Elektrotechnik und Informatik (ETI)
Filtern
Publikationstyp
- Beitrag in einer Konferenzveröffentlichung (21) (entfernen)
Sprache
- Englisch (21) (entfernen)
Schlagworte
- QUIC (3)
- Cyber Security (2)
- Transport Protocol (2)
- Ack Ratio (1)
- Algorithm learning (1)
- Author Keywords: Industrial robot, Automatic robot programming, SME, low volume, high variant, path planning, matching, machine vision (1)
- IEEE Keywords: Solid modeling, Path planning, Three-dimensional displays, Robot kinematics, Pipelines, Task analysis (1)
- INET OMNeT++ (1)
- INSPEC (Controlled Indexing): cameras, collision avoidance, industrial robots, mobile robots, small-to-medium enterprises (1)
- INSPEC (Non-Controlled Indexing): high variant manufacturing, industrial robots, complex robot tasks, medium-sized enterprises, camera-based path, planning overhead, fast program generation, collision-free path, specific robot task, SME (1)
Fachbereich / Studiengang
A communication over an Internet Protocol (IP) based network fails if an endpoint sends packets that are too big to reach their destination and if the sender is unable to detect that. The node on the path that drops these packets should respond with a Packet Too Big (PTB) message. However, multiple scenarios exist in which the sender will not receive a PTB message. Even if it does, it refrains from using the information in case it suspects that a potential attacker forged the message. In particular, we are not aware of any implementation of the secure transport protocol QUIC (e.g., used by HTTP/3) that processes PTB messages. In this paper, we present a novel parameterizable PTB detection algorithm for reliable transport protocols that does not depend on PTB messages. We further describe how to integrate our algorithm into QUIC, present results from an evaluation using the algorithm within a QUIC simulation model and, based on these results, suggest concrete parameter values.
S/MIME and OpenPGP use cryptographic constructions repeatedly shown to be vulnerable to format oracle attacks in protocols like TLS, SSH, or IKE. However, format oracle attacks in the End-to-End Encryption (E2EE) email setting are considered impractical as victims would need to open many attacker-modified emails and communicate the decryption result to the attacker. But is this really the case?
In this paper, we survey how an attacker may remotely learn the decryption state in email E2EE. We analyze the interplay of MIME and IMAP and describe side-channels emerging from network patterns that leak the decryption status in Mail User Agents (MUAs). Concretely, we introduce specific MIME trees that produce decryption-dependent net work patterns when opened in a victim’s email client.
We survey 19 OpenPGP- and S/MIME-enabled email clients and four cryptographic libraries and uncover a side-channel leaking the decryption status of S/MIME messages in one client. Further, we discuss why the exploitation in the other clients is impractical and show that it is due to missing feature support and implementation quirks. These unintended defenses create an unfortunate conflict between usability and security. We present more rigid countermeasures for MUA developers and the standards to prevent exploitation.
OpenPGP is one of the two major standards for end-to-end email security. Several studies showed that serious usability issues exist with tools implementing this standard. However, a widespread assumption is that expert users can handle these tools and detect signature spoofing attacks. We present a user study investigating expert users' strategies to detect signature spoofing attacks in Thunderbird. We observed 25 expert users while they classified eight emails as either having a legitimate signature or not. Studying expert users explicitly gives us an upper bound of attack detection rates of all users dealing with PGP signatures. 52% of participants fell for at least one out of four signature spoofing attacks. Overall, participants did not have an established strategy for evaluating email signature legitimacy. We observed our participants apply 23 different types of checks when inspecting signed emails, but only 8 of these checks tended to be useful in identifying the spoofed or invalid signatures. In performing their checks, participants were frequently startled, confused, or annoyed with the user interface, which they found supported them little. All these results paint a clear picture: Even expert users struggle to verify email signatures, usability issues in email security are not limited to novice users, and developers may need proper guidance on implementing email signature GUIs correctly.
