Throughput enhancement of MANETs: Packet fragmentation with hidden stations and BERs

Hidden nodes cause unexpected collisions and interferences in mobile ad-hoc networks (MANETs). However, little attention has been given to the behavior of hidden nodes, especially when channels are under random bit errors. Exchanging RTS/CTS can avoid hidden node problems. However, undesirable overheads in every packet cause performance degradation, and this motivates many researchers to design new algorithms with modification of the IEEE 802.11 standard. In this paper, the impact of hidden nodes is thoroughly modeled using Markov chains. This model incorporates critical network parameters (random bit errors, packet lengths, number of users) in noisy MANETs channels. Through rigorous analysis, the study suggests the optimal packet lengths to be used in various channel conditions with the existence of hidden nodes. These optimal fragments effectively reduce the number of retransmissions caused by collisions from hidden nodes and corrupted packets due to random bit errors, which significantly improves throughput in noisy MANETs channels without modification of the standards.