On-demand Multipath Routing Protocol In Mobile Ad Hoc Networks

Mobile ad hoc networks(MANETs)comprise a collection of wireless mobile nodes that can communicate with each other without any centralized architecture.Such characteristics allow people to build a MANET seamlessly in real time applications such as military and disaster recovery.Due to the limited communication range,the nodes exploit multiple hops to exchange information with any other nodes in the network.On-demand(reactive)routing protocols are very suitable for MANETs,as they discover a route whenever it is needed.Designing reactive routing protocols in a MANET is a complex task due to the constrained battery and bandwidth.Therefore,it is desirable to study efficient routing techniques to guarantee reliable communication promptly.To achieve this objective,this research work proposes four different reactive routing protocols,namely Mobility ADaptive AODV(MAD-AODV),Improved AOMDV(IAOMDV),Path discovery by Energy Avoiding redundant rebroadcast and Routing Load AODMV(PEARL-AOMDV),and Network Dynamics Oriented Ad Hoc On-Demand Multipath Distance Vector(NDO-AOMDV).The key contributions are summarized as follows.The MAD-AODV aims to reduce the delay in data delivery caused by frequent link failures.The MAD-AODV follows the basic AODV process,but improves the route maintenance phase of AODV so as to adapt to the dynamic node mobility.The MAD-AODV predicts the link failure based on signal strength.It repairs the link failure locally by discovering the alternate link,thus it reduces the reroute discovery time.The MAD-AODV measures the distance from itself to previous and next hop node using received signal strength of the Hello packet for alternate link establishment.The simulation results show that the MAD-AODV attains enhanced performance in terms of packet delivery ratio and throughput,along with a decrease of end-to-end delay,packet loss and communication overhead,compared with the traditional AODV.The second work proposes an IAOMDV routing protocol to reduce congestion and enhance the data delivery reliability in MANETs.The IAOMDV integrates two components which are queue length based router selection mechanism and Frequent Link Failure(FLF)measurement.The queue length based router selection mechanism selects a less congested node to disjoint routes from the source to the destination.The IAOMDV improves the routing performance by distributing the data packet to the destination through multiple disjoint paths.The Frequent Link Failure(FLF)measurement predicts the node mobility and informs the source node before the occurrence of link breakage to select next available routes for further data transmission.Moreover,the IAOMDV routing protocol achieves better load balancing and also reduces mobility impact on routing.The simulation results depict that the IAOMDV improves the Packet Delivery Ratio(PDR)and reduces the data latency.The PEARL-AOMDV routing protocols incorporate efficient route discovery packet rebroadcast scheme and efficient path selection scheme to reduce the redundant transmission and improve the routing reliability respectively.The PEARL-AODV significantly decreases the number of rebroadcasts of the RREQ packet by measuring the uncovered neighbor density.The efficient path selection scheme selects the routers based on three factors:hop count,routing load,and residual energy.Thus,it avoids routing of packets through nodes with low residual energy and less congestion.The simulation results show that the PEARL-AMODV attains less overhead and enhances routing performance compared to the traditional AOMDV.The NDO-AOMDV aims to achieve reliable data delivery using the data caching mechanism and optimal path selection algorithm.The caching mechanism enables the routers to cache data until the next-hop router successfully forwards it.When the transmission fails,the packet sender re-transmits the cached data to enhance the packet delivery.The NDO-AOMDV takes a routing decision based on the traffic flow involvements of each node and switches over to an available better path during packet loss or path overload.This process reduces the data delay and maintains high packet delivery ratio.The NS2 based simulation results show that the performance of the proposed method is superior to AOMDV in terms of packet delivery ratio and throughput.