|Title||Quality of Service Aware Source Initiated Ad-Hoc Routing|
|Publication Type||Master's thesis|
|Year of Publication||2004|
|Publisher||Electrical Engineering and Computer Science Department, Washington State University, WA, USA|
Mobil Ad-hoc networks (MANETs) are used in a variety of situations ranging from conference meetings to military operations. Among the important challenges facing MANETs is the development of suitable routing protocols. A Routing Protocol must be aware of the diversity in the network if the network is going to perform at its best. The routing protocol should be e\pmcient and able to adapt to rapidly changing topologies. A MANET is applicable to any situation requiring network communication in particular short-lived highly mobile networks. The routing protocol needs mechanisms that handle the mobility such that the communication disruption time is min- imized. Speci¯cally we desire a routing protocol with quality control mechanisms for every important stage in the lifetime of a route. This thesis proposes a Quality of Service (QoS) aware source initiated ad-hoc routing protocol (QuaSAR) that adds quality control to all the phases of an on-demand routing protocol. QuaSAR gathers statistical information from the network during route discovery, more speci¯cally battery power, signal strength, bandwidth and latency. In particular we use the signal strength to ¯nd stronger connected routes. The metrics are associated to the individual route and later used when a route to the destination is picked. QuaSAR includes proactive route maintenance features in addition to the reactive maintenance. The proactive mechanism of QuaSAR makes a Mobile Node (MN) on an active route aware of route critical incidents in progress and noti¯es the appropriate party of the development. Performance experiments were conducted to test the performance of QuaSAR using ns- 2 Network Simulator. Results were compared with Dynamic Source Routing (DSR). The following mobility models were used in the experiments: Random Waypoint, Gauss Markov, Manhattan Grid and Reference Point Group Model. We have demonstrated that using signal strength and proactive mechanisms as a means of routing will increase the throughput and the packet delivery ratio. The quality control mechanisms of QuaSAR enhance the performance experienced by MNs by giving it more stable paths thus minimizing the communication disruption time.