ML-Based Traffic Classification in an SDN-Enabled Cloud Environment

Round 1

Reviewer 1 Report

Section 2. Related work - please add some work on ensemble model for IoT network security, e.g. Genetic algorithm and artificial neural network for network forensic analytics - https://ieeexplore.ieee.org/document/9245140

3.1. Dataset - please compare your dataset to datasets used in related works, how these datasets differ and compare to each other

4. Methodology - please explain why did you use Naive Bayes, SVM, Random Forest, and J48 tree and ignore all other algoritms available in Weka

Figure 8. Weka’s ARFF file with Netmate features - is the data on this figure really important or can it be removed or shown in table format with appropriate descriptions?

5. Results and Discussion - please compare your results (reliability etc) with similar results from literature (related similar works)

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

In this work, authors proposed an efficient Machine Learning method for traffic  classification in the SDN/Cloud platform.The authors need to make the below few changes to enhance the quality of this manuscript:

1. Introduce key features and concerns related to traffic classification that effect security and privacy of network.

2. Elaborate more about classification methods used in field of SDN/Cloud platform, specifically.

3. Mention the detail of your approach of supervised learning through different four algorithms in comparative way.

4. Literature needs to be improved by incorporating some quality papers. 

5. The authors should focus on the points to be solved and clarify how unique and effective the proposed work is against the existing works.

6. There are some grammar mistakes in the manuscript. Do a thorough grammar check before next submission.

7. Regarding the various symbols and notations, a table should be provided, explaining them.

8. Figures 11 and 12 should be more clear for Avg. Call and Avg. Precision of the classifiers on both sets of features.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors propose a Machine Learning methodology for network traffic classification in SDN/Cloud platforms. The goal of the paper is to manage the network flows based on the specific application requirements in order to improve the QoS. the idea is very interesting and many researchers have worked in this field in the last few years. In general, the paper is well written and easy to read. However, there are some major issues that should be addressed:  

1) The authors claim that the main contribution of their work resides in the fact that they propose an ML network classification study in cloud networks managed by SDN. This is not enough to support the novelty of the proposed work. Its mandatory to clarify the exact innovation points of the proposed work. 

2) The proposed solution performs network traffic classification in virtualized infrastructures (i.e. cloud, SDN) but there is no connection between the output of the ML analysis and the SDN controller (Open Day Light) or the management framework of the cloud (OpenNebula).

3)  The proposed solution uses datasets for ML training and inference that are created by tools that are executed inside the VMs (i.e. tcpdump, isof etc.), this approach requires root execution rights to users' VMs. The authors should explain why they choose this methodology and how their approach ensures security and is compatible with data privacy regulations.

4)In nowadays, the holistic management approach of computational and networking resources is also facilitated by NFV technology. For completeness' sake, I suggest section 2 to include some references also to this technology. Some relevant papers could be the following:

- Zafeiropoulos, Anastasios, et al. "Benchmarking and Profiling 5G Verticals' Applications: An Industrial IoT Use Case." 2020 6th IEEE Conference on Network Softwarization (NetSoft). IEEE, 2020.

- Uzunidis, Dimitris, et al. "Intelligent Performance Prediction: The Use Case of a Hadoop Cluster." Electronics 10.21 (2021): 2690.

- Troia, Sebastian, et al. "Machine learning-assisted planning and provisioning for SDN/NFV-enabled metropolitan networks." 2019 European Conference on Networks and Communications (EuCNC). IEEE, 2019.

5) There are some errors in the text, good proofreading in necessary.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Authors have made changes according to my previous comments.

Reviewer 2 Report

Authors  have revised the manuscript. The revised paper is looking good. It needs a minor revision for all grammatical and punctuations related  errors.