Xavier Araújo

15 Feb - 20 Feb

OBJECTIVES: Solve mesh throughput problem. Understand how to implement an energy model in NS-3. Run multiple simulations automatically.
WORK: Before the implementation of the variable energy test scenario in NS-3, as was inittialy planned for this week, the attention was focused in solving a throughput problem which arose in the simulations exposed in the PDI report. This problem extended for more time than it was expected, and so, during this week, the attention was focused in solving it, to obtain accurate simulation results. The problem was discovered and fixed almost by the end of the week. After fixing this problem the attention was focused in discover how to implement an energy model in the NS-3 simulation. During this week were also made changes to the previous written code, to be able to execute multiple simulations automatically. This work was very important since it will not be viable to execute the simulations manually, when I need to run hundreds of simulations.
PROBLEMS: The main problem of this week was the throughput problem affecting the mesh network simulations. This was a serious problem, detected a few weeks ago, which was finally solved this week.

Close

22 Feb - 27 Feb

OBJECTIVES: Implementation in NS-3 of a test scenario of multi-hop networks with variable available energy.
WORK: After researching, it was discovered that the NS-3 already provides an energy framework to be used in the simulations. By making use of this energy framework, was configured a test scenario with variable available energy, in a multi-hop network. To program this test scenario, were used the mesh model and the energy model of NS-3. A little adaptation of NS-3 source code was needed, in order to interconnect the mesh model with the energy model. In this week was also written the sub-section of the thesis referring to the energy model, which explain its mode of operation.
PROBLEMS: The main problem of this week was to adapt the NS-3 source code to be able to interconnect the mesh model with the energy model, since it required to analyze the NS-3 source code and understand the modifications which had to be done.

Close

1 Mar - 6 Mar

OBJECTIVES: Specify a modification to TILIA channel assignment algorithm in order to make it energy aware.
WORK: The main concern of this week was to specify modifications to TILIA channel assignment algorithm in order to adapt it to networks in which the energy availibity is a concern. After some analysis it was decided that a basic modification should be done and tested, before making a complex modification. So it was decided that the modification to be done was to choose the parent of a certain node based on the energy parameter. This way the node will select as its parent, the candidate which has the highest amount of energy. We think that with this little modification it is possible to increase the network lifetime. This hypothesis should be tested before making more complex modifications. TILIA python script was analyzed to understand where these modifications should be placed.
PROBLEMS: The main problem of this week was to understand how TILIA channel assignment algorithm should be modified, in order to increase the network lifetime. A possible solution was proposed but it will probably be necessary to review it later.

Close

8 Mar - 13 Mar

OBJECTIVES: Modify the TILIA algorithm to obtain the first version of the E-TILIA algorithm, according to what was decided in the previous week. Integrate E-TILIA with the NS-3 simulation.
WORK: E-TILIA must be executed under a regular time basis, changing the nodes channel dynamically, in order to adapt to changes in the energy levels of the network. In this week the work focused in making the integration of the algorithm with the NS-3 simulation. It was necessary to keep track of the energy level of the nodes in order to know when to stop the simulation and when to deactivate a certain node, due to its lack of energy. It was necessary to make modifications in the NS-3 simulation in order to stop the applications of the nodes which run out of energy and to disable their network interface.
PROBLEMS: The main problem of this week was to integrate the E-TILIA algorithm, under a regular time basis, inside the simulation. Another problem was to change the network parameters to obey to the decisions made by the algorithm. The fact that the energy model doesn't deactivate the nodes when their energy is completely drained, was also a obstacle which had to be overcome.

Close

15 Mar - 20 Mar

OBJECTIVES: Discover the problem related to the energy model.
WORK: During the simulations of the energy model the results were not what was expected. The nodes who behave as relay node were expected to deplete all their energy much faster than the other nodes, since they receive and transmit much more information, but in the simulations all the nodes deplete their energy almost at the same time. This problem may be related to the default current, assumed by the energy model, for each wifi-radio state. So I tried to discover the current wasted by a Raspberry Pi in each one of the wifi-radio states, to understand if this is really the cause of the problem.
PROBLEMS: The main problem of this week was the simulation error obtained with the energy model. All the effort of this week had as goal to discover and fix this error, in order to obtain reliable data from the simulations made. It was also detected an higher number of lost packets after introducing the energy model, but I think that this may be related to the shutdown of the network interface of the nodes, when they run out of energy.

Close

22 Mar - 27 Mar

OBJECTIVES: Discover the problem related to the energy model. Adapt the simulation to random topologies.
WORK: This week I have tried to better understand the source of the error related to the energy model. Although I couldn't solve it yet, I am getting closer. I discovered that the problem is in the physical layer of the simulation, which is sending and receiving much more traffic than it was expected. By analyzing the simulation results I came up with some possible explanations for these results. This possible explanations will be tested in the next week. In this week I also made the code to generate random topologies, with some pre-defined characteristics, to test the E-Tilia algorithm. It were also made the necessary adaptions to the simulation code in order to adapt it to random topologies.
PROBLEMS: The main problem of this week was the error in the energy model. This problem has been dragging for quite some time but it has to be fixed in order to obtain accurate results. Another problem was to generate random topologies with a certain degree of connectivity and with a correct positioning of the gateways.

