SRIKANTH, T. K., AND TOUEG, S. Simulating authenticated broadcasts to derive simple faulttolerant algorithms. Rep. 84-623, Computer Science Dept., Cornell Univ., Ithaca, N.Y.,  Home/Search   Document Not in Database   Summary   Related Articles   Check

....while maintaining the same requirement, N 3t 1, on the number of processors and maintaining 300 C. DWORK et al. polynomial time complexity and polynomial message lengths. The modification is done by using a minor variation of a broadcast primitive, introduced by Srikanth and Toueg [20], which simulates the crucial properties of authentication. We first state these properties, then give the broadcast primitive, and finally describe the new agreement protocol. The broadcast primitive (and hence the agreement algorithm that uses the broadcast primitive) is defined in terms of ....

....r, then every other correct processor accepts m from p in superround max(r 1, GST) or earlier. The description of the BROADCAST primitive is given in Figure 1. The proof that the primitive has the Correctness and Unforgeability properties is identical to the proof of Sfikanth and Toueg [20]. We give the proof of the Relay property since it is slightly different than in [20] Say the correct processor q accepts m from p in superround r. Therefore, q must have received (echo, p, m, k) from at least N t processors by the end of the second round of superround r, so at least N 2t ....

[Article contains additional citation context not shown here]

SRIKANTH, T. K., AND TOUEG, S. Simulating authenticated broadcasts to derive simple faulttolerant algorithms. Rep. 84-623, Computer Science Dept., Cornell Univ., Ithaca, N.Y.,

....to determine whether a message it receives was modified by processors that relay the message. These assumptions are satisfied if digital signatures are employed by the sender of a message or if there is a direct link between every pair or processors and the simulated authentication technique of [Srikanth Toueg 84] is used to transmit messages. In either case, faulty processors are unable to masquerade as correct processors by sending messages and are unable to modify messages sent originally by correct processors before retransmitting them. The following protocol implements a Fireworks Agreement with b=G ....

Srikanth, T.K. and S. Toueg. Simulating authenticated broadcasts to derive simple faulttolerant algorithms. Technical Report TR 84-623, Department of Computer Science, Cornell University, Ithaca, New York, July 1984.

....of this are quantified in the Appendix. On the other hand, SE1 has fault tolerance degree N 1 because it was not necessary to stipulate an upper bound on the number of faulty processors. A second characterization of sufficient evidence , first used in the clock synchronization proto col of [Srikanth Toueg 84] 12, is based on the fact that if every processor broadcasts a message when its clock reads T, then provided there are at most k faulty processors, the k l st message received must be from a correct one or must follow a message from a correct one. SE2: Receipt of k 1 messages originated by ....

Srikanth, T.K. and S. Toueg. Simulating authenticated broadcasts to derive simple fault-tolerant algorithms. Technical Report TR 84-623, Department of Computer Science, Cornell University, Ithaca, New York, July 1984.

....clock synchronization, which is beyond the scope of this thesis. 1. 2 Previous Work Different variants of the clock synchronization problem have been the target of a vast amount of research from both practical viewpoint (e.g. 26, 6, 24, 28, 1, 15] and theoretical viewpoint (e.g. [16, 19, 7, 13, 33, 3], surveys [31, 30] and references therein) the exact definition of the problem depends both on the intended use of the clocks and on the specific underlying system. The large number of variants is justified by the wide spectrum of applications. One of the popular variants studied theoretically ....

T. K. Srikanth and S. Toueg. Simulating authenticated broadcasts to derive simple fault-tolerant algorithms. Distributed Computing, 2:80--94, 1987.

....make such a mistake, as we described previously. One solution is to attach a session identifier, possibly the identity of the session initiator, to the sensor value. This solution will increase the size of each message by O(log n) bits. This does not exceed the lower bound by Srikanth and Toueg [28] because they already allocate O(log n) bits for signatures. Detecting replay attacks. Beside distinguishing concurrent sessions initiated by different processors, it is equally important to detect any attempt to reuse past messages (from the same initiator) in a new run. The initiator must ....

T.K. Srikanth and S. Toueg. Simulating authenticated broadcasts to derive simple fault-tolerant algorithms. Distributed Computing, 2(2):80--94, 1987.

....They can stop, omit to send or receive messages, or send spurious messages and falsely claim to have received messages they did not actually receive. Such failures are also called malicious or Byzantine. Other failures models were studied in the literature but are beyond the scope of this paper [22, 23]. Many researchers have developed translations for failures within this hierarchy. 1 Coan [7] considered systems with asynchronous message passing (also called asynchronous systems) and developed a compiler that converts algorithms tolerant of crash failures into ones that tolerate arbitrary ....

....algorithms tolerant of crash failures into ones that tolerate send omission failures; it was later shown that his translation is not general in that it cannot be applied to all algorithms [15] 2 Neiger and Toueg [18] gave a translation for those failure type that is general. Srikanth and Toueg [23] showed how algorithms that use message authentication to mitigate arbitrary failures can be transformed into ones that do not use message authentication. Neiger and Toueg considered synchronous systems and developed a family of translations; some translate from crash to general omission failures, ....

T. K. Srikanth and Sam Toueg. Simulating authenticated broadcasts to derive simple fault-tolerant algorithms. Distributed Computing, 2(2):80--94, 1987.

