Joachim Posegga

Talaya Farasat, Joachim Posegga

Software vulnerabilities are a fundamental reason for the prevalence of cyber attacks and their identification is a crucial yet challenging problem in cyber security. In this paper, we apply and compare different machine learning algorithms for source code vulnerability detection specifically for Python programming language. Our experimental evaluation demonstrates that our Bidirectional Long Short-Term Memory (BiLSTM) model achieves a remarkable performance (average Accuracy = 98.6%, average F-Score = 94.7%, average Precision = 96.2%, average Recall = 93.3%, average ROC = 99.3%), thereby, establishing a new benchmark for vulnerability detection in Python source code.

Korbinian Spielvogel, Henrich Pöhls, Joachim Posegga

Message queuing brokers are a fundamental building block of the Internet of Things, commonly used to store and forward messages from publishing clients to subscribing clients. Often a single trusted broker offers secured (e.g. TLS) and unsecured connections but relays messages regardless of their inbound and outbound protection. Such mixed mode is facilitated for the sake of efficiency since TLS is quite a burden for MQTT implementations on class-0 IoT devices. Such a broker thus transparently interconnects securely and insecurely connected devices; we argue that such mixed mode operation can actually be a significant security problem: Clients can only control the security level of their own connection to the broker, but they cannot enforce any protection towards other clients. We describe an enhancement of such a publish/subscribe mechanism to allow for enforcing specified security levels of publishers or subscribers by only forwarding messages via connections which satisfy the desired security levels. For example, a client publishing a message over a secured channel can instruct the broker to forward the message exclusively to subscribers that are securely connected. We prototypically implemented our solution for the MQTT protocol and provide detailed overhead measurements.