On estimating joint interference for concurrent packet transmissions in low power wireless networks

Abstract

In a wireless network it is important to understand the nature of the joint interference generated at a receiver by multiple concurrent transmitters. This understanding helps developing packet scheduling algorithms. Prior experimental work using older generation mote-class radios (CC1000) have showed systematic deviations between estimation and direct measurement of the joint interference power, thus questioning whether the standard assumption that received signal powers are additive is applicable in practice. We, however, show via extensive experimentation that on newer generation radios (CC2420), the additive assumption is valid, particularly at the low power end.