Quick UDP Internet Connections (QUIC) is a novel transport protocol introducing known features in a new protocol design. To investigate these features and the design, we developed a QUIC implementation in the INET simulation model suite.
In this paper, we describe that implementation, its validation and a result achieved using the simulation model. The result shows the negative impact on throughput, when raising the acknowledgment ratio. We propose a solution and describe how it solves the issue.
A data sender in an IP based network is only capable to efficiently use a network path if it knows the packet size limit of the path, i.e., the Path Maximum Transmission Unit (PMTU). The IETF recently specified a PMTU discovery framework for transport protocols like QUIC. This paper complements this specification by presenting a search algorithm. In addition, it defines several metrics and shows results of analyses for the algorithm with various PMTU candidate sequences using these metrics. We integrated the PMTU discovery with our algorithm into a QUIC simulation model. This paper describes the integration and presents measurements obtained by simulations.
Due to the increasing connectivity of modern vehicles, collected data is no longer only stored in the vehicle itself but also transmitted to car manufacturers and vehicle assistant apps. This development opens up new possibilities for digital forensics in criminal investigations involving modern vehicles. This paper deals with the digital forensic analysis of vehicle assistant apps of eight car manufacturers. We reconstruct the driver’s activities based on the data stored on the smartphones and in the manufacturer’s backend.
For this purpose, data of the Android and iOS apps of the car manufacturers Audi, BMW, Ford, Mercedes, Opel, Seat, Tesla, and Volkswagen were extracted from the smartphone and examined using digital forensic methods following forensics guidelines. Additionally, manufacturer data was retrieved using Subject Access Requests. Using the extensive data gathered, we reconstruct trips and refueling processes, determine parking positions and duration, and track the locking and unlocking of the vehicle.
Our findings show that the digital forensic investigation of smartphone applications is a useful addition to vehicle forensics and should therefore be taken into account in the strategic preparation of future digital forensic investigations.
Modern implantable cardiologic devices communicate via radio frequency techniques and nearby gateways to a backend server on the internet. Those implanted devices, gateways, and servers form an ecosystem of proprietary hardware and protocols that process sensitive medical data and is often vital for patients’ health.
This paper analyzes the security of this Ecosystem, from technical gateway aspects, via the programmer, to configure the implanted device, up to the processing of personal medical data from large cardiological device producers. Based on a real-world attacker model, we evaluated different devices and found several severe vulnerabilities. Furthermore, we could purchase a fully functional programmer for implantable cardiological devices, allowing us to re-program such devices or even induce electric shocks on untampered implanted devices.
Additionally, we sent several Art. 15 and Art. 20 GDPR inquiries to manufacturers of implantable cardiologic devices, revealing non-conforming processes and a lack of awareness about patients’ rights and companies’ obligations. This, and the fact that many vulnerabilities are still to be found after many vulnerability disclosures in recent years, present a worrying security state of the whole ecosystem.
TLS is one of today's most widely used and best-analyzed encryption technologies. However, for historical reasons, TLS for email protocols is often not used directly but negotiated via STARTTLS. This additional negotiation adds complexity and was prone to security vulnerabilities such as naive STARTTLS stripping or command injection attacks in the past.
We perform the first structured analysis of STARTTLS in SMTP, POP3, and IMAP and introduce EAST, a semi-automatic testing toolkit with more than 100 test cases covering a wide range of variants of STARTTLS stripping, command and response injections, tampering attacks, and UI spoofing attacks for email protocols. Our analysis focuses on the confidentiality and integrity of email submission (email client to SMTP server) and email retrieval (email client to POP3 or IMAP server). While some of our findings are also relevant for email transport (from one SMTP server to another), the security implications in email submission and retrieval are more critical because these connections involve not only individual email messages but also user credentials that allow access to a user's email archive.
