Title: A Unifying Link Abstraction for Wireless Sensor Networks

Speaker: Joseph Robert Polastre

Abstract:

Recent technological advances and the continuing quest for greater
efficiency have led to an explosion of link and network protocols for
wireless sensor networks. These protocols embody very different
assumptions about network stack composition and, as such, have limited
interoperability. In principle, wireless sensor networks would benefit
from a unifying abstraction (or "narrow waist" in architectural
terms), and that this abstraction should be closer to the link level
than the network level. Through this abstraction, protocols may
transcend hardware generations and increase interoperability. This
dissertation proposes a specific unifying sensornet protocol (SP) that
provides best-effort single-hop communications.

The two goals of a unifying abstraction are generality and efficiency:
it should be capable of running over a broad range of link-layer
technologies and support a wide variety of network protocols, and
doing so should not lead to a significant loss of efficiency. Our
unified  abstraction, SP, differs from IP, the "narrow waist" of the
Internet, by permitting control information to flow between network
and link protocols in addition to data packets. Our approach is to
build a translucent interface with a minimal set of sufficient
primitives to build efficient protocols. SP includes mechanisms for
reliability, urgency, detecting congestion, and changing phase as well
as a shared neighbor table and message pool.

To investigate the extent to which SP is an effective unifying
abstraction, we implemented SP (in TinyOS) on top of two very
different radio technologies: B-MAC on Mica2 and IEEE 802.15.4 on
Telos. We also built a variety of network protocols on SP, including
examples of collection routing, dissemination, and aggregation.
Measurements show that these protocols do not sacrifice performance
through the use of our SP abstraction.

SP speeds up protocol development by separating concerns, and enables
cooperation between protocols for improved efficiency. Our results
show that protocols consumed up to 57% less energy by cooperating with
each other through our SP abstraction than executing independently. SP
serves as a common language to describe new protocols and a way to
systematically evaluate their merit. Finally, SP is the keystone to
developing a larger  wireless sensor architecture for building and
deploying sensor network applications. Although developed for wireless
sensors, SP's concepts are applicable to other wireless technologies,
like 802.11, and may motivate future revisions to the Internet
Architecture for greater portability and performance.