Computing Reviews

Performance analysis of opportunistic broadcast for delay-tolerant wireless sensor networks
Nayebi A., Sarbazi-Azad H., Karlsson G. Journal of Systems and Software83(8):1310-1317,2010.Type:Article
Date Reviewed: 11/08/10

Imagine that there is an interplanetary Internet (IPN) between Earth and Mars. The shortest distance between Earth and Mars is 100 million kilometers (km). With the speed of light, a signal will take about 667 seconds round trip. The current Internet protocol depends heavily on the round-trip time for its flow control and congestion control. Such a large round-trip delay makes the current protocol unworkable. Therefore, we have to devise a new kind of protocol that is tolerant of this long delay. Hence, what we need for an IPN is a delay-tolerant network (DTN).

The IPN was first studied by some Internet pioneers, including Vint Cerf. An IPN special interest group was formed under the Internet Society (ISOC). It is obvious that without support from the Pentagon and the National Aeronautics and Space Administration (NASA), the IPN would not be possible. The recent worldwide economic downturn and budget cuts make the IPN a very low priority, so it is safe to say that the IPN project has been indefinitely postponed.

Without real application scenarios, a DTN begins to find its way into other domains. This paper applies the DTN concept to wireless sensor networks (WSNs)--delay-tolerant WSNs (DTWSNs). It studies problems related to routing, data gathering, and neighbor discovery, within the scope of DTWSNs. A simple opportunistic broadcast algorithm is proposed, analyzed, and simulated. However, in the first step of the algorithm, a mobile node frequently and synchronously turns on and sends query packets. What does “synchronously” mean? The term is never mentioned again in the paper.

The paper does not refer to two requests for comments (RFC) about architecture and protocols for DTN that are closely related to this paper and the opportunistic broadcast algorithm (RFC4838 and RFC5050). Furthermore, the assumption that each sensor node must buffer all the data from other sensor nodes faces the problem of memory overflow. This should be considered and dealt with.

Apart from the comments above, this well-written paper is easy to follow and understand.

Reviewer:  R. S. Chang Review #: CR138565 (1105-0505)

Reproduction in whole or in part without permission is prohibited.   Copyright 2024 ComputingReviews.com™
Terms of Use
| Privacy Policy