It is envisioned that in the future, all kinds of devices ranging from resource-constraint wireless sensor nodes to powerful server-class computers will interact to form an Internet of Things (IoT). In such a setting, tiny devices will extend the Internet to the physical world and allow for a completely new class of applications. However, until today, no widespread deployment of such IoT applications can be observed. Major challenges to master issues of embedded programming, massive distribution, resource constraints, heterogeneity, and seamless integration with traditional Internet technologies. Orchestrating such a number of different devices to form an application can be arbitrarily complex. In the Internet and especially in Enterprise IT, the concept of Service-Oriented Architectures (SOA) has been applied successfully to address this orchestration problem. However, the technologies used to realize SOAs (such as Web Services, HTTP, XML, or BPEL) are to heavyweight to be used in resource-constraint networks. In this paper, we present a comprehensive, Web Service-compliant approach to use SOA technologies in WSNs. Our approach comprises self-description of sensor nodes, a light-weight Web Service transport protocol (called Lean Transport Protocol, LTP), an approach for the model-based programming of sensor nodes, and a solution for the integration with the Internet. We present an exhaustive simulation showing the first-rate performance of the approach.
|Journal||International Journal of Next-Generation Computing|
|Number of pages||17|
|Publication status||Published - 01.07.2010|