Tags:
create new tag
, view all tags
Monash WiSeNeT (or Wireless Sensor Network Testbed) has been developed in order to obtain real-world results of wireless sensor network congestion control and routing algorithms and techniques. It is a local software framework coded in Python to facilitate the implementation, testing, monitoring and debugging of various wireless sensor network projects and experiments. In the existing implementation of a wireless sensor network testbed at Monash University, traditional networking concepts such as layers and encapsulation were incorrectly utilised and many assumptions and shortcuts were taken to operate the testbed. Additionally, there was no clear indication of any standardisation. Amongst other things, this resulted in convoluted code, dependencies across many components, and poor resource management, and ultimately a system that could not easily be passed down to the next generation of students. Compared to this, WiSeNeTs key advantages are modularity, standardisation, reusability and easy-to-understand concepts. This sets WiSeNeT apart by offering a standard template for research into wireless sensor network algorithms without the need for major rework. It must be noted that the goal of WiSeNeT is not to be the fastest testbed, but to be a modular, simple to use and understand, robust testbed which can be reused across many projects. Built using layered concepts seen in theories such as the OSI model, WiSeNeT models all nodes in a wireless sensor network with a standardised model consisting of a link layer, a network layer, an application layer, and various other functions built around this approach (Figure 2). Its primary feature, and key difference from the existing testbed, is its modularity, which means that each layer and component within WiSeNeT has a clearly defined scope and is completely independent of the other layers, which therefore reduces the number of dependencies across the testbed. In order to achieve this, a standardised interface protocol has been developed in order to pass information between the different layers and components of the testbed.

wisenet-layered-approach.png Figure 2: WiSeNeT's layered approach. Note the absence of the transport layer. The reason is that we currently do not have any projects related to transport layer issues.

Modularity allows a simplification of the system landscape, resulting in an intuitive layout of the testbed which builds on common concepts and theories presented as part of studies in telecommunications and networking at Monash University. This leads to concepts and a code structure that is neat and easy to understand, and a system that is highly expandable and robust. On top of that, this approach has been improved through standardisation of various processes and concepts, rather than customising and reconfiguring testbeds for every new project in wireless sensor networks. For example, a standardised packet structure has been developed and future-proofed based on common concepts and probable future requirements of congestion control and routing algorithms. The scope of various processes and functions within the testbed has been clearly defined such that it simplifies thinking and analysis, and removes the need for assumptions, whilst aiding reusability. WiSeNeT has been coded in Python, which is a powerful object-oriented language with a number of improvements over C, one of which is its easier-to-understand concepts, especially in regards to threading. It has been baselined with ALOHA characteristics with exponential backoff, a basic and common approach to wireless sensor networks from which future research can build on top of. Useful functions from the existing testbed have been expanded to incorporate the new design, and in many cases simplified. One example is the logging capability, which previously captured just the minimum information required and now can be utilised to monitor and analyse almost any part of the testbed and network. The testbed employs heavy use of threading and therefore necessitated effective resource management of all shared resources and variables. It is also able to handle any amount of connected packet radios the maximum number tested to date is 13 radios connected to the same computer.

Topic revision: r3 - 2012-01-23 - AhmetSekercioglu
 
This site is powered by the TWiki collaboration platformCopyright © 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback