OVERVIEW
Energy-efficient, hands-free activation is becoming a “must have” feature in mobile devices, wearables, smart speakers, and many other battery-operated devices. Features such as voice control, gesture control, or face unlock demand an always-on user interface (UI) to be totally hands-free. While employing intensive signal-processing algorithms, such as those used for speech recognition and face detection, the supporting technology must consume very little power to conserve battery life.
Sensing is a key feature of smartphones, wearables, and IoT devices. The combined processing of sensory data from disparate sources, also known as sensor fusion, provides valuable contextual awareness for a host of user applications. Multiple MEMS sensors such as 9-Axis motion sensors, barometers, thermometers, and others call for a high-precision sensor fusion DSP to accurately and efficiently deduce motion and environmental data, and eliminate false readings.
The key to meeting these requirements is to choose a signal-processing IP that provides the following:
- A highly optimized arithmetic Instruction Set Architecture (ISA) and a rich control ISA that are efficiently utilized by the compiler as it handles sensor fusion software, which often mixes arithmetic and control code
- Ultra-low power consumption, enabling it to efficiently sustain always-on operation
- Minimal silicon area, which allows for very low leakage power and a cost-effective implementation
You are welcome to join our always-on technology group on LinkedIn.
Target Applications
As smartphones and tablets host an ever increasing variety of sensors, the benefit from fusing all of that sensor data increases accordingly. As always-on voice activation and gesture-based activation become standard, the need for an efficient implementation of both hardware and software proves mandatory.
With their small form-factor and increasingly complex UI needs, wearable devices often implement voice-based control to allow for the breadth of user interaction needed.
IoT devices usually have sensing, processing, and connectivity functions. Because sensing and processing are often always-on, it is essential to have an efficient implementation of those functions to allow for a decent battery life without compromising on the rich functionality they offer.