A Lightweight Middleware Platform for Distributed Computing on Wireless Sensor Networks

Abstract

The peculiar features of Wireless Sensor Networks (WSNs) suggest to exploit the distributed computing paradigm to perform complex tasks in a collaborative manner, in order to overcome the constraints related to sensor nodes limited capabilities. In this context, we describe a lightweight middleware platform to support the development of distributed applications on WSNs. The platform provides just a minimal general-purpose software layer, while the application components, including communication and processing algorithms, as well as the exchanged data, are described symbolically, with neither preformed syntax nor strict distinction between data and code. Our approach allows for interactive development of applications on each node, and requires no cross-compilation, a common practice that makes the development of WSN applications rigid and time-consuming. This way, tasks and behavior of each node can be modified at runtime, even after the network deployment, by sending the node executable code

Programming distributed applications with symbolic reasoning on WSNs

“…On the receiver side the payload is interpreted as if it had been typed at the command line interface. We exploited this feature in the implementation of a Timing-Sync protocol [5], and to remotely program deployed nodes in an interactive manner through a command line interface [6].…”

“…As an added consequence re-uploads of the application code could not be excluded should the hardware-abstraction layer require modifications. In some previous works we showed that programming WSNs with an extensible high-level language without compromising code execution efficiency is feasible by adopting a programming methodology based on Forth and we introduced a lightweight middleware for WSNs based on it [5], [6]. In this paper we show how our middleware can be easily extended with the ability to process symbolic data.…”

“…All the values are represented in the chosen formalism throughout the computation. As shown in [5] and [6], symbol processing is quite effective in our implementation since what is being sent is actually executable code that is either interpreted or compiled as it is received. A summary of all the words implementing the formalism is provided in Table I.…”

A Stream Processing System for Multisource Heterogeneous Sensor Data

“…In [10], the authors described the outlook of middleware for wireless sensor networks. Authors in [15] designed and implemented a lightweight middleware platform for distributed computation on wireless sensor networks. As wireless sensor network is becoming a new paradigm [16] in the realm of Internet of Everything, some studies have been made on facilities virtualization.…”