Abstract
VoIP services in general, and Session Initiation Protocol (SIP) ones in particular, continue to grow at a fast pace and have already become a key component of Next Generation Networks (NGN). Despite this proliferation, SIP-based services expose a large attack surface for perpetrators and especially those who seek to cause Denial of Service (DoS). While so far a plethora of works in the literature have been devoted to the detection of DoS attacks in SIP ecosystems, the focus is on those which exploit SIP headers neglecting the message body. In an effort to fill this gap, the work at hand concentrates on the detection of DoS attacks which instead capitalize on the Session Description Protocol (SDP) part of SIP requests. To this end, we not only scrutinize this ilk of attacks and demonstrate their effect against the end-user, but also develop an open source extensible SDP parser module capable of detecting intentionally or unintentionally crafted SDP segments parasitizing in SIP requests. Following a firewall-based logic,currently, the parser incorporates 100 different rules organized in 4 categories (policies) based on the corresponding RFC [1]. Through extensive experimentation, we show that our scheme induces negligible overhead in terms of processing time when working as a software module in either the SIP proxy or a separate machine in front of the latter.

Abstract
Voice over IP (VoIP) architecture and services consist of different software and hardware components that may be susceptible to a plethora of attacks. Among them, Denial of Service (DoS) is perhaps the most powerful one, as it aims to drain the underlying resources of a service and make it inaccessible to the legitimate users. So far, various detection and prevention schemes have been deployed to detect, deter and eliminate DoS occurrences. However, none of them seems to be complete in assessing in both realtime and offline modes if a system remains free of such types of attacks. To this end, in the context of this paper, we assert that audit trails in VoIP can be a rich source of information toward flushing out DoS incidents and evaluating the security level of a given system. Specifically, we introduce a privacy-friendly service to assess whether or not a SIP service provider suffers a DoS by examining either the recorded audit trails (in a forensic-like manner) or the realtime traffic. Our solution relies solely on the already received network logistic files, making it simple, easy to deploy, and fully compatible with existing SIP installations. It also allows for the exchange of log files between different providers for cross-analysis or its submission to a single analysis center (as an outsourced service) in an opt-in basis. Through extensive evaluation involving both offline and online executions and a variety of DoS scenarios, it is argued that our detection scheme is efficient enough, while its realtime operation introduces negligible overhead.

Abstract
Password management has long been a persistently challenging task. This led to the introduction of password management software, which has been around for at least 25 years in various forms, including desktop and browser-based applications. This work assesses the ability of two dozen password managers, 12 desktop applications and 12 browser plugins, to effectively protect the confidentiality of secret credentials in six representative scenarios. Our analysis focuses on the period during which a Password Manager (PM) resides in the RAM. Despite the sensitive nature of these applications, our results show that across all scenarios, only three desktop PM applications and two browser plugins do not store plaintext passwords in the system memory. Oddly enough, only one vendor recognized the exploit as a vulnerability, assigning it the at the time of writing reserved CVE-2023-23349, while the rest chose to disregard or underrate the issue.

Abstract
Being on a mushrooming spree since at least 2013, malware can take a large toll on any system. In a perpetual cat-and-mouse chase with defenders, malware writers constantly conjure new methods to hide their code so as to evade detection by security products. In this context, focusing on the MS Windows platform, this work contributes a comprehensive empirical evaluation regarding the detection capacity of popular, off-the-shelf antivirus and endpoint detection and response engines when facing legacy malware obfuscated via more or less uncommon but publicly known methods. Our experiments exploit a blend of seven traditional AV evasion techniques in 16 executables built in C++, Go, and Rust. Furthermore, we conduct an incipient study regarding the ability of the ChatGPT chatbot in assisting threat actors to produce ready-to-use malware. The derived results in terms of detection rate are highly unexpected: approximately half of the 12 tested AV engines were able to detect less than half of the malware variants, four AVs exactly half of the variants, while only two of the rest detected all but one of the variants.

