Small Large-Scale Wireless Networks: Mobility-Assisted Resource Discovery

Abstract

In this study, the concept of small worlds is investigated in the context of large-scale wireless ad hoc and sensor networks. Wireless networks are spatial graphs that are usually much more clustered than random networks and have much higher path length characteristics. We observe that by adding only few random links, path length of wireless networks can be reduced drastically without affecting clustering. What is even more interesting is that such links need not be formed randomly but may be confined to a limited number of hops between the connected nodes. This has an important practical implication, as now we can introduce a distributed algorithm in large-scale wireless networks, based on what we call contacts, to improve the performance of resource discovery in such networks, without resorting to global flooding. We propose new contact-based protocols for adding logical short cuts in wireless networks efficiently. The new protocols take advantage of mobility in order to increase reachability of the search. We study the performance of our proposed contact-based architecture, and clarify the context in which large-scale wireless networks can be turned into small world networks.