Energy-efficient differentiated directed diffusion (EDDD) in wireless sensor networks

A number of routing protocols [1] have been proposed for wireless sensor networks in recent years. Considering energy-efficiency as the primary objective, most of routing protocols focus on reducing the number of packet transmissions by clustering, leveraging geographical information, and so on. These routing protocols in sensor networks have the limitation of lacking application contexts for filtering or aggregation. To remedy this, Directed Diffusion (DD) [2], which utilizes application contexts in data dissemination, is proposed. However, DD cannot support time-sensitive traffic nor perform energy-balancing to increase network lifetime. To bridge this gap, this paper extends DD as follows: (1) real-time (RT) filters to provide better end-to-end (ETE) delay performance for real-time traffic, (2) best-effort (BE) filters to achieve global energy balance and to prolong network lifetime, (3) RT-repairs to fast recover node/link failure for RT traffic. The extended DD is dubbed energy-efficient differentiated directed diffusion (EDDD). Comprehensive simulation experiments show that EDDD has the following advantages: (1) differentiates dissemination service for RT and BE traffic, (2) achieves lower delay for RT traffic than DD, (3) exhibits substantially longer network lifetime than DD.