Research On Fuzzy Sighted Routing Algorithms In Manet

Mobile Ad-hoc network, known as MANET, originated from tactical communication in military in the 1970s. As a network that relies on no fixed infrastructure, self-organizes anywhere anytime, and works in a distributed manner, MANET sees a wide application prospect in civilian life such as disaster recovery, conferences, etc., due to the effort made by research organizations such as IETF MANET WG, and the support of mobile electronic devices with growing performance.With its multi-hop nature, as well as its dynamic topology and limited bandwidth, MANET requires a dedicated design for its routing protocols that differ from those used in wired networks. However, the growing trend of large size and frequently-changing topology confronts MANET with more extreme challenge in its routing protocol design. Routing protocols should be adaptive with respect to both network size and topology changes.Proactive, link-state routing protocols are the main focus of this dissertation, with a common feature of some of which generalized as follows: making use of"fuzzy sight technique", in which control message dissemination is restricted in time and/or space, in order to reduce routing overhead, thus allowing larger network and more dynamic topology.A discussion of the feasibility and applicability of fuzzy sight technique in MANET is consequently carried out, followed by the introduction of fuzzy-sighted routing algorithms, or FSRAs, with emphasis on flat FSRAs, which fall into two categories: DV-like flat FSRAs and LS-like flat FSRAs, namely, flat FSRAs that adopt the control message dissemination approaches used in traditional Distance Vector Algorithm (DVA) and Link State Algorithm (LSA), respectively. The advantage of LS-like flat FSRAs over DV-like flat FSRAs is discussed subsequently.As a consequence, the optimal"fuzzy manner"among the LS-like flat FSRAs is presented, based on the fact that it brings the least routing overhead when topology is frequently changing, and is adaptive to different network size. The routing algorithm, with this optimal fuzzy manner, is then improved to be more robust and efficient under low mobility circumstances, and to be self-adaptive with respect to topology changes.The improved algorithm can be implemented in a simple network with few nodes and few topology changes, as well as one with a large amount of nodes and a wide range of topology changes. Therefore, the algorithm meets the demand of MANET with growing network size and topology changes. Meanwhile, the analysis of this algorithm serves as a theoretical basis of its implementation and further research.