With the end of Moore's Law in sight, parallelism became the main means for speeding up computationally intensive applications, especially in the cases where large collections of tasks need to be performed. Network supercomputing -- taking advantage of very large numbers of computers in a distributed environment is an effective approach to massive parallelism that harnesses the processing power inherent in large networked settings. In such settings, processor failures are no longer an exception, but the norm. Any algorithm designed for realistic settings must be able to deal with failures. This paper presents a new message-passing algorithm for distributed cooperative work in synchronous settings where processors may crash, and where any broadcasts performed by crashing processors are unreliable. We specify the algorithm, prove that it is correct, and perform extensive simulations that show that its performance is close to similar algorithms that use reliable broadcast, and that its work compares favorably to the relevant lower bounds.

Internet supercomputing is an approach to solving partitionable, computation-intensive problems by harnessing the power of a vast number of interconnected computers. For the problem of using network supercomputing to perform a large collection of independent tasks, prior work introduced a decentralized approach and provided randomized synchronous algorithms that perform all tasks correctly with high probability, while dealing with misbehaving or crash-prone processors. The main weaknesses of existing algorithms is that they assume either that the average probability of a non-crashed processor returning incorrect results is inferior to 12, or that the probability of returning incorrect results is known to each processor. Here we present a randomized synchronous distributed algorithm that tightly estimates the probability of each processor returning correct results. Starting with the set P of n processors, let F be the set of processors that crash. Our algorithm estimates the probability pi of returning a correct result for each processor i ∈ P - F, making the estimates available to all these processors. The estimation is based on the (ε, δ)-approximation, where each estimated probability p̃iof piobeys the bound Pr[pi(1 - ε) ≤ p̃i≤ pi(1 + ε)] > 1 - δ, for any constants δ > 0 and ε > 0 chosen by the user. An important aspect of this algorithm is that each processor terminates without global coordination. We assess the efficiency of the algorithm in three adversarial models as follows. For the model where the number of non-crashed processors P - F is linearly bounded the time complexity T (n) of the algorithm is O(log n), work complexity W(n) is O(n log n), and message complexity M(n) is O(n log2n). For the model where P - F is bounded by a fractional polynomial we have T(n) = O(n1-alog n log log n), W(n) = O(n log n log log n), and M(n) = O(n log2n log log n). For the model where P - F is bounded by a poly-logarithm we have T(n) = O(n), W(n) = O(n poly log n), and M(n) = O(n log2n poly log n). All bounds are shown to hold with high probability.

Distributed cooperative computing in networks involves marshaling collections of network nodes possessing the necessary computational resources. Before the willing nodes can act in a concerted way they must first discover one another. This is the general setting of the Resource Discovery Problem (RDP). This paper presents a self-stabilizing algorithm that solves RDP in a deterministic synchronous setting. The solution approach is formulated in terms of evolving knowledge graphs, where vertices represent the participating network nodes, and edges represent one node's knowledge about another. Ideally, the diameter of such a graph is one, i.e., each node knows all others. The algorithm works in rounds as it evolves the knowledge graph with the goal of reducing its diameter. This is accomplished by nodes sharing their knowledge through gossip messages. We prove that the algorithm is self-stabilizing, i.e., it tolerates arbitrary perturbations in the nodes' local states and is guaranteed to solve the problem once such failures subside. The algorithm has stabilization time of O(D), and it takes at most 4D + 4 complete round to stabilize, where D is the diameter of the initial knowledge graph, and the corresponding message complexity is O,(|V|j ⋅D), where V is the set of participating nodes.

Internet supercomputing is an approach to solving partitionable, computation-intensive problems by harnessing the power of a vast number of interconnected computers. For the problem of using network supercomputing to perform a large collection of independent tasks, our prior work introduced the decentralized approach, and provided a synchronous algorithm that is able to perform all tasks with high probability (whp), while dealing with malicious behaviors under a rather strong assumption that the average probability of live (non-crashed) processors returning bogus results remains inferior to 1/2 during the computation. There the adversary is severely limited in its ability to crash processors that normally return correct results. This work develops an efficient synchronous decentralized algorithm that is able to deal with a much stronger adversary. We consider a failure model with crashes, where given the initial set of processors P, an adversary is able to crash any subset F of processors, where |F| ≤ f•n, for a constant f (0<f<1), under the constraint that there exists a subset H ⊆ P - F, with |H| = Ω(n), called the hardened set, such that the average probability of a processor in H returning a bogus result is inferior to 1/2. Here any processor may return bogus results, and H may be much smaller than P-F, while the average probability of processors in P-F returning a bogus result may be greater than 1/2. We develop an efficient randomized algorithm for n processors and t tasks (n≤t), where each live processor is able to determine locally when all tasks are performed, and obtain the results of all tasks. We prove that in Θ(t⁄n logn) rounds all live workers know the results of all tasks whp, and that these results are correct whp. The work complexity of the algorithm is Θ(tlogn), the message complexity is Θ(nlogn ), and the bit complexity is O(tn log3n).

