Faculty of Azad University of Parand (jan. 2006-Dec. 2008)
Shams Shafigh, Z. Abbasi, F. Noroozi, "Dynamic bandwidth allocation with fuzzy logic for average delay guarantee" 9th International Conference on Advanced Communication Technologies. 2007, ICACT2007, Korea [pdf]
Shams Shafigh, F. Noroozi, Z. Abbasi, H.Khoda bandeh, "Dynamic bandwidth allocation with minimum long fluctuations" 8th International Conference on Advanced Communication Technologies,feb. 2006, Phoenix Park, Korea [pdf]
F.noroozi, F.Ayazi, K.Ahmadi, H.R Khodabande Z.Abbasi "Outsourcing of IT Services and Its Affect On IT Development in Iran". 2nd IEEE international conference on ICT from theory to application Syria 2006.
S.Mohrehkesh, M.Rashtian, Z.Abbasi, H.R.Khodabandeh,“Evaluation of IT development in Iran”, 2nd International Conferences on Information and Knowledge Technology IKT2005, Tehran, Iran
Z.abasi, F.Noroozi, M.Vahedian, H.R.Khodabandeh "Application of International Internet Exchange Center in Iran" 10th Annual Computer Society of Iran Computer Conference, 15-17 February 2005, Tehran, Iran
Z.Abbasi, F.Noroozi, A.Sharifi,"Congestion Level prediction of Self Similar traffic based on Wavelet transform" , 3nd IEEE conference in Tunisa on March. 2005 [pdf] .
Z.Abbasi, M.Analouee, “Congestion level prediction under self-similar traffic using reinforcement learning”, The International Symposium on Telecommunication Conference (IST2003) 2003 [pdf] .
Zahra Abbasi
PhD Candidate
School of Computing and Informatics
Arizona State University
Contact
email: zahra dot abbasi at asu dot edu
Mailing Address: Brickyard Engineering (BYENG) Building Room 517, 699 S Mill Ave, Tempe, AZ 85281
Education
M.S. Computer Engineering, Iran University of Science and Technology,Tehran, Iran 2003
B.S. Computer Engineering, Shiraz University, Shiraz, Iran, 2001
1-Network Simulation and Bandwidth Management
In this research improvement of QoS parameters measurements in nodal packet switches based on wavelet modeling is concerned. Frequency domain analysis has been used for predicting pattern of traffic. The wavelet analysis deal with the input signal at different scales or resolutions We introduce and verify a novel scheme for the congestion prediction of burst traffic based on the Haar wavelet energy coefficients. simulation tools: NS-2, OPNET
As TCP has been designed based on short-range-dependent traffic models such as Poisson, In this research the congestion prediction of TCP revised under self-similar traffic models and was improved using reinforcement learning algorithm. This project mostly involved in traffic modeling and the use of reinforcement algorithm to predict the congestion. I used network simulator (NS-2) to evaluate my methods.
Aim to providing a simulation tool for analysis of a nation wide data network, some developing in NS-2 is done. First of all traffic models were investigated and the bests were chosen which were simulated in NS-2. Moreover a new self-similar traffic generator was developed and implemented in NS-2. After testing and adding MPLS and Diffserv components to NS-2, it was ready to simulate networks. Using MRTG, stochastic traffic information between every 2 nodes of Our desired real national data network are collected and simulated under NS-2. At the end performance parameters of our desired network under various conditions were tested and some recommendations to its improvement were suggested.
2-Information Technology Development