Hybrid Fault Models Research Articles

Asphalion: trustworthy shielding against Byzantine faults

Byzantine fault-tolerant state-machine replication (BFT-SMR) is a technique for hardening systems to tolerate arbitrary faults. Although robust, BFT-SMR protocols are very costly in terms of the number of required replicas (3f+1 to tolerate f faults) and of exchanged messages. However, with "hybrid" architectures, where "normal" components trust some "special" components to provide properties in a trustworthy manner, the cost of using BFT can be dramatically reduced. Unfortunately, even though such hybridization techniques decrease the message/time/space complexity of BFT protocols, they also increase their structural complexity. Therefore, we introduce Asphalion, the first theorem prover-based framework for verifying implementations of hybrid systems and protocols. It relies on three novel languages: (1) HyLoE: a Hybrid Logic of Events to reason about hybrid fault models; (2) MoC: a Monadic Component language to implement systems as collections of interacting hybrid components; and (3) LoCK: a sound Logic of events-based Calculus of Knowledge to reason about both homogeneous and hybrid systems at a high-level of abstraction (thereby allowing reusing proofs, and capturing the high-level logic of distributed systems). In addition, Asphalion supports compositional reasoning, e.g., through mechanisms to lift properties about trusted-trustworthy components, to the level of the distributed systems they are integrated in. As a case study, we have verified crucial safety properties (e.g., agreement) of several implementations of hybrid protocols.

Towards computational models of animal communications, an introduction for computer scientists

Animals communicate with each other via varieties of signal modalities such as visual, acoustic, semiochemicals, gestural, or communication behavior. Although animal communication has been studied as a signaler-receiver dyadic system until recently, it is actually a complex network due to the ubiquitous existence of third-party receivers in their signaling space, such as eavesdroppers, bystanders, and audiences. A question of fundamental importance in both science and humanity is whether or not animal signals are honest or deceitful (i.e., communication reliability) since it may shed light on the study of trust and altruism in human beings. The objective of this paper is twofold. First, the state-of-the-art research in animal communication (dyadic paradigm) and animal communication networks (network paradigm) and their potential inspiration for secure, resilient and pervasive computing are reviewed. Secondly, by arguing that competition, cooperation and communication are the three fundamental elements that natural selection acts upon in the evolution of organisms, and by comparing the study of cooperation (such as formulated as the Prisoner’s Dilemma (PD) game), with the study of animal communication (such as formulated with Sir Philip Sydney (SPS) game), it is suggested that the latter field may spawn important cross-disciplinary research with the potential influence similar to or beyond that generated by the PD game research. The paper also summarizes the recent extension to evolutionary game theory (EGT) with survival analysis and dynamic hybrid fault models by the author. Finally, seven open questions, with regard to animal communication inspired computing and communication, primarily from the interdisciplinary perspective are posed. As a side note, this is the second article of a two-part series in which I review the state-of-the-art research in behavior biology inspired computing and communication. Given the extensive scope of behavior biology, I only focus on two fields most relevant to computing and communication animal cognition (part one, Ma, 2014) and communication (this article). The previous part one reviewed animal-cognition inspired computing; part one also presented an overview of the behavior biology inspired computing. This part two is focused on animal communication; the relationship between cognition and communication is mainly discussed in this part also.

Dynamic Hybrid Fault Modeling and Extended Evolutionary Game Theory for Reliability, Survivability and Fault Tolerance Analyses

