| Elena Dubrova and Harald Sack. Probabilistic verification of multiple-valued functions. In Proc. Int. Symp. on Multiple-Valued Logic, Portland, Oregon, 2000. |
..... By we denote the support set of a Boolean function , which is the set of variables on which the function actually depends, i.e. 2. 2 Basic idea of the probabilistic method The probabilistic method based on arithmetic transform has been developed in [11] for Boolean functions an extended in [12] to the multiple valued case , positive integer greater than 2. In both cases, the combinational logic circuits and to be compared are transformed to the integer valued polynomials and . These polynomials are functions of type over a finite field of integers with being a prime. To compute hash ....
....with a particular input variable , If represents the value of in a given row of the truth table, then the corresponding term in the key polynomial is defined as . Parameter acts as a selector between and . For a more detailed description of arithmetic transform the reader is referred to [12]. 2.3 An new algorithm for computing hash codes Polynomials are not efficient as a data structure. Often, it takes more memory to store than to store . Therefore, we would like to avoid computing and storing a complete representation of . Instead, we would rather derive the polynomial only for ....
E. Dubrova, H.Sack, "Probabilistic Verification of Multiple-Valued Functions", 30th International Symposium on Multiple-Valued Logic, pp. 461-466, 2000.
....functionality. DSP functions are verified today by using the functional models as golden models and comparing their output with the output of RTL model. Because of the long simulation runs, this simulation based verification has very low coverage. In contrast, the probabilistic verification method [13, 14] is very efficient allowing a much larger coverage. 4. Novelty The novelty of the project is in combining different verification technologies to address different aspects of design and in different phases depending on their strengths and suitability. 5. Feasibility The ESDlab group at KTH has ....
E. Dubrova, H. Sack, "Probabilistic Verification of Multiple-Valued Functions", Proc. Int. Symp. om Multiple -Valued Logic, May 23-25 2000, Portland, Oregon.
....problem. In this section we show that the equivalence of Mod p DDs can be decided probabilistically in linear time, by extending the probabilistic equivalence test for Parity OBDDs [8] to the multiple valued case. Our extension employs the concept of multiple valued signatures introduced in [9] for identifying the equivalence of two multiple valued functions probabilistically. In the Boolean case, Parity OBDDs are a special case of OBDDs, allowing so called functional nodes labeled by an element of a basis of binary Boolean functions. It was shown that while the equivalence test for ....
....the probabilistic equivalence test for ParityOBDDs [8] to the multiple valued case. The equivalence of two Mod p DDs is determined by an algebraic transformation of the Mod p DDs in terms of polynomials over a finite field of integers modulo p. This algebraic transformation was introduced in [9]. Let GF (p k ) be a Galois Field with p k elements of characteristic p, p prime, k 0. Definition 5 Let P be a Mod p DD representing a multiple valued function f : M n M . With each node v 2 P we associate the polynomial p v : GF ( p k ) n GF (p k ) defined in the ....
[Article contains additional citation context not shown here]
E. Dubrova, H. Sack, Probabilistic verification of multiple-valued functions, Tech. Report 99-23, University of Trier, FB IV-Informatik, Nov. 1999.
No context found.
Elena Dubrova and Harald Sack. Probabilistic verification of multiple-valued functions. In Proc. Int. Symp. on Multiple-Valued Logic, Portland, Oregon, 2000.
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