Title: Dynamically reconfigurable evolutionary multi-context robust cellular array design

Authors: Sobhan Soleymani; Afzel Noore

Addresses: Lane Department of Computer Science and Electrical Engineering, West Virginia University, Morgantown, WV 26506, USA ' Lane Department of Computer Science and Electrical Engineering, West Virginia University, Morgantown, WV 26506, USA

Abstract: This paper presents a novel dynamically reconfigurable evolutionary design of multi-context multi-output near real-time functions for implementation using a 2D cellular array structure. A genetic algorithm generates solutions that reuse the cells in the cellular array to implement multiple multi-output functions, thereby minimising the area overhead and power consumed. Further constraints are introduced to yield solutions that are resilient to single cell failure. Test vectors to identify the faulty cell are presented. The experimental validation and results of the proposed approach is demonstrated on a 5 × 5 cellular array to implement context-specific arithmetic functions under fault-free and faulty cell scenarios. The upper and lower bounds for the total time delay in the reconfigurable 2D array is derived.

Keywords: fault-tolerant embedded design; evolvable arithmetic circuits; reconfigurable cellular arrays; 2D cellular arrays; multi-context design; genetic algorithms; dynamic reconfiguration; cellular array design; cell reuse; cell failure; resilience; fault tolerance.

DOI: 10.1504/IJCAD.2016.075889

International Journal of Circuits and Architecture Design, 2016 Vol.2 No.1, pp.1 - 12

Received: 25 Mar 2015
Accepted: 09 May 2015
Published online: 12 Apr 2016 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article