In Mobile Ad hoc Network, path between source and destination node changes too frequently due to unpredictable behavior and movement of mobile nodes. The data delivery to the intended destination becomes very challenging. The paths exist between source and destination node may be various types. Data delivery may be done with single or multiple paths. Single path sometimes not guaranteed about data delivery, so one of the better solution is multipath data delivery. Here, in this paper, we have considered Link Disjoint and Node Disjoint multipath for data delivery. For this AOMDV protocol with node and link disjoint is considered to evaluate performance. To evaluate their performance different node pause time considered with varying number of nodes. We have computed various QoS network metrics like throughput, average end-to-end delay, routing overhead to identify in which method of data delivery perform better and in what conditions.  The results obtain shows that Node Disjoint multipath method AOMDV performs better than Link Disjoint method. The simulation work carried out using Simulator NS-2.34.

In Mobile Ad Hoc Network, data delivery is very challenging through single path due to dynamic changes in the network topology. To cope up this issue multipath data delivery is very useful. Recently, many works have been carried out in this domain but few are addressing the queueing effect on multipath scenarios. In this paper, we have designed a network model that based on the existence of multipath between source and destination node and every node behave as M/M/1 queue. In order to do this we generate K (K=1, 2, 3…i) paths are available between each source toward the destination node. The traffic arrivals in each node follow poisson process with arrival rate λ packets/sec. The simulation work of this multipath scenario based on varying mean inter-arrival time. The effect of arrival rate on the performance of multipath network model is analysed and compared. To better understand the effect of arrival rate in application and network layer various QoS metrics are computed. Significant performance of individual node is noticed in the obtained results with various arrival rates.