Abstract
Packet sampling plays an important role in keeping storage and processing requirements at a manageable level in network management. However, because it reduces the amount of available information, it can also reduce the performance of some related tasks, such as detecting security events. In this context, this work explores how packet sampling impacts machine learning-based tasks, in particular, flow-based C2 TLS malware traffic detection using a deep neural network. Based on a proposed lightweight sampling scheme, the ongoing results show a small reduction in classification accuracy compared with analysing all the traffic, while reducing in 10 fold the number of packets processed.

Abstract
Huge efforts and resources are spent every year on prevention and recovery of cyberattacks targeting users, services and network infrastructures. Software-Defined Networking (SDN) is a technology providing advances to the field of security with the ability of programming the network, promoting high-performance solutions and efficient resource utilization at low costs, as the use of specialized hardware is avoided. The present paper aims at exploring the SDN paradigm to develop an SDN-based framework for prevention and mitigation of malicious attacks throuhgt the network. The framework design and proposal has concerns regarding the efficient use of network and computational resources, distributing the inspection of suspicious flows by distinct Intrusion Detection Systems. For this purpose, a load-balancing strategy for traffic inspection is devised, allowing to balance both the usage of resources and the analysis of traffic flows. In this way, this paper also sheds light on the usage of OpenFlow messages to build distributed SDN-based applications with the mentioned properties.

Abstract

Abstract
As software becomes more complex and assumes an even greater role in our lives, formal verification is set to become the gold standard in securing software systems into the future, since it can guarantee the absence of errors and entire classes of attack. Recent advances in formal verification are being used to secure everything from unmanned drones to the internet. At the same time, the usable security research community has made huge progress in improving the usability of security products and end-users comprehension of security issues. However, there have been no human-centered studies focused on the impact of formal verification on the use and adoption of formally verified software products. We propose a research agenda to fill this gap and to contribute with the first collection of studies on people's mental models on formal verification and associated security and privacy guarantees and threats. The proposed research has the potential to increase the adoption of more secure products and it can be directly used by the security and formal methods communities to create more effective and secure software tools. © C. Carreira et al.

Abstract

Abstract
Vulnerability detection and repair is a demanding and expensive part of the software development process. As such, there has been an effort to develop new and better ways to automatically detect and repair vulnerabilities. DifFuzz is a state-of-the-art tool for automatic detection of timing side-channel vulnerabilities, a type of vulnerability that is particularly difficult to detect and correct. Despite recent progress made with tools such as DifFuzz, work on tools capable of automatically repairing timing side-channel vulnerabilities is scarce. In this paper, we propose DifFuzzAR, a new tool for automatic repair of timing side-channel vulnerabilities in Java code. The tool works in conjunction with DifFuzz and it is able to repair 56% of the vulnerabilities identified in DifFuzz's dataset. The results show that the tool can indeed automatically correct timing side-channel vulnerabilities, being more effective with those that are controlflow based.