Download now Free registration required
By bringing the physical world and the virtual one closer, one is able to expand the possibilities of accelerating and easing embedded system design. In case of Wireless Sensor Network (WSNs), which require testing of the developed applications in hundreds to thousands of nodes, this is even more useful. The developers often find it difficult to efficiently manage test beds that have huge number of nodes. To tackle this problem, they mostly make use of simulation frameworks to create virtual sensor nodes. This helps them to provide levels of abstraction to specify the applications which will be executed on the nodes. Though this serves the purpose to some extent, but due to complete of direct interfaces with the physical environment, this kind of simulation doest not offer complete solution.This paper presents a hybrid simulation framework for WSN application development. This framework is programmed to interconnect a virtual network with the physical network to simulate the networks as a whole. The WSN applications can be modeled by developers using high level abstractions. This can be further used for multi-platform automatic code generation. The paper uses MathWorks tools to describe and test an extensible framework for platform independent application modeling and hybrid simulation for sensor network algorithms. MathWorks tools come with rich libraries for digital signal processing and control algorithm behavior simulation, which make the task easier.
- Format: PDF
- Size: 320.8 KB