Modeling and Optimization of some Network design problems under Uncertainty

Abstract

This research explores network design optimization problems, specially, wireless network newlinedesign and addresses important issues regarding energy efficiency, reliability, and network newlinelifetime in presence of uncertainty. Constrained optimization problems serve as the newlinefoundation, embodying the intricate interplay of system constraints, physical limitations newlineand functional requisites. The research employs various methodologies and algorithms to newlineoptimize wireless sensor networks (WSNs) lifetime under uncertain conditions, presenting newlineinnovative solutions to enhance their performance and longevity. newlineThe study commences by recognizing the complex nature of constrained optimization newlineproblems in network design and data distribution, particularly in scenarios marked by non newlinelinear and non-convex constraints. The third and fourth chapter of thesis focuses on newlinedistributed network design problem involving uncertainty in parallel routing and the newlinemaximization of connectivity time in WSNs under uncertain parameters. The next newlinecontigues chapter addresses the problem related with cluster formation, cluster head newlineselection under uncertain parameter and the impact of choosing bad CHs in network newlinelifetime. The optimal number of clusters and cluster reliability play critical roles in newlinedetermining the overall lifetime and performance of WSNs. Factors like energy efficiency, newlinecommunication overhead and network coverage directly depends on optimal clustering newlinewhereas data aggregation, fault tolerance and network lifetime are also related with cluster newlinereliability. In the last section of this study problems related with optimal cluster and cluster newlinereliability has been discussed. newlineThe third and fourth chapter addresses two distinct challenges in network design and newlineWSNs. The third chapter focuses on a distributed network design problem with newlineuncertainty, employing the Kuhn-Tucker (K-T) optimality algorithm for non-linear, convex newlineXX newlineAbstract newlineoptimization. Triangular fuzzy numbers are used to express uncertain traffic rates, newlinetransf