Questions submitted to AIMday

How to reliably count the number of co-channel signals in the presence of spatially correlated noise?

The problem is co-channel signal enumeration / signal subspace dimension estimation for non-Gaussian signals in the presence of spatially

correlated noise. Traditional information theoretic methods strictly only apply to Gaussian signals in additive white Gaussian noise. In practise

the presence of non-Gaussian signals is okay, but correlated noise is a big problem and leads to significant over-estimation. How can this be

countered? An extension is to consider adaptive operation with a changing number of co-channel signals.

How to match the power consumption requirements of various sensors with our energy harvesting product?

Energy harvesting devices, specifically regenerate kinetic energy into electricity, intermittent power output, and the current is typically low. We would like a discussion to maximise useable power output, and explore applicable scenarios where sensors can become self-powered. Our aim is to remove the problem of having to replace the battery, or at least extend the battery life.

Are there parts of the RF spectrum not currently exploited for sensing which could be with new technology and signal processing methods?

The RF spectrum is getting very crowded and only so much can be done within the bands currently allocated. Exploiting third party signals could be done but only if they are ever present and of the correct strength. The other possibility is to use parts of the spectrum where very little goes on. These may have been ignored for being sub-optimal for sensing but could be unlocked by new technology or signal processing to make up for sparse data etc.

I'd like to discuss methods of measuring vehicle suspension deflection and performance.

I'm currently building a suspension monitoring system for a commercial product but feel that a technology using accelerometers would be more discrete and could provide additional benefits. However I am open to any pertinent technology.

I would be interested to know how I could monitor a Mountain Biker or Trials rider

Specifically in terms of trying to produce an avatar of the rider and bike moving through a terrain - and in particular to be able to see the attitude of the bike and the body movement of the rider when manipulating the bike on tricky technical terrain.

Integrating sensors into clothing for wearable technology applications has numerous advantages. These include making the technology less intrusive, ensuring it can't be forgotten and making it more convenient to use/wear. Is it possible to develop a textile-based temperature and humidity sensor using polymers or biopolymers?

n/a

Is it possible to develop a wearable device that can detect and log when a garment is being worn (versus not being worn)?

The device should be unobtrusive, have a long battery life (i.e. weeks rather than days, the longer the better) and provide basic connectivity to download the data (with data being stored locally until it can be downloaded). It should either be worn by an individual or be attached in some way to the garment being monitored.