Description
In wireless communication environments, backoff is traditionally based on the IEEE binary exponential backoff (BEB). Using BEB results in a high delay in message transmission, collisions, and ultimately wasting the limited available bandwidth. As each node has to obtain medium access before transmitting a message, in dense networks, the collision probability in the medium access control (MAC) layer becomes very high, when a poor backoff algorithm is used. The logarithmic algorithm proposes some improvements to the backoff algorithms that aim to efficiently use the channel and to reduce collisions. The algorithm under study is based on changing the incremental behavior of the backoff value. The BEB is used by the local area networks (LANs) standards, IEEE 802.11, MAC. BEB uses a uniform random distribution to choose the backoff value; this often leads to reducing the effect of a window-sized increment. This paper carries out a deeper study and analysis of the logarithmic backoff algorithm that uses logarithmic increments, instead of an exponential extension of the window size to eliminate the degrading effect of random number distribution. Results from simulation experiments reveal that the algorithm subject under study achieves higher throughput and less packet loss, when in a mobile ad hoc environment.