Traditional approaches to network supercomputing employ a master process and a large number of potentially undependable worker processes that must perform a collection of tasks on behalf of the master. In such a centralized scheme, the master process is a performance bottleneck and a single point of failure. This work develops an original approach that eliminates the master and instead uses a decentralized algorithm, where each worker is able to determine locally that all tasks have been performed, and to collect locally the results of all tasks. The failure model assumes that the average probability of a worker returning a wrong result is inferior to 1/2. A randomized synchronous algorithm for n processes and n tasks is presented. The algorithm terminates in Θ(log n) rounds, and it is proved that upon termination the workers know the results of all tasks with high probability, and that these results are correct with high probability. The message complexity of the algorithm is Θ(n log n), and the bit complexity is O(n2 log3 n).

In recent years, two distinct electronic voting technologies have been introduced and extensively utilized in election procedures: direct recording electronic systems and optical scan (OS) systems. The latter are typically deemed safer, as they inherently provide a voter-verifiable paper trail that enables hand-counted audits and recounts that rely on direct voter input. For this reason, OS machines have been widely deployed in the United States. Despite the growing popularity of these machines, they are known to suffer from various security vulnerabilities that, if left unchecked, can compromise the integrity of elections in which the machines are used. This article studies general auditing procedures designed to enhance the integrity of elections conducted with optical scan equipment and, additionally, describes the specific auditing procedures currently in place in the State of Connecticut. We present an abstract view of a typical OS voting technology and its relationship to the general election process. With this in place, we lay down a ldquotemporal-resourcerdquo adversarial model, providing a simple language for describing the disruptive power of a potential adversary. Finally, we identify how audit procedures, injected at various critical stages before, during, and after an election, can frustrate such adversarial interference and so contribute to election integrity. We present the implementation of such auditing procedures for elections in the State of Connecticut utilizing the Premiere (Diebold) AccuVote OS; these audits were conducted by the UConn VoTeR Center, at the University of Connecticut, on request of the Office of the Secretary of the State. We discuss the effectiveness of such procedures in every stage of the process and we present results and observations gathered from the analysis of past election data.

Special purpose trusted computing devices are currently being deployed to offer many services for which the general purpose computing paradigm is unsuitable. The nature of the services offered by many of these devices demand high security and reliability, as well as low cost and low power consumption. Electronic Voting machines is a canonical example of this phenomenon. With electronic voting machines currently being used in much of the United States and several other countries, there is a strong need for thorough security evaluation of these devices and the procedures in place for their use. In this work, we first put forth a general framework for special purpose trusted computing devices. We then focus on Optical Scan (OS) electronic voting technology as a specific instance of this framework. OS terminals are a popular e-voting technology with the decided advantage of a user-verified paper trail: the ballot sheets themselves. Still election results are based on machine- generated totals as well as machine-generated audit reports to validate the voting process. In this paper we present a security assessment of the Diebold AccuVote Optical Scan voting terminal (AV-OS), a popular OS terminal currently in wide deployment anticipating the 2008 Presidential elections. The assessment is developed using exclusively reverse-engineering, without any technical specifications provided by the machine suppliers. We demonstrate a number of security issues that relate to the machine's proprietary language, called AccuBasic, that is used for reporting election results. While this language is thought to be benign, especially given that it is essentially sandboxed by the firmware to have only read access, we demonstrate that it is powerful enough to (i) strengthen known attacks against the AV-OS so that they become undetectable prior to elections (and thus significantly increasing their magnitude) or, (ii) to conditionally bias the election results to reach a desired outcome. Given the discovered vulnerabilities and attacks we proceed to discuss how random audits can be used to validate with high confidence that a procedure carried out by special purpose devices such as the AV-OS has not been manipulated. We end with a set of recommendations for the design and safe-use of OS voting systems.