No abstract

Improving the performance of Mobile Ad hoc NET-works (MANETs) is challenging. The MANETs communications are presented with many constraints such as dynamic topology, unreliability of wireless links, and limited bandwidth and energy resources. The routing in MANETs can be particularly difficult due to aggravated congestion problem. This paper proposes a new QoS (Quality of Service) routing protocol combined with the flow control mechanism. This routing protocol selects the routes having more resources in an intelligent manner and does not rely on diffusion. The proposed routing protocol uses a new metric to find the route with higher transmission rate, less latency and better stability. In order to reinforce the congestion avoidance, a flow control mechanism is employed to adapt the transmission rate for each route. The proposed routing solution is modeled by ant systems. The numerical results are obtained using the Network Simulator (NS2.31). These results confirm that the joint consideration of flow control and routing can significantly improve the network performance compared to other QoS routing protocols such as the QoS-AODV (Ad hoc On Demand Vector with QoS) protocol.

Mobile ad hoc networks consist of freely moving nodes responsible of not only forwarding packets for other nodes but can also perform extensive computations. One of the most critical issues in these networks is the significant differences in term of processing and energy capacity between the nodes, inducing a load imbalance. Thus, sharing the load between the overloaded and idle nodes is a necessity in ad hoc networks. In this paper, we present a new load balancing algorithm based on clustering where a subset of nodes 'clusterheads' is elected to maintain some balance within their respective clusters while minimizing the overall communication cost. Our primary goal is to minimize the total execution time of the tasks by distributing the workload among nodes. Another goal is to extend the overloaded nodes lifetime inducing a stability of the network. The simulation results have shown that network performance can be reached by distributing load to idle nodes within the network.