Daniel Khankin | Searcharxiv

arXiv2024

Evaluation of Posits for Spectral Analysis Using a Software-Defined Dataflow Architecture

Spectral analysis plays an important role in detection of damage in structures and deep learning. The choice of a floating-point format plays a crucial role in determining the accuracy and performance of spectral analysis. The IEEE Std 754\textsuperscript{TM} floating-point format (IEEE~754 for short) is supported by most major hardware vendors for ``normal'' floats. However, it has several limitations. Previous work has attempted to evaluate posit format with respect to accuracy and performance. The accuracy of the posit has been established over IEEE~754 for a variety of applications. For example, our analysis of the Fast Fourier Transform shows 2x better accuracy when using a 32-bit posit vs. a 32-bit IEEE754 format. For spectral analysis, 32-bit posits are substantially more accurate than 32-bit IEEE~754 floats. Although posit has shown better accuracy than IEEE~754, a fair evaluation of posit with IEEE~754 format using a real hardware implementation has been lacking so far. A software simulation of posit format on an x86 CPU is about $\mathbf{69.3\times}$ slower than native IEEE~754 hardware for normal floats for a Fast Fourier Transform (FFT) of $\mathbf{2^{28}}$ points. We propose the use of a software-defined dataflow architecture to evaluate performance and accuracy of posits in spectral analysis. Our dataflow architecture uses reconfigurable logical elements that express algorithms using only integer operations. Our architecture does not have an FPU, and we express both IEEE~754 and posit arithmetic using the same integer operations within the hardware. On our dataflow architecture, the posit format is only $\mathbf{1.8\times}$ slower than IEEE~754 for a Fast Fourier Transform (FFT) of $\mathbf{2^{28}\approx 268}$ million points. With this implementation, we empirically propose a new lower bound for the performance of posit compared to IEEE~754 format.

cs.AR↗

arXiv2019

Random Spanning Trees for Expanders, Sparsifiers, and Virtual Network Security

This work describes probabilistic methods for utilizing random spanning trees generated via a random walk process. Goyal et al. showed that the union of random spanning trees approximates the expansion of every cut of a graph. First, we generalize the method by Goyal et al. for weighted graphs and show that it is possible to approximate the expansion of every cut in a weighted graph with the union of random spanning trees generated by a random walk on a weighted graph. Second, we show that our union of random spanning trees is a spectral sparsifier of the graph. Moreover, we show that $O(\log n /ε^2)$ random spanning trees are required in order to spectrally approximate a bounded degree graph. This result closes a previously open question on the number of random spanning trees required for saprsification. Third, we show that our random spanning trees based construction provides security features for virtual networks, in the context of Software-Defined Networking. Network virtualization coupled with Software-Defined Networking allows new on-demand management capabilities. We demonstrate such a service, namely, on-demand efficient monitoring or anonymity. The proposed service is based on network virtualization of expanders or sparsifiers over the physical network. The defined virtual (or overlay) communication graphs coupled with a multi-hop extension of Valiant randomization based routing lets us monitor the entire traffic in the network, with a very few monitoring nodes. We propose methods that theoretically improve services provided by existing monitoring or anonymity networks, and optimize the degree of monitoring or anonymity.

cs.NI↗