Loading…

Understanding protocol performance and robustness of ad hoc networks through structural analysis

There has recently been renewed interest among various research communities in understanding the structure of social and infrastructure networks. Motivated by this line of research, we conduct an in-depth structural analysis of large ad hoc networks derived by placing nodes randomly as well as by pl...

Full description

Saved in:
Bibliographic Details
Main Authors: Barrett, C., Drozda, M., Engelhart, D.C., Anil Kumar, V.S., Marathe, M.V., Morin, M.M., Ravi, S.S., Smith, J.P.
Format: Conference Proceeding
Language:English
Subjects:
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:There has recently been renewed interest among various research communities in understanding the structure of social and infrastructure networks. Motivated by this line of research, we conduct an in-depth structural analysis of large ad hoc networks derived by placing nodes randomly as well as by placing them in realistic urban environments, a scenario that is rapidly gaining interest (K. Jain, et al, 2003). We use structural analysis in two illustrative settings. First, we use it to study the performance of network protocols. Our results indicate that structural analysis of interference graphs that model ad hoc networks can yield a good first order prediction of the overall protocol performance. Second, we study the robustness of a network to random node and edge failures. This study is important in the context of ad hoc networks wherein one expects nodes/edges to fail due various natural or system dependent reasons. The experimental results presented in this paper show the following: (i) structural properties of ad hoc networks depend crucially on the spatial distribution of the nodes. (ii) Structural properties of the network significantly affect the performance of protocols. (iii) Graph theoretic measures can provide good first order insights into the network protocol performance. (iv) The measures are also useful in characterizing the robustness of such networks.
ISSN:2160-4886
DOI:10.1109/WIMOB.2005.1512886