We used EAST to analyze 28 email clients and 23 servers. In total, we reported over 40 STARTTLS issues, some of which allow mailbox spoofing, credential stealing, and even the hosting of HTTPS with a cross-protocol attack on IMAP. We conducted an Internet-wide scan for the particularly dangerous command injection attack and found that 320.000 email servers (2% of all email servers) are affected. Surprisingly, several clients were vulnerable to STARTTLS stripping attacks. In total, only 3 out of 28 clients did not show any STARTTLS-specific security issues. Even though the command injection attack received multiple CVEs in the past, EAST detected eight new instances of this problem. In total, only 7 out of 23 tested servers were never affected by this issue. We conclude that STARTTLS is error-prone to implement, under-specified in the standards, and should be avoided.
Background: Modern healthcare devices can be connected to computer networks and many western healthcareinstitutions run those devices in networks. At the same time, cyber attacks are on the rise and there is evidence thatcybercriminals do not spare critical infrastructure such as major hospitals, even if they endanger patients. Intuitively,the more and closer connected healthcare devices are to public networks, the higher the risk of getting attacked.
Methods: To asses the current connectivity status of healthcare devices, we surveyed the field of German hospitalsand especially University Medical Center UMCs.
Results: The results show a strong correlation between the networking degree and the number of medical devices.The average number of medical devices is 25.150, with a median of networked medical devices of 3.600. Actual keyusers of networked medical devices are the departments Radiology, Intensive Care, Radio-Oncology RO, NuclearMedicine NUC, and Anaesthesiology in the group of UMCs. In the next five years, the usage of networked medicaldevices will increase significantly in the departments of Surgery, Intensive Care, and Radiology. We detected a strongcorrelation between the degree of connectivity and the likelihood of being attacked.The survey answers regarding the cyber security status reveal a lack of security basics in some of the inquiredhospitals. We did discover successful attacks in hospitals with separated or subsidiary departments. A fusion ofcompetencies on an organizational level facilitates the right behavior here. Most hospitals rated themselvespredominantly positively in the self-assessment but also stated the usefulness of IT security insurance.Conclusions:Concluding our results, hospitals are already facing the consequences of omitted measures within theirgrowing pool of medical devices. Continuously relying on historically grown structures without adaption and trustingmanufactures to solve vectors is a critical behavior that could seriously endanger patients.
OOXML and ODF are the de facto standard data formats for word processing, spreadsheets, and presentations. Both are XML-based, feature-rich container formats dating back to the early 2000s. In this work, we present a systematic analysis of the capabilities of malicious office documents. Instead of focusing on implementation bugs, we abuse legitimate features of the OOXML and ODF specifications. We categorize our attacks into five classes: (1) Denial-of-Service attacks affecting the host on which the document is processed. (2) Invasion of privacy attacks that track the usage of the document. (3) Information disclosure attacks exfiltrating personal data out of the victim's computer. (4) Data manipulation on the victim's system. (5) Code execution on the victim's machine. We evaluated the reference implementations – Microsoft Office and LibreOffice – and found both of them to be vulnerable to each tested class of attacks. Finally, we propose mitigation strategies to counter these attacks.
Smart wearable devices become more and more prevalent in the age of the Internet of Things. While people wear them as fitness trackers or full-fledged smartphones, they also come in unique versions as smartwatches for children. These watches allow parents to track the location of their children in real-time and offer a communication channel between parent and child.
In this paper, we analyzed six smartwatches for children and the corresponding backend platforms and applications for security and privacy concerns. We structure our analysis in distinct attacker scenarios and collect and describe related literature outside academic publications. Using a cellular network Man-in-the-Middle setup, reverse engineering, and dynamic analysis, we found several severe security issues, allowing for sensitive data disclosure, complete watch takeover, and illegal remote monitoring functionality.
