Computer Science & Software Engineering Mobile and Wireless Computing (CITS7219)

Faculty Home  |  CSSE Home  |  csentry  |  CITS7219  |  help7219

Mobile and Wireless Computing 2009 - Written paper on wireless sensor networks

Submit short written answers (half to a page) on each of the questions below. Questions are based on the recommended reading for each lecture and the laboratory exercises for the Wireless Sensor Networks component of CITS7219.

Question 1: In the paper ‘The Hitchhiker's Guide to Successful Wireless Sensor Network Deployments’ from SenSys 2008, the authors report on experience with the development, testing and deployment of wireless sensor network applications. Select any one of these three categories and, in your own words, summarise the challenges reported in that area.

Question 2: Search the web for several currently available sensor network nodes. For example, CSIRO Fleck, Crossbow Mica2 and MicaZ, SunSPOT, BTNode. For two different nodes, make a summary list of its characteristics including the properties of its radio and microcontroller sensors supported, method of programming, and its price.

Question 3: Using the flooding simulator introduced in laboratories, explain how the density of the network affects the time taken to complete the flood. Using either the random or a regular topology, experiment with different network densities by adjusting the numbers of nodes and the landscape size. For each parameter setting, measure how long the flood takes to settle. You will need to run each test at least 2 or 3 times since the flooding algorithm is highly non-deterministic, and so the results will be different each time. Record your results in a table such as the one below. Why does the flood message sometimes fails to reach all nodes? (Hint: there are a number of different causes for this – try to identify at least one cause)

Topology (random or regular)

Number of Nodes (and source node position)

World size

Number of nodes reached by the flood

Time(s) taken for all nodes to receive the flood

Time(s) taken to settle (number of steps)

Question 4: Adjust the landscape size and number of nodes until you have a setting for which the broadcast storm effect is significant. That is, where time to settle is much greater than time to reach all nodes. Identify a few settings that have the broadcast storm effect, and a few settings that do not. Then select one of the settings you discovered in Question 3 that has the broadcast storm effect. For your chosen setting, adjust the retransmission probability to optimise the performance of the flood protocol. The goal is to make the settling time closer to the reception time. What is the optimal retransmission probability for your setting? Can you suggest a general way of selecting a suitable retransmission probability for other settings?

Question 5: You have been running single experiments to gain an understanding of this system. In order to prove any of the hypotheses you made in the experiments above, you would need to run multiple simulations, and analyse the statistical spread of your results. The program FloodExperiments.java provides a framework for running multiple experiments, writing the output to standard out in comma separated form, so that you can use a spreadsheet or Matlab for analysis and graphing. Modify FloodExperiments to run a generalised form of one or more of the previous experiments. Do the general results support your hypothesis. Present your results using tables and graphs, and with an argument justifying your conclusions.

Question 6: Consider the problem of sharing information between people in a city centre. For example, people could share information on traffic congestion, good shopping opportunities, or environmental conditions (pollution). Choose any two of the opportunistic networking protocols described in the paper Opportunistic Networking: Data Forwarding in Disconnected Mobile Ad Hoc Networks by Pelusi et al (see reading list on the CITS7219 schedule). Give a brief explanation of how each of the protocol strategy you selected could be used for this application scenario. State which protocol would be best for this scenario, giving reasons.

Question 7: The directed diffusion protocol uses several elements: producers, consumers, interests, data, messages, gradients and reinforcements. In your own words, explain each of these terms.

Question 8: Following the instructions in the Diffusion tutorial, run the simulator for One Phase Pull Diffusion with noisy transmission. Re-run the experiment for 4 or 5 different runs. That is, stop the application and restart java Simulation. Describe the results of each experiment. What do you observe? In what way do different runs of the protocol differ from one another? Make a table to summarise the results of each experiment.

Question 9: Summarise the main advantages and disadvantages of three data dissemination protocols you have studied: flooding, directed diffusion and trickle. Consider properties such as energy efficiency, robustness in noisy WSN environments, and simplicity of implementation.

Question 10: For a wireless sensor network application of your choice, describe two situations of interest. Describe each of these situations using the situation representation language presented in the paper Cardell-Oliver and Liu, Representation and Recognition of Situations in Sensor Networks (see reading list on the CITS7219 schedule)

---

The fine print

• The deadline for this exercise is 12noon Friday 23rd October (end of the semester). The exercise contributes 30% of your assessment in Mobile and Wireless Computing this year.
• Your submission is expected to be “the effort of the submitting student who may have discussed general principles with others in order to understand the topic, but who is responsible for submitting an individual piece of work;” (see http://web.csse.uwa.edu.au/internal_pages/school_policies/plagiarism2 )
• Use the web-based submission program, cssubmit, to submit:
• A single file (doc or pdf) listing the questions above with your answers.

Please post any requests for clarification about the requirements of the exercise to help7219, so that all students may remain equally informed.

Good luck,

Rachel Cardell-Oliver (unit co-lecturer)

Assignment published on CITS7219 web pages September 2009 (minor corrections 9 October 2009)

Top of Page

CRICOS Provider Code: 00126G