Real-time monitoring of sinusoidal signals
Sinusoidal signals are encountered in many technology areas and methods. Finding a way to track these outputs is crucial. Current signal monitoring methods are highly specified, but the development of more flexible, generalised techniques would be beneficial in order to facilitate the Internet of Things (IoT) and other sensor networks.
Oxford researchers have developed the Prism object; a flexible and efficient signal monitoring methodology, which operates at the lowest theoretical bounds of variance. This new approach has widespread application across any field where it is necessary to monitor sinusoidal signals rapidly and in real-time.
A sine wave is a mathematical function describing a repetitive oscillation and is ubiquitous across mathematics, physics and engineering. In signal processing, sinusoidal waves are regularly encountered in mechanical, electrical or acoustic systems, arising from the underlying physics or from the sensing technique applied. Finding methods to digitally track these waveforms is vital.
A sine of things to come
With the advent of the IoT (Internet of Things), the deployment of sensors has increased and systems have become more interconnected. With larger systems, sensors must become self-validating, more flexible and easier to deploy. Current sensors generally employ fixed algorithms that are applied to data streams to monitor specific parameters, and do not lend themselves to the networked nature of the IoT. This drive for greater sensor flexibility has led researchers at the University of Oxford to develop the Prism methodology.
Prism: Precise, repeat integral signal monitoring
The Prism is a recursive FIR (Finite Impulse Response) filtering object, which generates two orthogonal outputs, from which frequency, amplitude and phase may be calculated using simple and computationally efficient techniques. Critically, the resulting variances are close to the theoretical lower bounds, thus yielding high accuracy. Further simple and efficient techniques enable the isolation and tracking of multiple sinusoid components in any signal.
Prism design is simple, facilitating a modular approach to signal processing. Prism networks may be created or modified in real time in response to new goals or changing signal properties e.g. adapting to a detected fault.
The main benefits of the Prism objects are as follows:
- A stable FIR technique, with a recursive calculation for low computational overhead
- Simple filter design facilitating real-time modification of Prism networks to match evolving signal properties, for self-optimisation or self-validation
- Variance close to the theoretical minimum (Cramér-Rao Lower Bound)
- Tracking of multiple sinusoid components in a single signal
- A simple, efficient and powerful toolset for any application using sinusoidal signals.
The Prism method is covered by a UK priority patent application and Oxford University Innovation Ltd. is looking to talk to potential partners to aid in the commercialisation of the technology.
about this technology