Coventry University logo

Welcome to Coventry University website

Skip Navigation


Research

My Current Research

My Phd Project  

My current research involves the investigation of information retrieval techniques in wireless sensor networks. A review on the more popular approaches to information extraction has been completed and compiled into a Cogent Computing internal technical report. Further, an in-depth investigation into the TinyDB query processing system has led to a number of small, useful WSN applications. These were on display at the National Science Week event at Bablake School.

I am now investigating in-network, complex query processing and how that can be facilitated using logical abstractions (some of which are used in today’s macroprogramming systems). As a first step, experiments utilizing these abstractions were tested in simulation using the SenSor simulator. The simulations compared the efficiency of the approach against other common query processing techniques in use today. More details on this ongoing research are available on my PhD project page.

My previous WSN research

Prior to my Research Assistant role my Masters work was spent on the ASQue Project. There the focus was on creating a formal specification of a parser for the ad-hoc sensor networks query language, ASQue, using the B method. The aim of ASQue is the design of a language that allows its validation and verification at compile time. The B specification, by definition, was a top-level specification so details of the language were not implemented. Rather, it addressed the behaviour required by the parser and created a design flexible enough to eventually allow multiple implementations. Subsequent work on the ASQue project involved the implementation of this B design. The code generated as a result of the B implementation phase formed the basis for a compiler front-end developed in C.

The increased availability and affordability of wireless technology has led to a rise in the number and breadth of sensor networks, networks that allow individual nodes to collaborate and coordinate themselves to achieve a larger task. Given the complexity that this scenario presents, there is a need to gather the information from the area of the network where it is generated and to be able to send information where it is needed.

ASQue

Although some research promotes the agent-based approach to information extraction our view is that separating the user from the internal workings of any information extraction process is critical to its usability and utility. An applicative querying tool is seen as the more practicable approach where a user can be certain of what a query means and its output without having to know how the query is actually executed.

The ASQue query language described in detail here was proposed in an effort to allow the easy construction, execution and processing of queries within ad-hoc sensor networks and has been preliminarily specified using a form of first order logic. The focus here is correctness of queries at compile time. To that end a formal specification and validation are proposed as a way of improving the likelihood that queries constructed and issued to the sensor network are correct, clear and reliable. Building a formally correct compiler for ASQue is important in facilitating this process.

ASQue parser – Specification to Implementation
A formal specification brings with it a number of advantages. It reduces a user’s uncertainty in the system, allows correction early on in the development process (prior to implementation of a system) and can lead to a higher quality error-free product. Formal methods have been used successfully in a number of areas of system development. In the area of wireless sensor networks they have been used to specify and in some cases validate a number of WSN algorithms.

The B method was selected as the technique for creating the formal specification, for a number of reasons. First, it provides support for all stages of development of a system, from specification to implementation. Second, it allows generation of actual code by providing refinement steps that allow the user to introduce more detail while being confident in the correctness of the new details being added. For details of this design and implementation download the following publications.

Publications

T.Daniel, S.Mount, R.M.Newman, E.I.Gaura (2006)Towards Trusted Compilation for Querying in Wireless Sensor Networks, Proceedings of the Eighteenth International Conference on Systems Engineering (ICSE '06), Coventry University, UK, pp. 89-95.  PDF

T.Daniel, S.Mount, R.M.Newman, E.I.Gaura (2006)Towards a Trusted Compiler for a Query Language for Wireless Sensor Networks, Proceedings of the Second International Symposium on Leveraging Applications of Formal Methods, Verification and Validation (ISOLA 2006), Paphos,Cyprus, pp. 277-282. PDF 

top of page