Published online by Cambridge University Press: 05 December 2014
Wireless sensor networks (WSNs) use the same medium (which is the air) for wireless transmission that the nodes in a wireless local area network use. In order for nodes in a local area network to communicate properly, standard access protocols like IEEE 802.11, IEEE 802.15.4, and ZigBee, are available. However, these and other protocols cannot be directly applied to the wireless sensor area networks. The major difference is that, unlike the devices participating in local area networks, sensors are equipped with a small source of power (usually a battery), which drains very quickly. In addition, sensor nodes have limited resources of memory and computational power. Hence, there is a need to design new protocols for medium access control (MAC). This chapter investigates some of the major MAC protocols available [1–7].
Medium access control in wireless networks
There exists a very good set of standard protocols for wired and wireless area networks, which are proven to operate efficiently. There are many reasons for not using the wired and traditional wireless MAC protocols for WSNs. The standard protocols in wired local area networks use the well-known carrier sense multiple access with collision detection (CSMA/CD) scheme for individual stations to access the medium. This protocol, the IEEE 802.3, is known as Ethernet [1–46]. In this scheme, each station senses a medium for a random amount of time. If no activity is detected it starts its own transmission. If it detects any activity in the medium it defers its transmission until the activity ceases. If it senses a collision of packets from different nodes in the networks it backs off a random exponential amount of time and then starts contending for transmission using the backoff exponential algorithm. Slotted ALOHA networks traditionally use a time division multiple access (TDMA) scheme [9–10]. In this scheme, the time is divided into an equal number of slots such that a node is allowed to transmit only in its allotted slot. Here, the disadvantage is that a node is allowed to transmit if and only if it owns the slot; otherwise it has to listen to the medium for the data intended for it. Therefore, if the node has no data to send the bandwidth is simply wasted because there will be no activity for the amount of slot time allotted to a node with no data.