Close

30 Mar - 03 Apr

OBJECTIVES: Discover the problem related to the energy model.
WORK: This week I was focused in solving the problem related to the energy model. This was very important because it was not possible to make progress in the thesis until this problem was fixed. Fortunately this week I was able to find the source of the error and to fix it. So now I am getting the simulation results which I was expecting. This week I also fixed some mistakes present in the random topologies simulation. A substantial amount of time was also dedicated to the thesis writing.
PROBLEMS: The main problem of this week was the error in the energy model which after several weeks was finally solved.

Close

06 Apr - 10 Apr

OBJECTIVES: Test the algorithm with random topologies.
WORK: This week I studied the TILIA script to understand where it should be modified. After I this I modified the script, creating the first version of the eTILIA algorithm. After test the algorithm with some random topologies I realized that the random topologies generated were not the best so I started to improve the random topology generator. Improve random topology generator. Trying to understand where we could modify tilia to obtain etilia. Do the modifications and test them.
PROBLEMS: The main problem of this week was to understand how should TILIA script be modified. The impromevement of the random topology generator was also a problem.

Close

13 Apr - 17 Apr

OBJECTIVES: Improve the random topology generator. Test eTILIA algorithm with random topologies.
WORK: This week I realized that the procedure of the random topology generator to discover the neighbours was not appropriate. So this week I developed a new strategy to obtain this information.
PROBLEMS: The main problem of this week was to discover the neighbour nodes in the random topology generator.

Close

20 Apr - 24 Apr

OBJECTIVES: Conclude the random topology generator. Test eTILIA algorithm with random topologies.
WORK: This week was concluded the random topology generator. With the new adopted strategy, the neighbours are discovered through a short indepedent simulation. So now I am generating good random mesh topologies to test eTILIA. After this goal was completed I started to simulate eTILIA with random topologies. So this week I was able to obtain the first valid simulations results and I could already see some improvements in terms of energy efficiency.
PROBLEMS: The main problem of this week was to conclude the random topology generator in order to obtain good random mesh topologies.

Close

27 Apr - 01 May

OBJECTIVES: Test eTILIA algorithm with random topologies. Modify the HWMP routing protocol from NS-3. Thesis writing.
WORK: This week we continued to test the eTILIA algorithm with random topologies, and concluded that to obtain accurate simulation results we had to modify the HWMP protocol of NS-3. The idea was to modify it in order to take into consideration the parent choice made by the eTILIA algorithm.
PROBLEMS: The main problem of this week was to modify the source code of NS-3 to adapt the HWMP protocol to the parent choice made by eTILIA.

Close

04 May - 08 May

OBJECTIVES: Thesis writing.
WORK: This week I focused on the thesis writing.
PROBLEMS: The main problem of this week was to organize some of the sections of the thesis.

Close

11 May - 15 May

OBJECTIVES: Thesis writing. Modify the HWMP routing protocol from NS-3.
WORK: This week I focused on the thesis writing and on solving the routing problem of the simulations. I already modify the source code of NS-3 but the routes that I add are being deleted too soon. The next step is to solve this problem in order to obtain accurate simulations.
PROBLEMS: The main problem of this week was to solve the routing problem of the simulations.

Close

18 May - 22 May

OBJECTIVES: Algorithm testing with regular and random topologies.
WORK: This week I focused on testing the eTILIA algorithm with both regular and random topologies. The analysis of the simulation results was the main priority during this week.
PROBLEMS: The main problem of this week was to understand the results given by the simulations.

Close

25 May - 29 May

OBJECTIVES: Algorithm testing with regular and random topologies. Thesis writing.
WORK: This week I focused on testing the eTILIA algorithm with both regular and random topologies. The analysis of the simulation results was the main priority during this week. A certain amount of time was also dedicated to the thesis writing.
PROBLEMS: The main problem of this week was to understand the results given by the simulations.

Close

01 Jun - 05 Jun

OBJECTIVES: Algorithm testing with regular and random topologies. Adapt the parameters of the simulation. Thesis writing.
WORK: This week I focused on testing the eTILIA algorithm with both regular and random topologies, and adapting the simulation parameters to better characterize the system. A certain amount of time was also dedicated to the thesis writing.
PROBLEMS: The main problem of this week was the adaptation of the parameters of the simulation according to the characteristics of the system.

Close

08 Jun - 12 Jun

OBJECTIVES: Algorithm testing with random topologies. Simulation results analysis. Thesis writing and review.
WORK: This week I focused on testing the eTILIA algorithm with random topologies and analyse the results. A certain amount of time was also dedicated to the thesis writing and review.
PROBLEMS: The main problem of this week was to analyse the simulation results.

Close

15 Jun - 19 Jun

OBJECTIVES: Thesis writing and review.
WORK: This week was dedicated to the thesis writing and review.
PROBLEMS: -

Close

22 Jun - 26 Jun

OBJECTIVES: Thesis writing and review.
WORK: This week was dedicated to the thesis writing and review.
PROBLEMS: -

Close