Previous [ 1] [ 2] [ 3] [ 4] [ 5] [ 6] [ 7] [ 8] [ 9] [ 10] [ 11] [ 12] [ 13] [ 14] [ 15] [ 16] [ 17] [ 18] [ 19] [ 20] [ 21]


Journal of Information Science and Engineering, Vol. 26 No. 1, pp. 197-214 (January 2010)

An Analytical Study of People Mobility in Opportunistic Networks*

1Institute of Information Science
Academia Sinica
Taipei, 115 Taiwan
2Department of Computer Science
University of Massachusetts Amherst
Amherst, MA 01003 U.S.A.

An opportunistic network is a type of Delay Tolerant Network (DTN) in which communication opportunities are intermittent. Moreover, an end-to-end path between the source and the destination may never have existed, disconnection and reconnection are common occurrences, and link performance is highly variable or extreme. With numerous emerging opportunistic networking applications, strategies that can facilitate effective data communication in such challenging environments have become increasingly desirable. In particular, knowing the fundamental properties of opportunistic networks will soon be the key to the proper design of opportunistic routing schemes and applications. In this study, we investigate opportunistic network scenarios based on two public network traces, namely, the UCSD and Dartmouth traces. Our contribution is twofold. First, we identify the censorship issue in network traces that usually leads to a strongly skewed distribution of the measurements. Based on this knowledge, we then apply the Kaplan- Meier Estimator to calculate the survivorship of network measurements. The survivorship feature is used to design our proposed censorship removal algorithm (CRA) for recovering censored data. Second, we perform an in-depth analysis of the UCSD and Dartmouth network traces. We show that they exhibit strong self-similarity, and can be modeled as such. We believe these newly revealed characteristics will be important in the future development and evaluation of opportunistic networks.

Keywords: delay tolerant networks, survival analysis, Kaplan-Meier estimator, self-similarity, censorship removal algorithm

Full Text () Retrieve PDF document (201001_14.pdf)

Received January 11, 2008; revised May 7, 2008; accepted June 12, 2008.
Communicated by Chin-Laung Lei.
* A preliminary version of this paper [19] appeared in the 26th Annual Joint Conference of the IEEE Computer and Communications Societies (IEEE INFOCOM 2007). This material was based upon work supported by the National Science Council under grant No. NSC-95-2221-E-001-025.