Abstract
In case of high dynamic topology, reactive routing protocols provide quick convergence by faster route discoveries and route maintenance. Frequent broadcasts reduce routing efficiency in terms of broadcast cost; B-k, and expected time cost; E[t]. These costs are optimized using different mechanisms. So, we select three reactive routing protocols; Ad-hoc On-demand Distance Vector (AODV), Dynamic Source Routing (DSR), and DYnamic Manet On-demad (DYMO). We model Expanding Ring Search (ERS); an optimization mechanism in the selected protocols to reduce B-k and E[t]. A novel contribution of this work is enhancement of default ERS in the protocols to optimize B-k and E[t]. Using NS-2, we evaluate and compare default-ERS used by these protocols; AODV-ERS1, DSR-ERS1 and DYMO-ERS1 with enhanced-ERS; AODV-ERS2, DSR-ERS2 and DYMO-ERS2. From modeling and analytical comparison, we deduce that by adjusting Time-To-Live (T T L) value of a network, efficient optimizations of B-k and E[t] can be achieved.