Federated learning is a privacy-enforcing machine learning technology but suffers from limited scalability. This limitation mostly originates from the internet connection and memory capacity of the central parameter server, and the complexity of the model aggregation function. Decentralized learning has recently been emerging as a promising alternative to federated learning. This novel technology eliminates the need for a central parameter server by decentralizing the model aggregation across all participating nodes. Numerous studies have been conducted on improving the resilience of federated learning against poisoning and Sybil attacks, whereas the resilience of decentralized learning remains largely unstudied. This research gap serves as the main motivator for this study, in which our objective is to improve the Sybil poisoning resilience of decentralized learning.

We present SybilWall, an innovative algorithm focused on increasing the resilience of decentralized learning against targeted Sybil poisoning attacks. By combining a Sybil-resistant aggregation function based on similarity between Sybils with a novel probabilistic gossiping mechanism, we establish a new benchmark for scalable, Sybil-resilient decentralized learning.

A comprehensive empirical evaluation demonstrated that SybilWall outperforms existing state-of-the-art solutions designed for federated learning scenarios and is the only algorithm to obtain consistent accuracy over a range of adversarial attack scenarios. We also found SybilWall to diminish the utility of creating many Sybils, as our evaluations demonstrate a higher success rate among adversaries employing fewer Sybils. Finally, we suggest a number of possible improvements to SybilWall and highlight promising future research directions.

An essential step of software development is obtaining an understanding of the behaviour of a system. Accurate state models of system behaviour might help software developers build such an understanding. There exist several techniques for automatically inferring models on system behaviour using log analysis, but these do not scale well for systems that produce large amounts of logs.
In this study, we present an approach for inferring concise state models of system behaviour using log analysis, called MASM (Markov Algorithm for inferring concise State Models), which implements a Markov chain. We argue this approach has the potential for higher scalability than existing techniques, as it attempts to approximate a state model by exploiting the properties of Markov chains. This is achieved by considering a sequence of log statements as a stochastic process, which enables MASM to infer a naive state model from a set of log statements, which is then minimized using the properties of Markov chains.
We evaluated MASM in an empirical study on the XRP Ledger Consensus Protocol. In this empirical study, MASM was evaluated on accuracy at different compression rates, and scalability. The results indicate that MASM shows several knee points in terms of accuracy, suggesting several optimal compression rates. We also found that the run-time of MASM scales linearly with the size of the dataset. Finally, we discovered that MASM is unable to outperform a random compression algorithm, which compresses the model by randomly merging states, in terms of accuracy. However, further research is required on measuring the readability of the produced models to derive definitive conclusions on the performance of MASM compared to a random compression algorithm.