Esmaeil Kheirkhah ¹, Sayyed Mehdi Poustchi Amin ², Hediyeh AmirJahanshahi Sistani ², Haridas Acharya ³

Affiliations
1 Department of Computer Engineering, Mashhad Branch, Islamic Azad University, IR
2 Department of Computer Studies and Research, Symbiosis International University, Pune, IN
3 Allana Institute of Management Science, Pune University, Pune, IN

Abstract

We studied Brute-force SSH attacks carried out on six different universities campus networks by using Honeypot Techniques. Brute-force password guessing attacks against SSH, FTP and telnet servers are the most common form of attack to compromise servers facing the internet. A key factor to avoid disruption of these networks is to defend it against Brute-force attacks. We focused on the attempts to gain remote access to our SSH Honeypots Plus Tools and techniques employed. There are striking similarities in the methods used to attack these dissimilar systems. The evidence shows that, pre-compiled lists of usernames and passwords that are widely shared form the basis for brute-force attacks. When the passwords were analysed, it was found that in the event of actual malicious traffic what was commonly understood to be strong password did not protect the systems from being compromised. The data from the study were used to evaluate the efficacy of a variety of techniques designed to defend the systems against these attacks. Table 17 lists some commonly recommendation for the protection of SSH servers.

Keywords

Honeypot, Honeynet, Internet Attacks, Network Security, SSH, Brute-force, Malware

Full Text

   

Hediyeh Amir Jahanshahi Sistani ¹, Haridas Acharya ²

Affiliations
1 Department of Computer Studies and Research, Symbiosis International University, Pune, IN
2 Allana Institute of Management Science, Pune University, Pune, IN

Abstract

Packet classification is an important function for a range of network management and control applications, e.g. firewalls, policy based networking, traffic accounting, and network address translation. All the packets have to be classified in proper order for emerging broadband internet services, applications such as Internet TV/Radio, gaming, Video on Demand (VoD) and e-businesses, who are in constant demand for a higher degree of transmission bandwidth, complex security, and specific Quality of Service (QoS). Packet treatment needs multiple classifications which help to justify, which flow an incoming packet belongs to, and with which rules the packet's header fields are to be treated. So these systems ought to be compared with every packet header, in comparison, against a large set of rules, and while analysing such related packets, often incur delay. Therefore, this problem can be reduced with the help of fast packet classification algorithm. This paper presents a modified version of the packet classification algorithm, called DimCut Packet Classification Algorithm using Recursive Dimensional Cutting. In order to get extended DimCut algorithm, in addition, some discovery&result can also be obtained. Modifications could also extend the DimCut algorithm by adding some new heuristics ideas, new implementing techniques and parallel programming that classifies packets based on five header fields which we propose in this paper.

Keywords

Cutting, Dimcut, Firewalls, Heuristics, Packet Classification, Rules

Full Text