ataei-thesis
0 likes31 views
The document discusses work on a thesis about artificial intelligence in wireless networks. It includes introducing network slicing and neural methods, reading documentation on reinforcement learning tools like Gym and Sacred, running simulations with the Nokia wireless suite and observing results with MongoDB, implementing Q-learning in Python and Matlab, and working to optimize resource allocation and improve efficiency. The goal is to apply reinforcement learning techniques to problems in frequency allocation for wireless networks.
ataei-thesis
- 1. Politecnico di Milano - AmirhoseinAtaei 1 25/09/2020 (sep) - Introduce the network slicing in 5g and 6g also neural methods - Introduce the nokia wireless suite full thesis Artificial intelligence in the air
- 2. Politecnicodi Milano - AmirhoseinAtaei 2 Considering INI Network slicing https://ieeexplore.ieee.org/abstract/document/8476595
- 3. Politecnicodi Milano - AmirhoseinAtaei 3 Nokia github
- 4. 09/19/2020 (oct) - Read the gym and sacred documentation also deep learning - Run the nokia wireless suite with observer Politecnico di Milano - AmirhoseinAtaei 4 full thesis Artificial intelligence in the air
- 5. GYMENVIRONMENT SACRED CONFIGURATION Politecnicodi Milano - AmirhoseinAtaei 5 Wireless-suite develop
- 6. Politecnicodi Milano - AmirhoseinAtaei 6 REINFORCEMENTLEARNING Machine learning https://youtu.be/kopoLzvh5jY
- 7. ADD TO PATH Politecnicodi Milano - AmirhoseinAtaei 7 Package should be add to path list
- 8. IMPORTMONGODB AS OBSERVER Politecnicodi Milano - AmirhoseinAtaei 8 Should be import mongo observer
- 9. PORT AND HOST DATA BASEPATH Politecnicodi Milano - AmirhoseinAtaei 9 Configuration of mongodb
- 10. CREATE DB AND USER TO ACCESS RUN MONGOSHELL Politecnicodi Milano - AmirhoseinAtaei 10 Running mongodb and it’s shell
- 11. COMPASSGUI Politecnicodi Milano - AmirhoseinAtaei 11 See the data in GUI mongodb compass
- 12. RUN LAUNCH AGENTACCORDINGCONFIGFILES Politecnicodi Milano - AmirhoseinAtaei 12 Run the simulation
- 13. REWARD VALUE AS RESULT Politecnicodi Milano - AmirhoseinAtaei 13 Get the reward number
- 14. STORED DATA Politecnicodi Milano - AmirhoseinAtaei 14 Evaluate all the datas that are saved
- 15. MAKE ENVIRONMENT Politecnicodi Milano - AmirhoseinAtaei 15 Gym environment
- 16. GYMENVIRONMENTEXAMPLE Politecnicodi Milano - AmirhoseinAtaei 16 Test open ai gym
- 17. Politecnicodi Milano - AmirhoseinAtaei 17 Matlab reinforcement learning https://it.mathworks.com/products/reinforcement-learning.html
- 18. Politecnico di Milano - AmirhoseinAtaei 18 26/10/2020 (oct) - Exploit all the specification of q-learning python program - improve the time efficiency full thesis Artificial intelligence in the air
- 19. Politecnicodi Milano - AmirhoseinAtaei 19 Virtualy run on top of powerful HW Powerful resource
- 20. Politecnicodi Milano - AmirhoseinAtaei 20 Q-learn https://www.youtube.com/watch?v=e3L4VocZnnQ Find and solve the problem of UMTS-OLPCcode Related to number of episodes
- 21. Politecnicodi Milano - AmirhoseinAtaei 21 Run the episodes Output plots
- 22. Politecnico di Milano - AmirhoseinAtaei 22 10/11/2020 (nov) - Solve the problems of q-learning python program - change the configuration and compare result full thesis Artificial intelligence in the air
- 23. Politecnicodi Milano - AmirhoseinAtaei 23 Fix the problem of plots Solve the problem and now We have two plot with different scale in figure one
- 24. Politecnicodi Milano - AmirhoseinAtaei 24 Change the result with different parameter Comparing results with change config file
- 25. CONFIGparameterslike carrier frequency Politecnicodi Milano - AmirhoseinAtaei 25 Parameters that we can change
- 26. Politecnico di Milano - AmirhoseinAtaei 26 01/12/2020 (dec) - find split in time frequency file - implement qlearn in time frequency program full thesis Artificial intelligence in the air
- 27. Politecnicodi Milano - AmirhoseinAtaei 27 Frequency resource allocation FRA