanwhile deciding the working mode of the node's radio transceiver, which is used to promise effectively energy consumption of sensor nodes in WSN. An energy-efficient MAC protocol of WSN can convince the requirement of efficient data communication under low-power consumption condition. This paper regards energy efficiency as the optimization target, and presents some reasonable and effective algorithms to improve the energy-efficient performance of WSN. And this paper focuses on the following three topices, they are: how to adjust the tradeoff between energy consumption and time delay for MAC protocols, how to extend the lifetime of WSN, and how to decrease the energy consumption for data packets transmission.Firstly, we discussed the optimization objective of WSN MAC protocol designs based on the basic concept of MAC protocol, taking fully consider of the differences of MAC protocols between the traditional network and WSN, and the specific and novel restrictions for MAC protocol designs. Then we studied the typical MAC protocols of WSN, called slotted access MAC protocols, and demonstrated that there was a significant gap between the optimal performance and current energy efficiency algorithm. For example, the energy-efficient of current protocol are fixed duty cycle which brings the difficulty of flexible adjusting between energy consumption and time delay; the boundary nodes in virtual cluster decrease the netowrk lifetime significantly; the extra energy consumed by frequently switching of the node's radio transceiver leading to increase the data tansmission overhead.Secondly, concentrate on the issue of adjusting the tradeoff between energy consumption and time delay, unlike the fixed duty cycle, we provided a dynamic duty cycle protocol called queue delay sensor MAC(QDSMAC) protocol, which exploited the literation of the network traffic load as a parameter. To determine the changing steps of the network traffic load, the queue delay of the node buffer and the remain energy are combined and then designed as the algorithm of the average delay in buffer queues. According to the information of the changing steps of the network traffic load, the proposed QDSMAC protocal adjusts the duty cycle immediately to achieve the optimum tradeoff between energy consumption and time delay. By broadcasting control packets priority, QDSMAC protocal is able to avoid the data packets loss caused by changing duty cycle.Thirdly, we discuss the issue of the boundary node energy consumption in the cluster-shaped synchronization structure, and provided a virtual cluster aggregation (VCA) algorithm. As the boundary nodes at the junction among multiple virtual clusters have to keep the listening time longer enough to ensure the configuration of the network; thus they consume more energy than the ordinary nodes, which will generate blocking and destroy the topology of the network.The proposed VCA algorithm can calculates the total number of nodes in a virtual cluster by a style of tree structure. Since each senor node knows the size of own virtual cluster, when the adjacent cluster's size changes, the nodes start to aggregate other virtual clusters the smaller size as that of other virtual cluster. The proposed VCA algorithm reduces the number of boundary nodes effectively and extends the network timelife significantly.Finally, an energy-efficient data transimission algorithm called fragment identifier message passing(FIMP), which is used to decrease the extra energy caused by frequently state swicthing during data packets transmission. By using data packets fragments transmission, it is possible to avoid extra energy consumed by long data packets retransmission. Generally speaking, both the state of the node's radio transceiver frequency switching and the control packets transmission comsume extra energy. The proposed FIMP algorithm firstly confirms many data packet fragments by an acknowledgment which includes the information of data packets fragment identifier; and data packets recombinant with the data packets fragment identifier is able to avoid duplicate transmission and reduce the energy consumption per data packet.