Scalable location management for geographic routing in mobile ad hoc networks
Philip, Sumesh J.
MetadataShow full item record
While many solutions have been proposed for routing in mobile ad hoc networks, only a few have considered the issue of scalability of these protocols in networks having node membership in the order of thousands spread over a large area. Geographic routing using source-destination locations has been widely suggested as a scalable alternative to conventional routing approaches in mobile ad hoc networks. However, efficient location management algorithms are required to discover a destination's location before data transfer can be attempted using geographic routing. To be deemed scalable with respect to network size, mobility and traffic, the signalling overhead due to location management must be kept low so that the performance of geographic routing is minimally affected. In this research, we introduce a novel location management protocol known as Scalable Location Management (SLALoM), which outlines a scheme for partitioning a given terrain into ordered regions for location management. Our detailed analysis shows that under random node mobility and communication requirements, SLALoM improves upon the asymptotic location management cost compared to existing location management schemes. As an optimization, we use the concept of local forwarding to introduce a scheme called Efficient Location Forwarding (ELF) that mitigates the location update cost of SLALoM. We show that, while the asymptotic overhead cost by such an improvisation matches that of SLALoM, ELF outperforms SLALoM in practice. Noting that a two-level hierarchy leads to an overall reduction in the location management cost, we investigate the use of multilevel hierarchy to further minimize the signalling cost and make efficient use of the limited bandwidth of the wireless channel. We propose a novel grid ordering scheme known as Hierarchical Grid Location Management (HGRID) that yields only a logarithmic increase in the location update cost with respect to the number of nodes in a uniformly and randomly distributed ad hoc network. We also show that, under a specific framework, all the proposed protocols are scalable with respect to mobility and network size. We carry out extensive simulations to quantitatively compare the performance of the protocols under practical considerations that could not be incorporated into the analysis, and to study how location management can affect geographic routing. (Abstract shortened by UMI.)