Journal
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS
Volume 67, Issue 6, Pages 1925-1935Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCSI.2020.2973676
Keywords
Latches; Radiation hardening (electronics); Transistors; Redundancy; Hardware; Digital storage; Double-node upset; latch; radiation hardened; reliability; process variations
Categories
Funding
- National Natural Science Foundation of China [61604133, 61501275]
- Young Creative Talents Training Plan of General Universities of Heilongjiang Province of China [UNPYSCT-2017152]
Ask authors/readers for more resources
Double-node upsets induced by the charge sharing effects are emerging as a major reliability issue in nanometer latch design. Although the existing robust latches can provide a good tolerance for double-node upsets, the implementation of these hardened latches incurs in considerable hardware penalties in terms of delay, area, and power, because they rely on traditional hardening techniques such as space redundancy. In this paper, a novel low-complexity (with respect to hardware redundancy in terms of area, delay, and power) radiation hardened (RH) latch is proposed; this latch is based on the dual interlocked storage cell (DICE). Based on the radiation upset mechanism, in particular the upset polarity of the transient pulse, the proposed RH latch can effectively reduce the number of protected nodes (sensitive nodes), as well as the number of transistors, thus reducing circuit overhead. At the same, a single node upset in any sensitive node and a double-node upset in any sensitive node pair can be recovered, because at least two stable nodes can retain the values even when a double-node upset occurs. The results are based on mapping designs to TSMC 65 nm commercial CMOS process design kit (PDK) and demonstrate that the proposed RH latch incurs in significant reductions in terms of propagation delay, circuit area, and power-delay-area product (PDAP), compared with existing hardened latches. The tolerance functionality of the proposed latch is successfully validated by fault injection. In addition, the impact of variations and process corners is assessed using Monte Carlo (MC) simulation. The simulation results confirm that the proposed latch can also tolerate all upsets, even under extreme values of process variations.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available