....sent by a processor is correct relative to the protocol and the processors state. In contrast, the messages sent by a faulty processor in the Byzantine model is in no way constrained. More recent work has provided mechanisms that restrict the behavior of faulty processors in the Byzantine model [13, 9, 1]. Still, these mechanisms require n 3t and lower bounds have been established [1] that show that such mechanisms multiply the running time of algorithms designed for the weaker models. 1.2 Our Results One contribution of this paper is the introduction of a new family of models of failure in ....

....q, we can require not only that there be a sufficient number T (usually equal to n Gamma t) of witnesses to q having sent v but also no more than b witnesses to any other value. Thus, a quorum of at least b 1 processors can veto v. This is in contrast to several existing protocols, e.g. [9, 13, 12], that only require a minimum number of positive witnesses. In order to prevent Byzantine processors from vetoing the actions of nonfaulty processors, it is necessary that a veto quorum include at least one nonByzantine processor. But, if we tried to implement the veto in the pure Byzantine model ....

[Article contains additional citation context not shown here]

T.K. Srikanth and S. Toueg. Simulating authenticated broadcasts to derive simple fault-tolerant algorithms. Distributed Computing, 2(2):80--94, 1987. This article was processed using the L a T E X macro package with LLNCS style

.... given to the interactive consistency problem, also called the Byzantine Generals Problem, 7] Requirements for the existence of a solution have been explored, 3] 4] 6] 9] and numerous algorithms for reaching agreement in the presence of malicious faults have been derived [1] 2] 6] 10] [12]. As Pease, Shostak, and Lamport in their original paper pointed out, an algorithm for reaching agreement need not reveal which units are faulty; it matters only that the faultfree units compute the same interactive consistency vector. Our objective is to explore the feasibility of identi cation ....

T.K.Srikanth and S.Toueg, Simulating authenticated broadcasts to derive simple fault-tolerant algorithms, Distributed Computing 1 (1987), 8094.

....fault tolerance of protocols which require message authentication only in certain rounds has been investigated in [Bor95] The protocols given there were maximally fault tolerant for a given number of authenticated rounds, but not very message efficient. A different approach has been taken in in [ST87]. There, authenticated messages are simulated by nonauthenticated subprotocols. This allows to transform authenticated protocols easily into non authenticated protocols while keeping some of their properties. But with this technique, the good fault tolerance properties of authenticated protocols ....

T.K. Srikanth and Sam Toueg. Simulating authenticated broadcasts to derive simple fault-tolerant algorithms. Distributed Computing, 2:80--94, 1987.

....algorithm is almost optimal if processors decide within a constant additive or multiplicative number of rounds of the decision time of an optimal algorithm. The possibility of such an algorithm in the multiplicative case is suggested by studies of translations between models of failures [1,2,4,15,18]. For example, Moses and Tuttle showed the existence of optimal algorithms that tolerate crash failures. Neiger and Toueg [15] showed how algorithms tolerant of crash failures could be converted to tolerate general omission failures by doubling the number of rounds used; local computation time was ....

T. K. Srikanth and Sam Toueg. Simulating authenticated broadcasts to derive simple fault-tolerant algorithms. Distributed Computing, 2(2):80--94, 1987.

....or true. One process g is distinguished, and has associated with it a boolean value B. It is required that: 1. If g is Reliable, the decision value of each Reliable process is B. 2. All Reliable processes eventually reach the same decision. Faults may make Reliable processes Unreliable. Program [12, 37] We assume authenticated communication: messages sent by Reliable processes are correctly received by Reliable processes, and Unreliable processes cannot forge messages on behalf of Reliable processes. Agreement is reached within N 1 rounds of communication, where N is the maximum number of ....

T. Srikanth and S. Toeug, "Simulating authenticated broadcast to derive simple fault tolerant algorithms", Distributed Computing, 2(2), pp. 80-94, 1987.

....message authentication, a mechanism that restricts the faulty behavior of processes that are subject to arbitrary failures. A precise definition of the properties of message authentication, and a mechanism for providing them without digital signatures in point to point networks appears in [ST87b] There are methods for automatically increasing the fault tolerance of algorithms. This is achieved by translations that transform any given algorithm tolerant of a certain type of failure into an algorithm that tolerates a more severe type of failure. Such translations are given in ....

....There are methods for automatically increasing the fault tolerance of algorithms. This is achieved by translations that transform any given algorithm tolerant of a certain type of failure into an algorithm that tolerates a more severe type of failure. Such translations are given in [Bra87,Coa87,ST87b,NT90,BN91,BN92] They can be used to transform any algorithm tolerant of crash failures into one tolerant of arbitrary failures, in both synchronous and asynchronous systems. Much of the theoretical work on fault tolerant algorithms focused on synchronous models, including the so called ....

T. K. Srikanth and Sam Toueg. Simulating authenticated broadcasts to derive simple fault-tolerant algorithms. Distributed Computing, 2(2):80--94, 1987.

No context found.

T. K. Srikanth and S. Toueg. Simulating authenticated broadcasts to derive simple fault-tolerant algorithms. Distributed Computitg1`2:80 941`1987.

No context found.

Srikanth, T. K., and S. Toueg, Simulating Authenticated Broadcasts to Derive Simple Fault Tolerant Algorithms. Distributed Computing , 2,2 (August 1987), 80--94. 5

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC