Search Results

Now showing 1 - 3 of 3

Citation - WoS: 3
Citation - Scopus: 3
Array Bp-Xor Codes for Hierarchically Distributed Matrix Multiplication
(IEEE, 2022-03-01) Arslan, Şuayb Şefik
A novel fault-tolerant computation technique based on array Belief Propagation (BP)-decodable XOR (BP-XOR) codes is proposed for distributed matrix-matrix multiplication. The proposed scheme is shown to be configurable and suited for modern hierarchical compute architectures such as Graphical Processing Units (GPUs) equipped with multiple nodes, whereby each has many small independent processing units with increased core-to-core communications. The proposed scheme is shown to outperform a few of the well–known earlier strategies in terms of total end-to-end execution time while in presence of slow nodes, called stragglers. This performance advantage is due to the careful design of array codes which distributes the encoding operation over the cluster (slave) nodes at the expense of increased master-slave communication. An interesting trade-off between end-to-end latency and total communication cost is precisely described. In addition, to be able to address an identified problem of scaling stragglers, an asymptotic version of array BP-XOR codes based on projection geometry is proposed at the expense of some computation overhead. A thorough latency analysis is conducted for all schemes to demonstrate that the proposed scheme achieves order-optimal computation in both the sublinear as well as the linear regimes in the size of the computed product from an end-to-end delay perspective.
Citation - WoS: 4
Citation - Scopus: 6
Data Repair-Efficient Fault Tolerance for Cellular Networks Using Ldpc Codes
(IEEE, 2022-01-01) Haytaoglu, Elif; Kaya, Erdi; Arslan, Şuayb Şefik
The base station-mobile device communication traffic has dramatically increased recently due to mobile data, which in turn heavily overloaded the underlying infrastructure. To decrease Base Station (BS) interaction, intra-cell communication between local devices, known as Device-to-Device, is utilized for distributed data caching. Nevertheless, due to the continuous departure of existing nodes and the arrival of newcomers, the missing cached data may lead to permanent data loss. In this study, we propose and analyze a class of LDPC codes for distributed data caching in cellular networks. Contrary to traditional distributed storage, a novel repair algorithm for LDPC codes is proposed which is designed to exploit the minimal direct BS communication. To assess the versatility of LDPC codes and establish performance comparisons to classic coding techniques, novel theoretical and experimental evaluations are derived. Essentially, the theoretical/numerical results for repair bandwidth cost in presence of BS are presented in a distributed caching setting. Accordingly, when the gap between the cost of downloading a symbol from BS and from other local network nodes is not dramatically high, we demonstrate that LDPC codes can be considered as a viable fault-tolerance alternative in cellular systems with caching capabilities for both low and high code rates.
Citation - WoS: 4
Citation - Scopus: 7
On the Distribution Modeling of Heavy-Tailed Disk Failure Lifetime in Big Data Centers
(IEEE, 2021-06-01) Arslan, Şuayb Şefik; Zeydan, Engin
It has become commonplace to observe frequent multiple disk failures in big data centers in which thousands of drives operate simultaneously. Disks are typically protected by replication or erasure coding to guarantee a predetermined reliability. However, in order to optimize data protection, real life disk failure trends need to be modeled appropriately. The classical approach to modeling is to estimate the probability density function of failures using nonparametric estimation techniques such as kernel density estimation (KDE). However, these techniques are suboptimal in the absence of the true underlying density function. Moreover, insufficient data may lead to overfitting. In this article, we propose to use a set of transformations to the collected failure data for almost perfect regression in the transform domain. Then, by inverse transformation, we analytically estimated the failure density through the efficient computation of moment generating functions, and hence, the density functions. Moreover, we developed a visualization platform to extract useful statistical information such as model-based mean time to failure. Our results indicate that for other heavy-tailed data, the complex Gaussian hypergeometric distribution and classical KDE approach can perform best if the overfitting problem can be avoided and the complexity burden is overtaken. On the other hand, we show that the failure distribution exhibits less complex Argus-like distribution after performing the Box–Cox transformation up to appropriate scaling and shifting operations.

Bilgisayar Mühendisliği Bölümü Koleksiyonu

Browse

Filters

Settings

Sort By

Results per page

Search Results