The transformation technique relies on the comparison of parity values computed in two ways. The fault detection structures are developed and they not only detected subsystem faults but also corrected faults introduced in the data processing system. Concurrent parity va More
The transformation technique relies on the comparison of parity values computed in two ways. The fault detection structures are developed and they not only detected subsystem faults but also corrected faults introduced in the data processing system. Concurrent parity values techniques are very useful in detecting numerical error in the data processing operations, where a single fault can propagate to many output faults. Parity values are the most effective tools used to detect faults occurring in the code stream. In this paper, we present a methodology for redundant systems that allows to detect faults. Checkpointing is the typical technique to tolerate such faults. This paper presents a checkpointing approach to operate on encoded data. The advantage of this method is that it is able to achieve very low overhead according to the specific characteristic of an application. The numerical results of the multiple checkpointing technique confirm that the technique is more efficient and reliable by not only distributing the process of checkpointing over groups of processors. This technique has been shown to improve both the reliability of the computation and the performance of the checkpointing.
Manuscript profile
In this paper, we aim to encode the set of all positive integers so that the codewords not only be uniquely decodable but also be an instantaneous set of binary sequences. Elias introduces three recursive algorithms for universal coding of positive integers where each c More
In this paper, we aim to encode the set of all positive integers so that the codewords not only be uniquely decodable but also be an instantaneous set of binary sequences. Elias introduces three recursive algorithms for universal coding of positive integers where each codeword contains binary representation of the integer plus an attachment portion that gives some information about the first part [1]. On the other hand, Fibonacci coding which is based on Fibonacci numbers is also introduced by Apostolico and Fraenkel for coding of integers [2]. In this paper, we propose a new lossless recursive algorithm for universal coding of positive integers based on both recursive algorithms and Fibonacci coding scheme without using any knowledge about the source statistics [3].The coding schemes which don’t use the source statistics is called universal coding, in these universal coding schemes we should use a universal decoding scheme in the receiver side of communication system. All of these encoding and decoding schemes assign binary streams to positive integers and conversely, without any need of use to probability masses over positive integers. We show that if we use Fibonacci coding in the first part of each codeword we can achieve shorter expected codeword length than Elias Omega code. In addition, our proposed algorithm has low complexity of encoding and decoding procedures.
Manuscript profile