Improving the Efficiency of Embedded Data Logging on NAND Flash for IoT Systems
Scott Fazackerley, Ramon Lawrence · 2021
A key requirement for many Internet of Things (IoT) systems is the collection and storage of time-based sensed data. Edge-devices are typically embedded systems with limited capabilities and power constraints. The fundamental challenge of storing time series data on embedded systems with raw flash memory has received minimal attention. This work develops an efficient technique for providing record-level consistency that significantly improves upon the typical, sequential storage approach. Simulation results demonstrate a reduction in the memory operations performed, time spent on data storage, energy consumed, and memory wear on the device.