Abstract:
In WBAN, large numbers of sensors are deployed inside and outside of human body. People with the chronic disease require long-term hospitalization. To avoid from longerm hospitalization. WBAN is the ideal way. Main objective which leads to design of IEEE 802.15.4 is to offer low data rate, little cost and Jess power protocol for communication in a sensor network. In IEEE 802.15.4, two basic techniques' are used to access the medium or channel. These techniques are named as slotted ALOHA and CSMA/CA. Slotted Aloha and CSMA/CA based techniques provides benefits of low computations and low overheads, but suffers collisions. The reason is blind transmission strategy in the beginning of time slots recently. Typically, WBAN nodes have small size, low energy capacity, low storage capacity and low processing capabilities of data. On the other hand, energy efficiency is still from key issues in WBAN. Machine learning has been functional effectively in wireless sensor networks to improve their perfonnance. In this document. I propose Adaptive wake up intervaL and access of channel based on Learning to increase the through put, reduce the energy consumption, and Jess delay under different scenarios, and convergence speed. These parameters are analysed for many different starting values of initial wake up. Learning Based Self Adaptive (LBSA MAC) based wake up intervals helps to minimize the unnecessary energy waste dynamic traffic in WBAK.Extensive simulations done in Ns-2 shows that LBSA-MAC performed better than the A Heuristic-based self-adaptive MAC for Resource constrained WBAN, simply IEEE 802.15.4, and TAP-MAC.
