A Scalable Pulse Protocol For Structural Health Monitoring

This paper presents a scalable and energy-aware pulse-based networking architecture, which is an alternative to traditional packet-based networking in low-information and energy-constrained applications. Various networked sensing applications, notably in the Structural Health Monitoring domain, require the collection of information which is essentially binary in nature (occurrence or absence of a localized event). For such applications, an energy-efficient pulse-position-based MAC and routing approach, called Cellular Pulse Networking (CPN), has been demonstrated earlier as an energy-light alternative to traditional packet-based approaches. Such a protocol encodes the essential information, namely presence or absence of a monitored event and location of origin of the event, using a few pulses and their position with respect to a synchronized time frame structure. It has been shown in prior work that the energy usage in the pulse networking approach is significantly smaller than the corresponding packet-based approach when the information needs are small. However, the pulse networking approach, as has been discussed till date, does not scale well because the frame structure, which is central to the architecture, increases quadratically with increase in the number of cells in the network. It is to be noted that cells are essentially localized clusters of network nodes which are assumed to have the same location indicated by their unique cell ID, thus representing the lowest granularity of spatial localization. In the current work, we present Scalable Cellular Pulse Networking (SCPN), which aims to retain the energy benefits of CPN while improving the delivery delay performance for large cell counts. This is obtained by employing a new frame, which is shown to scale linearly with cell count. Using simulations, we show the advantages of the energya-ware SCPN approach in terms of its inherent energy-efficiency, improved networking performance and more graceful scaling as the cell count increases. We also demonstrate important trade-offs as well as scenarios where more information can be packed in using the new scheme with energy benefits and delay at par with the baseline CPN approach.