In this report, we consider the problem of maximizing the throughput of Byzantine agreement. Byzantine agreement is a classical problem in distributed computing, with initial solutions presented in the seminal work of Pease, Shostak and Lamport [1, 2]. There has also been more recent work on designing multicast algorithms that can survive Byzantine attacks (e.g., [3]). Many variations on the By...

Distributed machine learning algorithms enable processing of datasets that are distributed over a network without gathering the data at a centralized location. While efficient distributed algorithms have been developed under the assumption of faultless networks, failures that can render these algorithms nonfunctional indeed happen in the real world. This paper focuses on the problem of Byzantin...

Permanent faults : This type of failure is persistent: it continues to exist until the faulty component is repaired or replaced. Examples of this fault are disk head crashes, software bugs, and burnt-out power supplies. Any of these faults may be either a fail-silent failure (also known as a fail-stop) or a Byzantine failure. A fail-silent fault is one where the faulty unit stops functioning an...

We propose a distributed spectrum decision protocol resilient to primary user emulation attacks (PUEA) in dynamic spectrum access (DSA) networks. PUEA is a type of denialof-service attack that can result in unreliable and/or disconnected DSA networks by depriving legitimate secondary users of spectrum access. We first propose an individual detection mechanism for secondary users to achieve prel...

Mobile ad-hoc networks (MANET) is a network mode that does not depend on infrastructure and central access. The fast flexible networking of MANET renders its wide applications in specific scenarios. However, rapidly changing topology open channels bring potential security problems. In this paper, we proposed an active-routing authentication scheme (AAS) based the characteristics active routing ...

Traditional Byzantine consensus does not work in P2P network due to Sybil attack while the most prevalent Sybil-proof consensus at present still can’t resist adversary with dominant compute power. This paper proposed a two-layered opinion dynamics model named SkyHash for hash consensus over P2P network. For hash consensus, failures are constrained due to the difficulty to create collision with ...

Asynchronous distributed machine learning solutions have proven very effective so far, but always assuming perfectly functioning workers. In practice, some of the workers can however exhibit Byzantine behavior, caused by hardware failures, software bugs, corrupt data, or even malicious attacks. We introduce Kardam, the first distributed asynchronous stochastic gradient descent (SGD) algorithm t...

International Computer Science Institute, Berkeley, CA 2007 – Present Postdoctoral Researcher, Networking Group, working with Prof. Scott Shenker, Dr. Petros Maniatis, and Dr. Sylvia Ratnasamy Diverse replication for single-machine Byzantine-fault tolerance: Investigate exploiting cores in many-core systems to defend against software attacks. Explore different isolation and software diversity m...

We propose a Byzantine Fault Tolerant variant of the Raft consensus algorithm, BFTRaft, inspired by the original Raft[1] algorithm and the Practical Byzantine Fault Tolerance algorithm[2]. BFT Raft maintains the safety, fault tolerance, and liveness properties of Raft in the presence of Byzantine faults, while also aiming towards to Raft’s goal of simplicity and understandability. We have imple...

Byzantine Agreements were first formally proposed by Lamport [8], for use in finding faulty processors in distrubuted computer networks. In this article we will use the process of finding faults to uncover traitors in secure communications networks. Maintaining reliability in trusted networks is an important application of Byzantine Agreements. Thus by introducing the fundamentals of the Byzant...