Design Issues in an Energy Autonomous IoT Sensor Node
Department of Electronics Engineering, Aligarh Muslim University, Aligarh, India.
E-mail address: firstname.lastname@example.org
The design of an energy autonomous IoT node is challenging in the presence of cost and form factor constraints. The low cost of a sensor node will help its popularity and extent of penetration. Secondly, the low form factor will ensure that these nodes will not disturb the deployment area. Moreover, the power consumption of these nodes has to be eventually reduced to a level where energy harvesting can be effectively employed, thereby, making them energy autonomous. It has been noticed that the wireless power is much more than the processing or sensing power in a high throughput IoT node. The power of these always ON nodes can be reduced by trading off transmitted power with processing power. The processing power can then be reduced by employing near threshold operation. The write power of flash and static power consumption of SRAM can be curtailed by replacing these memories by the upcoming non-volatile memories like STT-MRAM or RRAM etc. The microcontroller power is also lowered by switching its supply voltage from nominal to near threshold level and to employ multi-core microcontrollers for higher throughput always active applications. In the end, it is necessary to reduce the power consumption of these nodes by using devices beyond CMOS like Tunnel FET and Spin devices. This talk will cover techniques that can be effectively employed to eventually reduce the energy of the constituent blocks of the IoT node by using either CMOS or beyond CMOS alternatives such that the overall power consumption level of the node is such that energy harvesting becomes practical without much area overhead (cost).