- Have any questions?
- +91 9176206235
- info@phdsupport.org
PhD in Congestion identification in Machine Learning
Ph.D Research Topics in 5G LTE Network
May 27, 2021PhD in Security Issues in Wireless system Using IOT
May 27, 2021
PhD in Congestion Control in Computer Networks Using Machine Learning Approach
Machine learning is a subfield of artificial intelligence which defines the capability of a machine to imitate intelligent human behavior. It is a part of artificial intelligence where computer algorithm improves congestion control of TCP in wired/wireless networks. Decision tree boosting is best method for application appear .the high flexibility, adaptability provides extends traditional approach usedin multiple fields including network operation and management based on Congestion Control in Computer Networks using Machine Learning Approach
RESEACH APPROACH:
In machine learning we used decision tree based algorithm for providing usual result for classify the problems. To make a reliable estimate we randomly need to divide the database into two divisions that consist of learning sample as one part and its ROC curve, AUC, and error rate are evaluated on the validation sample. The main criteria to evaluate this protocol which are bandwidth usage in the case of wireless links and TCP in case of the wired network. There RL-based CC algorithm are much feasible on NS3 simulator which separates the calculation on simulation result and analysis. Thus by the decision tree the machine learning algorithm the research process will be proposed.
LATEST ISSUES:
- The model induced by boosting will still be too complex for some of the practical applications.
- Drawback of storing and update continuously about 40 variables based on different statistic packets.
- In The realistic network communication the engineering issues are to be significant for the RL-based CC algorithms.
- RL-Based CC algorithm requires high storage space especially for the continuous environments there it meets a size based issues.
- The standard regression trees is that suffer from a high variance in sometimes being competitive in terms of accuracy compared to other algorithms
PhD in Congestion Control in Computer Networks using Machine Learning Approach
PROPOSED SOLUTION:
A way to increase the throughput of TCP over wireless links is to prevent it from reducing its rate when it faces a loss due to a link error.
- Proposed a generalization based kanerva coding technique provides a methodology for learning agents to automatically adjust and manage the level for every prototype.
- Kanerva coding technique leads quick on-line policy development with minimal computation and memory space.
- Increasing the size of the training datashet will helps to increase the accuracy rate of the results.
- Thefile transfer in the wireless links with protocols helps to transfer three times faster than with standard TCP.
FUTURE PROPOSALS:
There still a possible computational efficiency requires to consider the impact of reduction of accuracy on behavior of the protocol.
- To improve the performance and robustness we need further research based on Congestion Control in Computer Networks using Machine Learning Approach to deal issues like computation time,data storage and pre-designed parameters.
- In future network environment are complicated with need for addressing such complexity and flexibility.
- Robust domain knowledge is needed to realize light weight learning based CC algorithms because for making decisions with concern demands on memory and future storage.
- The light weighted and efficient learning based models with general learning based platforms are need to reduce the future demand by implementing alternative research.
Our experts provide complete guidance for PhD in Congestion Control in Computer Networks using Machine Learning .
