Survival of the Fittest: Data Dissemination with Selective Negotiation in Wireless Sensor Networks
Xiaolong Zheng,Jiliang Wang,Wei Dong,Yuan He,Yunhao Liu
DOI: https://doi.org/10.1109/MASS.2013.79
2013-01-01
Abstract:Data dissemination is a building block of wireless sensor networks (WSNs). In order to guarantee the reliability, many existing works rely on a negotiation scheme, making senders and receivers negotiate the schedule of transmissions through a three-way handshake procedure. According to our observation, however, negotiation incurs long dissemination time and seriously defers the network wide convergence. On the other hand, the flooding approach, which is conventionally considered to be inefficient and energy-consuming, may facilitate data dissemination if appropriately designed. This motivates us to pursue a delicate tradeoff between negotiation and flooding in the data dissemination process. In this paper, we propose SurF (Survival of the Fittest), a data dissemination protocol which selectively adopts negotiation and leverages flooding opportunistically. How to capture and utilize the opportunities when negotiation should be used is a challenging issue. SurF incorporates a time-reliability model to estimate the time efficiencies of the two schemes (flooding vs. negotiation) and dynamically selects the fittest one to facilitate the dissemination process. We implement SurF based on TinyOS 2.1.1 and evaluate its performance with 40 TelosB nodes. The results show that SurF, while retaining the dissemination reliability, reduces the dissemination time by 40% in average, compared with the state-of-the-art protocols.