Design and Optimization of MAC Layer Schemes for Maximizing Throughput in Wireless Networks

Abstract

In today's world, when communication has become a basic need of life, im- proved user experience and network throughput is crucial from the perspec- tive of both customers and Internet service providers. With the increasing applications of wireless communication in city man- agement, entertainment, social media, education, industry, and many more, the requirements of data rate have grown exponentially. The diversity of technologies and applications make it impossible to nd a single generalized solution. In our study, at rst, the performance of the medium access control layer (MAC) of wireless sensor network has been investigated for energy harvest- ing enabled nodes through the Markov model. Based on the ndings of the analysis and solar energy harvesting pro le, an adaptation of backo pa- rameters has been proposed to maximize average network throughput while meeting the application delay and reliability constraints throughout the day. The proposed scheme exhibits a 72 % improvement in network performance. Next, we explore an energy-e cient and environment-friendly communica- tion technology, visible light communication (VLC), in an indoor scenario. We begin with the modi cation of the MAC frame structure, which is a rel- atively less explored area of VLC MAC. A reservation-based frame structure (RBT) has been proposed to overcome the extra delay in contention-based iv access and to improve the average network throughput. The performance of the network under the proposed frame structure has been evaluated through a Markov model of the network. The result exhibits on an average 80% de- crease in delay and a 170 % increase in reliability. Further, we explore a hybrid VLC/RF network which is an in-practice deployment of indoor VLC networks to deal with the small coverage range of VLC access point (VLC- AP). To improve the network performance of the multi-VLC and RF AP network, we optimize slot allocation in both uplink and downlink directions. The optimization problem exploits more than one AP per user during a single frame period to maximize the average network throughput. This collective resource (slot) optimization strategy, improves the network throughput by more than 30% as compared to a single VLC-AP based scheme (SV O RF), in which a single VLC-AP is utilized for a user and RF-AP is utilized to o oad tra c when VLC resources are not su cient. Incorporating power allocation along with the slot allocation for downlink transmission exhibits more than 65% improvement in both fairness factor and average sum rate. In the last, we investigate another important class of wireless technolo- gies; molecular nano-networks. A multi-molecular gradient eld is generated and utilized for distinguishing hop distance instead of a complicated address- ing mechanism. Bacteria based transmission scheme is used for transferring information across single and multiple hops following the multi-molecular gradient eld. The proposed scheme provides a high success probability for larger information transfer in a relatively shorter time due to an increased thrust of message-carrying bacteria towards the sink node.