NOVEL ALGORITHMS FOR IMPROVING LIFETIME AND TACKLING LINK RUPTURE IN WIRELESS NETWORKS

Abstract

A flexible data communication system is commonly referred to as wireless network which does not make use of wires to transmit data, rather uses air as medium to transmit and receive data. Some applications of the wireless networks are sensor networks, smart homes, cellular networks, telemedicine and MANETs. Wireless sensor networks (WSN) are related to much imprecision and various discrepancies that are inborn to it. Despite of various challenges, WSN finds wide applicability in environmental monitoring, agricultural field, industrial usage, and military applications and so on. Sensors have limited energy, as they are battery driven and are of small size. WSN, when deployed in hostile environments, it is impractical to change batteries. So, the proper consumption of energy plays a major role in deciding performance of WSN. Wireless ad hoc network is a collection of nodes that act like a host as well as a router. Nodes move randomly and organize themselves arbitrarily. As a result the network topology changes rapidly and unpredictably. Communication among nodes can be point-to-point or multi-hop. Point-to-Point communication is possible when points are within the radio range of each other. However, in multi-hop communication a packet reaches the destination through multiple number of intermediate nodes, in this case they act as relay nodes. These relay nodes transmit their own traffic as well as traffic from other nodes. In this thesis, a Two -Step Uniform Clustering (TSUC) algorithm is proposed, where clustering is done in two phases. In first phase standard clustering is done and in second phase isolated nodes are clustered if they are in communicating range of each other. The aim is to provide connectivity to the nodes in every part of the network. TSUC algorithm increases networks lifetime and throughput by re-clustering isolated nodes. Results obtained after simulation showed that the TSUC algorithm performed better than the other existing clustering algorithms. In the stationary sink setup of WSN, the nodes which are situated near to the sink are more prone to drain their energy fast and become dead due to concentration of data towards sink. So, we proposed optimal rendezvous points for mobile sink (ORP-MS) to address stationary sink originated problem. The proposed scheme incorporates mobile sink to the above proposed clustering algorithm (TSUC) and the rendezvous points for mobile sink are evaluated for each round. This further decreases the consumption of energy by cluster head and isolated nodes during communication with sink. During last decades, WSN has become an important base set-up for the Internet of Things (IoT). IoT has exponentially escalated the number of interconnected entities around the world with instant connectivity. Although, framing an efficient model for integrating WSNs into IoT is an exciting issue due to scalability and interoperability of IoT. Finally, the proposed TSUC algorithm is integrated with IoT for the prevention of forest fire by using two-layer architecture of sensor network assisted by IoT enabled Unmanned Aerial Vehicles (UAVs). A mobile ad hoc network (MANET), commonly referred to as wireless ad hoc network, as i discussed above, it is a constantly changing, scattered and dynamic network of mobile devices connected via wireless links. Node moves randomly in dynamic topology of the network. Relative mobility of the nodes with respect to each other, in the network is the pivot in influencing efficient execution of MANET. Consequently, mobility along with its perceptible parameters like velocity, angle of movement, stop time, etc., plays a dominant role in deciding performance of the network. In this thesis, we proposed Modified Mobility Factor to withstand with relative speeds of the communication entities and their current separation to escalate the lifetime of the link. Finally, to minimize the effect of constant random movement of the nodes, a modified routing protocol using cuckoo search algorithm (CSA) has been proposed and represented as (MAODVCS). This facilitates selecting healthy, minimally congested and shortest route for MANET. For achieving this, we defined link value index (LVI) and link quality index (LQI), which enable selection of routes having nodes with high data delivery rates. In addition, fitness function is defined that considers link’s features as well as node’s features (primarily residual energy) to optimize the route using cuckoo search and levy flight algorithm. The optimized routes obtained are further repaired using modified leap frog algorithm.