Ralph Ledwa

An 82μA/MHz microcontroller with embedded FeRAM for energy-harvesting applications

In recent years energy-harvesting technology has become much more mature. The energy source is free and widely available. However, there are some limitations that energy-harvesting devices need to cope with, to build a reliable system. First of all the peak current of typical energy-harvesting power sources is quite limited [1] and the systems active power must be as low as possible. A second limitation is the possibility of interruption or fluctuation of power source. For example, solar energy is only available during the day, and thermal or vibrational energy sources may change depending on environmental conditions. As a consequence, systems must retain state and information despite possible power outages. Flash or EEPROM as todays standard non-volatile memory is slow in write access and has high peak current demands due to the required charge pump for erase and write cycles. Flash technology has limited write endurance in the order of 105 write cycles requiring a Wear Leveling algorithm [2] for large cycle state storage. SRAM could be used for fast, low-power read/write [3], but it requires a separate battery for state retention.

Risk management in embedded devices using metering applications as example

Along with the rise in use of everyday life electronic products that collect and communicate personal data, there is an increasing need for adequate security. The use of ultra-low-power MCUs in such applications provides a clear advantage in terms of energy consumption. However, given their general-purpose nature and low-power needs, security has not been the main focus in the past. This work places emphasis on methodologically analyzing open security gaps at a system level and providing a score for each vulnerability found. Such vulnerability scores help prioritize the efforts towards building a secure system and balancing the trade-off between suitable protection and minimal cost. The work presented uses as an example an abstraction of metering applications implemented using a general purpose microcontroller. The presented approach makes use of the Common Vulnerability Scoring System open framework to quantify the impact of possible vulnerabilities and prioritize their remediation based on their relevancy.