Abstract
One of the most critical building blocks of the reliable operation of the Internet is the Border Gateway Protocol (BGP) that is used to exchange routing messages, signaling active and defective routing paths. During large-scale catastrophic incidents, such as conventional military operations or cyberwarfare, the stability of the Internet is affected, causing the announcements of defective routing paths to increase substantially. This work studies the relation between major incidents, such as armed conflicts in a country scale, and the corresponding network outages observed in the core of the Internet infrastructure as announced by BGP. We focus on the Russo-Ukrainian war as a timely and prominent use case and examine geolocalized BGP data for a 2-month period. Our methodology allows us to cherry-pick long-term network outages among temporary interruptions of service in this specific time window, and pinpoint them to the areas of the operations. Our results indicate that there is a high correlation between the start of military operations and network outages in a city and country level. Furthermore, we show that the last few days before the start of the operations network outages rise as well, indicating that preparatory cyberattack activities take place. No less important, network outages remain at much higher than usual levels during the operations, something that can be attributed to infrastructure destruction possibly backed by cyberattacks.

Abstract
The Transport Layer Security (TLS) protocol was introduced to solve the lack of security and privacy in the early versions of the world wide web. However, even though it has substantially evolved over the years, certain features still present privacy issues. One such feature is the Server Name Indication (SNI) extension, which allows multiple web servers to reside behind a provider hosting multiple domains with the same IP address; at the same time it allows third parties to discover the domains that end users visit. In the last few years, the Encrypted Server Name Indication (ESNI) Internet draft is being developed by the Internet Engineering Task Force (IETF); this encrypted variant of the extension was renamed to Encrypted Client Hello (ECH) in latest versions. In this paper, we measure the adoption of both these versions, given that they have substantial differences. By analyzing the top 1M domains in terms of popularity, we identify that only a small portion, less than 19%, supports the privacy-preserving ESNI extension and practically no domain supports ECH. Overall, these results demonstrate that there is still a long way to go to ensure the privacy of end users visiting TLS-protected domains which are co-located behind a common Internet-facing server.

Abstract
Emergency calling services are a cornerstone of public safety. During the last few years such systems are transitioning to VoIP and unified communications, and are continuously evolving under the umbrella of organizations, including NENA and EENA. The outcome of this effort is NG911 or NG112 services operating over the so-called Emergency Services IP network (ESInet). This work introduces and meticulously assesses the impact of an insidious and high-yield denial-of-service (DoS) attack against ESInet. Contrariwise to legacy SIP-based DoS, the introduced assault capitalizes on the SDP body of the SIP message with the sole purpose of instigating CPU-intensive transcoding operations at the ESInet side. We detail on the way such an attack can be carried out, and scrutinize on its severe, if not catastrophic, impact through different realistic scenarios involving a sufficient set of codecs. Finally, highlighting on the fact that 911 or 112 calls cannot be dropped, but need to be answered as fast as possible, we offer suggestions on how this kind of assault can be detected and mitigated.

Abstract
Over the last decade online services have penetrated the market and for many of us became an integral part of our software portfolio. On the one hand online services offer flexibility in every sector of the social web, but on the other hand these pros do not come without a cost in terms of privacy. This work focuses on online services, and in particular on the possible inherent design errors which make these services an easy target for privacy invaders. We demonstrate the previous fact using a handful of real-world cases pertaining to popular online web services. More specifically, we show that despite the progress made in raising security/privacy awareness amongst all the stakeholders (developers, admins, users) and the existence of mature security/privacy standards and practices, there still exist a plethora of poor implementations that may put user’s privacy at risk. We particularly concentrate on cases where a breach can happen even if the aggressor has limited knowledge about their target and/or the attack can be completed with limited resources. In this context, the main contribution of the paper at hand revolves around the demonstration of effortlessly exploiting privacy leaks existing in widely-known online services due to software development errors.

