Link-correlation-aware data dissemination in wireless sensor networks

Abstract

Reprogramming is a crucial technique for software deployment and update in wireless sensor networks (WSNs). Bulk data dissemination is an important building block for enabling wireless reprogramming. Dissemination for reprogramming is required to be fast and energy efficient. Recent studies show that wireless link correlation can have a significant impact on the performance of bulk data dissemination. In this paper, we propose correlated dissemination (CD), a fast and energy-efficient bulk data dissemination protocol for reprogramming in WSNs. The main contribution of CD is that it disseminates data according to a novel structure called correlated tree, which considers both link qualities and link correlations. The construction of correlated tree is lightweight and allows for more efficient data transmission. Moreover, it selectively and dynamically employs rateless codes to improve the performance, especially in circumstances where link correlation is inherently weak. We implement CD based on TelosB testbed with both TinyOS and OpenWSN (with 802.15.4e MAC). Evaluation results show that, compared with previous works, CD greatly improves the dissemination performance in terms of completion time, transmission cost, and energy efficiency.