Vulnerabilities in private networks are difficult to detect for attackers outside of the network. While there are known methods for port scanning internal hosts that work by luring unwitting internal users to an external web page that hosts malicious JavaScript code, no such method for detailed and precise service identification is known. The reason is that the Same Origin Policy (SOP) prevents access to HTTP responses of other origins by default. We perform a structured analysis of loopholes in the SOP that can be used to identify web applications across network boundaries. For this, we analyze HTML5, CSS, and JavaScript features of standard-compliant web browsers that may leak sensitive information about cross-origin content. The results reveal several novel techniques, including leaking JavaScript function names or styles of cross-origin requests that are available in all common browsers. We implement and test these techniques in a tool called CORSICA. It can successfully identify 31 of 42 (74%) of web services running on different IoT devices as well as the version numbers of the four most widely used content management systems WordPress, Drupal, Joomla, and TYPO3. CORSICA can also determine the patch level on average down to three versions (WordPress), six versions (Drupal), two versions (Joomla), and four versions (TYPO3) with only ten requests on average. Furthermore, CORSICA is able to identify 48 WordPress plugins containing 65 vulnerabilities. Finally, we analyze mitigation strategies and show that the proposed but not yet implemented strategies Cross-Origin Resource Policy (CORP)} and Sec-Metadata would prevent our identification techniques.
The Portable Document Format, better known as PDF, is one of themost widely used document formats worldwide, and in order to en-sure information confidentiality, this file format supports documentencryption. In this paper, we analyze PDF encryption and showtwo novel techniques for breaking the confidentiality of encrypteddocuments. First, we abuse the PDF feature ofpartially encrypteddocuments to wrap the encrypted part of the document withinattacker-controlled content and therefore, exfiltrate the plaintextonce the document is opened by a legitimate user. Second, we abusea flaw in the PDF encryption specification to arbitrarily manipulateencrypted content. The only requirement is that a single block ofknown plaintext is needed, and we show that this is fulfilled bydesign. Our attacks allow the recovery of the entire plaintext of en-crypted documents by using exfiltration channels which are basedon standard compliant PDF properties.We evaluated our attacks on 27 widely used PDF viewers andfound all of them to be vulnerable. We responsibly disclosed thevulnerabilities and supported the vendors in fixing the issue
Teaching People to program is a crucial requirement for our society to deal with the complexity of 21st-century challenges. In many teaching systems, the student is required to use a particular programming language or development environment. This paper presents an intelligent tutoring system to support blended learning scenarios, where the students can choose their programming language and development environment. For that, the system provides an interface where the students request test data and submit results to unit test their algorithms. The submitted results are analyzed by a machine learning system that detects common errors and provides adaptive feedback to the student. With this system, we are focusing on teaching algorithms rather than specific programming language semantics. The technical evaluation tested with the implementation of Mean and Median algorithm shows that the system can distinguish between error cases with an error rate under 20%. A first survey, with a small group of students, shows that the system helps them detect common errors and arrive at a correct/valid solution. We are in the process of testing the system with a larger group of students for gathering statistically reliable data.
Camera based path planning for low quantity - high variant manufacturing with industrial robots
(2019)
The acquisition costs for industrial robots have been steadily decreasing in past years. Nevertheless, they still face significant drawbacks in the required effort for the preparation of complex robot tasks which causes these systems to be rarely present so far in small and medium-sized enterprises (SME) that focus mainly on small volume, high variant manufacturing. In this paper, we propose a camera-based path planning framework that allows the fast preparation and execution of robot tasks in dynamic environments which leads to less planning overhead, fast program generation and reduced cost and hence overcomes the major impediments for the usage of industrial robots for automation in SMEs with focus on low volume and high variant manufacturing. The framework resolves existing problems in different steps. The exact position and orientation of the workpiece are determined from a 3D environment model scanned by an optical sensor. The so retrieved information is used to plan a collision-free path that meets the boundary conditions of the specific robot task. Experiments show the potential and effectiveness of the the framework presented here by evaluating a case study.
Fast Constant Time Memory Allocator for Inter Task Communication in Ultra Low Energy Embedded Systems