Abstract
Over the last decade, VoIP services and more especially the SIP-based ones, have gained much attention due to the low-cost and simple models they offer. Nevertheless, their inherently insecure design make them prone to a plethora of attacks. This work concentrates on the detection of resource consumption attacks targeting SIP ecosystems. While this topic has been addressed in the literature to a great extent, only a handful of works examine the potential of Machine Learning (ML) techniques to detect DoS and even fewer do so in realtime. Spurred by this fact, the work at hand assesses the potential of 5 different ML-driven methods in nipping SIP-powered DDoS attacks in the bud. Our experiments involving 17 realistically simulated (D)DoS scenarios of varied attack volume in terms of calls/sec and user population, suggest that some of the classifiers show promising detection accuracy even in low-rate DDoS incidents. We also show that the performance of ML-based detection in terms of classification time overhead does not exceed 3.5 ms in average with a mean standard deviation of 7.7 ms.

Abstract
Covert channels pose a significant threat for networking systems. In this paper, we examine the exploitation of Session Description Protocol (SDP) information residing in Session Initiation Protocol (SIP)requests with the aim to hide data in plain sight.While a significant mass of works in the literature cope with covert communication channels, only a very limited number of them rely on SIP to realize its goals. Also, none of them concentrates on SDP data contained in SIP messages to implement and evaluate such a hidden communication channel. Motivated by this fact, the work at hand proposes and demonstrates the feasibility of a simple but very effective in terms of stealthiness and simplicity SIP-based covert channel for botnet Command and Control (C&C). As a side contribution, we assess the soundness and the impact of such a deployment at the victim's side via the use of two different types of flooding attacks.

Abstract
This paper focuses on network anomaly-detection and especially the effectiveness of Machine Learning (ML) techniques in detecting Denial of Service (DoS) in SIP-based VoIP ecosystems. It is true that until now several works in the literature have been devoted to this topic, but only a small fraction of them have done so in an elaborate way. Even more, none of them takes into account high and low-rate Distributed DoS (DDoS) when assessing the efficacy of such techniques in SIP intrusion detection. To provide a more complete estimation of this potential, we conduct extensive experimentations involving 5 different classifiers and a plethora of realistically simulated attack scenarios representing a variety of (D)DoS incidents. Moreover, for DDoS ones, we compare our results with those produced by two other anomaly-based detection methods, namely Entropy and Hellinger Distance. Our results show that ML-powered detection scores a promising false alarm rate in the general case, and seems to outperform similar methods when it comes to DDoS.

Abstract
Attackers always find ways to elude the employed security mechanisms of a system, no matter how strong they are. Nevertheless, audit trails - which as a rule of thumb are kept by any service provider - store all the events pertaining to the service of interest. Therefore, audit trail data can be a valuable ally when it comes to the certification of the security level of a given service. This stands especially true for critical realtime services such as multimedia ones, which nowadays are on the rise. This work proposes a practical, simple to implement yet powerful solution based on the Hellinger Distance metric for conducting audit trail analysis destined to expose security incidents. Our solution relies on a set of different features existing in the app layer protocol for session handling in order to classify the analyzed traffic as intrusive or not. Taking the well-known Session Initiation Protocol (SIP) as an example, we thoroughly evaluate the effectiveness of the proposed detection scheme in terms of accuracy under various realistic scenarios. The outcomes reveal competitive detection rates in terms of false positives and negatives and can be used as a reference for future works in the field.

Abstract
Network audit trails, especially those composed of application layer data, can be a valuable source of information regarding the investigation of attack incidents. Nevertheless, the analysis of log files of large volume is usually both complex (slow) and privacy-neglecting. Especially, when it comes to VoIP, the literature on how audit trails can be exploited to identify attacks remains scarce. This paper provides an entropy-driven, privacy-preserving, and practical framework for detecting resource consumption attacks in VoIP ecosystems. We extensively evaluate our framework under various attack scenarios involving single and multiple assailants. The results obtained show that the proposed scheme is capable of identifying malicious traffic with a false positive alarm rate up to 3.5%.