A Model Driven Framework for Modeling and Verification of Underwater Wireless Sensor Networks

Abstract

The domain of underwater wireless sensor networks (UWSNs) has recently captured the significant attention of researchers from both industry and academia due to its substantial enhanced capabilities in ocean surveillance, marine tracking, and device deployment for detecting underwater targets. However, wide-scale implementation of UWSNs is a long way off for a number of reasons. One of the most critical reasons is that ocean-centric applications are both expensive and time-consuming to deploy offshore and to test in the field. Therefore, verification of network requirements prior to actual deployment requires a quick and reusable solution. Although there are tools (simulators) for the analysis of UWSNs infrastructure, protocol and algorithms, these operate on lower abstraction level with higher complexities where knowledge of low-level code is essential. On the other hand, Model-Driven Engineering (MDE) is system development approach to provide higher level of abstraction. Although it has been utilized in the domain of UWSNs for system level design, the early analysis and verification of system design is usually performed at lower abstraction level in isolation. As a result, a major gap between design and verification has emerged significantly. To fill this gap, a user-friendly and higher abstraction layer is required that allows design and verification aspects to be modelled at the same level as system design. In order to bridge this gap, this thesis proposes a Model-Driven Framework for Underwater Wireless Sensor Networks (MFUWSN). Particularly, it permits the modeling of system design and verification aspects of UWSNs at higher abstraction level altogether. To achieve this, a UML Profile for Underwater Wireless Sensor Networks (UPUWSN) is developed to model design and verification aspects of UWSNs. A transformation engine named “UWSN Transformation Engine (UWSNTE)" has been developed as part of the research to automatically transform high-level MFUWSN source models into low-level C/C++ code. As a result, the AquaSim NS3 simulator can be used to do early analysis and verification of UWSNs design in the early stages of development. The applicability of framework is established through two benchmark case studies i.e., Monitoring Offshore Oil & Gas Reservoirs and Smart Cites Underwater. The results indicate that the proposed framework significantly simplifies the system design and verification of UWSNs as both design and verification aspects are managed altogether at higher abstraction level.