We introduce a new layered modeling architecture consisting of dynamic hybrid fault modeling and extended evolutionary game theory for reliability, survivability, and fault tolerance analyses. The architecture extends traditional hybrid fault models and their relevant constraints in the Agreement algorithms with survival analysis, and evolutionary game theory. The dynamic hybrid fault modeling (i) transforms hybrid fault models into time- and covariate-dependent models; (ii) makes real-time prediction of reliability more realistic, and allows for real-time prediction of fault-tolerance; (iii) sets the foundation for integrating hybrid fault models with reliability and survivability analyses by integrating them with evolutionary game modeling; and (iv) extends evolutionary game theory by stochastically modeling the survival (or fitness) and behavior of `game players.' To analyse survivability, we extend dynamic hybrid fault modeling with a third-layer, operational level modeling, to develop the three-layer survivability analysis approach (dynamic hybrid fault modeling constitutes the tactical and strategic levels). From the perspective of evolutionary game modeling, the two mathematical fields, i.e., survival analysis and agreement algorithms, which we applied for developing dynamic hybrid fault modeling, can also be utilized to extend the power of evolutionary game theory in modeling complex engineering, biological (ecological), and social systems. Indeed, a common property of the areas where our extensions to evolutionary game theory can be advantageous is that the risk analysis and management are a core issue. Survival analysis (including competing risks analysis, and multivariate survival analysis) offers powerful modeling tools to analyse time-, space-, and/or covariate-dependent uncertainty, vulnerability, and/or frailty which `game players' may experience. The agreement algorithms, which are not limited to the agreement algorithms from distributed computing, when applied to extend evolutionary game modeling, can be any problem (game system) specific rules (algorithms or models) that can be utilized to dynamically check the consensus among game players. We expect that the modeling architecture and approaches discussed in the study should be implemented as a software environment to deal with the necessary sophistication. Evolutionary computing should be particularly convenient to serve as the core optimization engine, and should simplify the implementation. Accordingly, a brief discussion on the software architecture is presented.

Frailty modelling for risk analysis in network security and survivability

The common usage of the term frailty, which literally means the condition (quality) of being frail or a fault, may be too ambiguous to be very useful for characterising network security. Nevertheless, the mathematical concept of frailty, which originated in the study of aging and demography, may offer a powerful approach for analysing the risks in network security and survivability. This is because network security analysis is essentially a risk analysis problem, and identifying risks, especially those that are uncertain, latent, unobserved or unobservable (UUUR) (Ma, 2008), and further analysing the potential dependence between those risks are two critical challenges for risk analysis in any field. Frailty modelling excels in addressing these two challenges. When it is integrated with dynamic hybrid fault models (DHF) and extended evolutionary game theory (EEGT) (Ma and Krings, 2011), frailty modelling should provide powerful approaches for analysing risks in network security and survivability.

Exploiting omissive faults in synchronous approximate agreement

In a fault-tolerant distributed system, it is often necessary for nonfaulty processes to agree on the value of a shared data item. The criterion of Approximate Agreement does not require processes to achieve exact agreement on a value; rather, they need only agree to within a predefined numerical tolerance. Approximate Agreement can be achieved through convergent voting algorithms. Previous research has studied convergent voting algorithms under mixed-mode or hybrid fault models, such as the Thambidurai and Park Hybrid fault model, comprised of three fault modes: asymmetric, symmetric, and benign. This paper makes three major contributions to the state of the art in fault-tolerant convergent voting. (1) We partition both the asymmetric and symmetric fault modes into disjoint omissive and transmissive submodes. The resulting five-mode hybrid fault model is a superset of previous hybrid fault models. (2) We present a new family of voting algorithms, called Omission Mean Subsequence Reduced (OMSR), which implicitly recognize and exploit omissive behavior in malicious faults while still maintaining full Byzantine fault tolerance; (3) We show that OMSR voting algorithms are more fault-tolerant than previous voting algorithms if any of the currently active faults is omissive.

New hybrid fault models for asynchronous approximate agreement

An important problem in fault-tolerant distributed systems is maintaining agreement between nonfaulty processes in the presence of undiagnosed faults. To achieve agreement, processes exchange their local "opinions" of a particular value, and then vote on the values received to arrive at a "consensus". Approximate agreement defines a condition in which it is not necessary for consensus values to be identical. Rather, it is only necessary that they agree to within a predefined tolerance. Approximate agreement can be achieved through a sequence of convergent voting rounds, in which the range of values held by nonfaulty nodes is reduced in each round. Research has revealed simple expressions for the convergence rate and fault tolerance of a broad family of convergent voting algorithms called Mean-Subsequence-Reduced (MSR) algorithms. These results were derived under the Thambidurai and Park (1988) hybrid fault model comprised of asymmetric, symmetric, and benign faults. However, these results apply only to synchronous systems, in which there is a known finite bound on computation and communications times. We extend the previous results to asynchronous systems, in which no such bound exists. In addition, we introduce two new hybrid fault models which further differentiate between omissive faults and transmissive faults. The new fault models permit tighter bounds on the fault-tolerance of asynchronous systems to be derived.