To sleep or not to sleep: understanding the social behavior of lifetime-aware networks

Network lifetime is one of the key characteristics for evaluating wireless sensor networks (WSNs) in an application-specific way based on the availability of sensor nodes, wireless radio coverage, and wireless connectivity. Basically it shows in a resource constrained environment the consumption of every limited resource must be considered. A large number of energy efficient protocols and algorithms have been proposed in WSNs, mainly by introducing a sleep mode (SM) state to prolong the lifetime of a sensor network. The network nodes or links can be switched between working and sleep modes dynamically according to the real-time traffic situations. While there are far less critical discussions on what can be the negative effects of SMs on network lifetime in terms of hardware reliability such as failure rate. The duration of SMs tends to increase hardware lifetime, while the frequency of power state transitions tends to decrease it. In this paper, we extend the lifetime concepts in WSNs to wired network to reveal the side-effects of SMs on the hardware reliability. We have extensively studied the lifetime behavior of network links in a backbone network scenario as well as identified the sensitive social factors impacting the network lifetime. This novel research dimension is thought-provoking and opening a new conversation for researchers who are working in the areas of sustainable communications and computing to rethink and redesign the energy efficient approaches so as to address their possible side-effects on hardware reliability for the next stage of their implementation.