Field-programmable gate array technologies are creating a new range of challenges for pervasive and ubiquitous systems. Revisiting and extending approaches borrowed from the purely software domain is a fundamental opportunity in this scenario. In particular, this paper addresses code mobility, a well-established approach used to dynamically adapt a distributed system based on the actual application needs, and extends it to a deep code mobility concept, allowing 'logical' hardware components to be migrated across a pervasive infrastructure. The work presents the architecture and the prototype implementation of a reconfigurable computing framework providing full support to deep code mobility through an abstraction layer which exposes a portable view of the underlying reconfigurable hardware. The paper then thoroughly discusses two application scenarios, hardware-accelerated distributed data mining and autonomous online testing, confirming the impact of deep code mobility in real-world pervasive computing contexts. © 2018 Inderscience Enterprises Ltd.
An abstraction layer enabling pervasive hardware-reconfigurable systems / Cilardo, Alessandro; and Mazzocca, A.; Prinetto, Paolo. - In: INTERNATIONAL JOURNAL OF EMBEDDED SYSTEMS. - ISSN 1741-1068. - 10:5(2018), pp. 366-377. [10.1504/IJES.2018.095022]
An abstraction layer enabling pervasive hardware-reconfigurable systems
Cilardo;A. and Mazzocca;PRINETTO, PAOLO
2018
Abstract
Field-programmable gate array technologies are creating a new range of challenges for pervasive and ubiquitous systems. Revisiting and extending approaches borrowed from the purely software domain is a fundamental opportunity in this scenario. In particular, this paper addresses code mobility, a well-established approach used to dynamically adapt a distributed system based on the actual application needs, and extends it to a deep code mobility concept, allowing 'logical' hardware components to be migrated across a pervasive infrastructure. The work presents the architecture and the prototype implementation of a reconfigurable computing framework providing full support to deep code mobility through an abstraction layer which exposes a portable view of the underlying reconfigurable hardware. The paper then thoroughly discusses two application scenarios, hardware-accelerated distributed data mining and autonomous online testing, confirming the impact of deep code mobility in real-world pervasive computing contexts. © 2018 Inderscience Enterprises Ltd.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.