A Radio Eye

Ken Tapping, 11th September, 2013

The human eye is an amazing thing. Its basic design, which we share with most of the animal kingdom, took Mother Nature many millions of years to evolve. It’s not surprising therefore that it has taken us time and effort to develop a technological equivalent, even though we knew what we were trying to emulate. A major project at our observatory is to develop a radio eye, to image radio waves in the same way our eyes image light.

Our eyes comprise two main components: a lens to collect light and to project an image, and an array of sensors to convert that image into data our brains can assimilate. Because each sensor element measures the brightness and colour of one point in the image, the more sensors there are, the more detailed the image we can see.

Our first attempt to copy nature was the photographic camera. The lens projected an image on a sheet of plastic or glass that was coated with silver compounds sensitive to light. Each grain of chemical was affected by the light and recorded the brightness of the light at that point in the image. One problem was that before the image could be seen, the film had to be processed. This was nothing like real-time imaging.

The invention of the digital camera changed that. In these the image is projected on an array of electronic sensors, which can be read out, processed and displayed immediately, and recorded as computer data files. The bit of image measured by one sensor is called a “picture element”, or pixel, and the more of them we have, the more detail we see in the image. Modern digital cameras have sensor arrays with more than 12 million pixels, that is 12 megapixels. This number will continue to rise.

Even though radio waves are the same sort of thing as light waves, just longer, it has taken a long time to develop a radio equivalent of the digital camera, or human eye. However, we are now doing exactly that, as part of Canada’s contribution to a large, international radio telescope project, called the Square Kilometer Array, or SKA.

Because radio waves are much longer than light waves, they require much larger devices to form the images. Lenses get too big and heavy, and since they have to be transparent, they can only be supported round the edges. Fortunately, we can use concave mirrors instead. These have a disadvantage in that they form an image in front of the mirror, so the sensor array blocks part of the light reaching it. However, because the light or radio waves do not have to go through the mirror, it can be well supported from behind. This is why dish-shaped antennas are standard tools in radio astronomy; we can make them big.

The sensor array being developed at our observatory consists of many small antennas, which can be used over a wide range of radio frequencies and crowded closely together. Achieving this was a challenge. However, from the front the antenna array looks simple, like a lot of open-ended box-shaped tubes jammed together. When you look at it from the side, the sheer complexity of the device becomes clear. Because it would be impractical to run hundreds of cables down from the antenna, the signals have to be processed at the antenna. The result is a slab of complex signal processing electronics, including computers, mounted on the back of the antenna array. We are building the device for radio astronomical observations, but it will be useful for lots of other things, such as radar, remote sensing and interference rejection. It is an excellent example of how research sciences produce technologically useful spinoff.