Light fantastic

Abstract

Holographic data storage (HDS) systems have the potential to revolutionise the way we store electronic data. The method of writing and reading data in a holographic data storage system In their Letter, authors from Soongsil University, Korea, explore the commercial opportunities for these systems, assess their current limitations and outline how the obstacles may be overcome. Their iterative detection method for HDS systems uses readback signals as a source of extrinsic information, leading to an improvement in channel-detection output. Fuelled by the drive for bigger, better and faster data storage systems, holographic data storage systems have attracted a lot of attention. HDS systems offer the possibility of pushing past the areal density of current forms of storage media; in other words, the number of bits of information that can be stored on a given physical size of storage medium. Whilst conventional storage systems such as CD's, DVDs or magnetic hard disk drives store information on their surface, HDS stores data in three dimensions. To store information, data is formatted into ‘pages’, which then form holograms in an optical medium. To record data, one laser beam passes through a 2D array, which is in fact the page composer, a device to pick up the input information (the ‘signal’ beam or ‘object’ beam), and meets another reference beam in the material. The signal carrying beam interferes with the reference beam inside the recording medium to produce the holograms. To read data, the recorded hologram is re-illuminated with the original reference beam to form reconstructed data. This is then imaged onto a pixelated detector array. These holograms can be superimposed, and by changing the wavelength, angle of the hologram, and various other factors, thousands of pages can be stored. The authors claim, “HDS systems can theoretically provide an achievable areal density of around 40 Tb/in2, very short access times (<50 µs), extremely fast input/output rates (>10 Gb/s), and enormous search capabilities at very high rate (>100 Gb/s)”. However, as with any new technology, there are technical challenges to face in order for HDS systems to realise their full potential. Defects in the materials involved in delay system optimisation, crosstalk between superimposed pages (inter page interference) due to their superposition, and imperfections introduced during the fabrication of the laser source, mean page composers and high speed read-out detectors can cause severe misalignment. The use of the authors’ iterative detection model, along with the support of extrinsic information, improves performance in situations with misalignment and 2D intersymbol interference (a case where a pixel is smeared vertically and horizontally onto its neighbours). By looking at the surrounding 8 pixels, they are able to more accurately detect information about the smeared pixel in question. This, in turn, allows a channel detector to make more accurate decisions. Alongside their work overcoming the drawbacks of 2d intersymbol interference, more work must be done before HDS systems are ready for large-scale commercial use. However, there is undeniable potential. In comparison to HDD or SSD, vast amounts of data can be stored more easily, and more cheaply using HDS systems. It is also more energy efficient whilst maintaining superior performance; this makes it a natural solution for back-up and disaster recovery applications: “we expect that with its unique advantages, combined with improvements of the architecture, materials, signal processing techniques, fabrication and replication, HDS will develop into commercialised products for both enterprises and personal applications”. HDS systems have many advantages and potential uses