Project Aims
Simulation is heavily used in WSN research, but results are only as realistic as the models that they are built around. To ensure a close correlation between simulation and practical results, this research has developed communication, sensing, energy and timing models. WSNsim (a simulator for WSNs) has been developed as a model-centric simulator built around a structured ‘unified’ stack for communications, energy management and intelligent sensing.
WSNsim: A Simulator for WSNs
WSNsim (Wireless Sensor Network Simulator) was developed to debug, evaluate and improve algorithms developed at the University. WSNsim is an in-house object orientated discrete-event simulator for WSNs, developed using Microsoft Visual Studio .net 2005 (due to limited support and documentation, it is not currently available as open-source).
 
Left: The Structure of WSNsim, Right: A Network Under Simulation in WSNsim
As shown above, WSNsim gives considerable emphasis to the range of environmental and physical models that it encompasses. Furthermore, the use of a ‘unified’ stack for communications, energy management and intelligent sensing further increase the attention that WSNsim gives to these areas, while also allowing the simulation of code designed using a structured layered process.
Environmental and Physical Modelling
To support WSNsim, a range of environmental and physical models were investigated, encompassing communication, energy, sensing and timing.
Modelling Wireless Communication and Propagation
The communication model considers both path loss (using an empirical path loss model) and packet reception (by considering BERs at a per-byte level). Energy models are proposed for energy stores (batteries and supercapacitors), energy sources (photovoltaics and vibration energy harvesters), and energy consumers (radio transceivers, microcontrollers, and peripherals).
Modelling Energy Stores (Top Left), Energy Sources (Top Right) and Sensors (Bottom)
Sensor models account for errors and inaccuracies in sensed data by modelling the sensor hardware. Timing models consider the differences between the ‘true’ time, and the nodes’ perceptions of time.
Type: Postgraduate Research
Research Groups: Electronic Systems and Devices Group, Pervasive Systems Centre, Electronics and Electrical Engineering
Theme: Simulation, Modelling and Evalution
Dates: 1st October 2004 to ?
KeywordsFundingPrincipal InvestigatorsOther Investigators |
