Date:6 April 2017
Optical communication is on the frontiers of communication. It’s communication using patterns of light and South Africans are leading the charge.
Optical communication is the technology at the very back-bone of our society. When we connect over the Internet, make calls or send data by our smartphones, we ultimately link into the fibre-optic network around the globe.
Light is sent down this glass fibre, which is no more than a tenth of a human hair in diameter, and detected on the other side. To encode information into the light, we make use of its core properties, for example, using many colours, each to carry information, or using intensity modulation, switching light on and off in a manner analogous to the old Morse code.
Today our state-of-the-art networks enable a global average connection speed of 5,1 Mbps, and up to 32,2 Mbps for global average peak connections. This is according to the Akamai State of the Internet Report (Q3 2015). The report further tells us that South Africa fares the worst in the entire Europe, Middle East and Africa (EMEA) region, wherein Sweden had the highest average connection speed at 17,4 Mbps.
South Africa had the lowest, at 3,7 Mbps. Romania had the highest average peak connection speed at 72,9 Mbps. South Africa had the lowest at 18,9 Mbps. [see “Back to pigeon for SA” below]
Every decade, we have seen disruptive technologies emerge to drive capacity forward: high-quality glass in the 80s to minimise loss and noise, Erbium-doped fibre amplifiers in the 90s to extend the reach of fibre from kilometres to across continents and, in the early part of this century, using colours (wavelengths) to pack more information in. Yet, despite these advances – each crucial to allow capacity to keep ahead of demand – the optical networks we have now are reaching a capacity limit. It is simply not possible to put more light (signal) into the optical fibres. One way to address this is simply to lay more fibre.
Paradoxically, though, the largest cost of the optical communication networks is not the technology, but the labour cost to dig the holes. For this reason, many researchers around the globe are working on a revolutionary technology to overcome our pending data crunch.
Optical communication: Using patterns of light
One approach is to exploit an unused property of the light that we already have: its pattern. When you shine a laser pointer on a wall it looks like a round blob of light. More accurately, the brightness of the light follows a Gaussian profile: more light appears in the middle than on the edges (like the Bell curve for grading students).
Now, light comes in many patterns, and importantly, these patterns can carry information. How many patterns are there? The answer: infinity. In principle, at least, it should be possible to pack an inordinately large amount of information into the light we already have, and send this down optical fibre or even across an open path in the atmosphere or space. This was first suggested more than three decades ago, but only today has the technology matured enough for this to be viable.
This large time lag is not unusual in science. The accepted rule of thumb is that, if you want to know what the next technology will be, look back 20 years in the scientific literature. To see how patterns of light could carry information, imagine each pattern as a channel for information, like a pipe carrying water.
At the moment we use only one pattern in our optical communication systems – the very same Gaussian pattern of the laser pointer. If we could use N patterns, then each would carry the same information as the existing network, so the information capacity would increase by N. So, 1 000 patterns means 1000x faster Internet. To make this work, the patterns must all be combined into a single stream of light, but each pattern in that stream is its own channel that is independently modulated for information transfer. This is how information is transferred using colours: like the rainbow, many colours travel together, each modulated separately, but can be split later into distinct groups.
In this sense, while they travel together, they don’t mix. The reason for this is that they are unique, so even though they are combined they can be distinguished from one another. Imagine a bowl of red and green sweets: they are together, but they don’t mix in the sense that red and green sweets do not combine to form a new coloured sweet. The combining of the patterns before entering the fibre is called multiplexing, and the splitting at the other end, demultiplexing.
The entire link is then called space division multiplexing, or sometimes mode division multiplexing. Of course, one could say that this is nothing more than a very fancy way of waving flags for communication! Flag waving, used extensively in the shipping world, also uses a spatial pattern to represent information. The only difference is that here the pattern is encoded into the light itself.
The challenge is to do this with patterns of light (See Fig 1). One of the technologies used for this is digital holograms (see Popular Mechanics Nov. 2008, pg. 74-78). Here miniature LCD displays (like your television at home) are used to change light from one pattern to another, which we call spatial light modulators (SLM) because they change (modulate) light at every point (spatially) to form a new pattern.
All the patterns required can be created in this manner in a rewritable fashion. Once a pattern set is chosen, a more efficient glass optic can be used to multiplex and demultiplex the light. The challenge is that patterns do mix. Imperfections in the fibre and distortions in the air cause patterns to overlap in a way that makes it difficult to tell them apart – like looking at an outdoor scene through a fuzzy window. This remains a topic of intense research.
Another issue is the so-called “last mile”, where communication systems over long distance must change to short distance solutions at the end-user. For example, the long-haul fibre (glass) must link to systems through air (Wi-Fi). The issue is that patterns that transmit well in one medium do not work very well in the other.
Example: The “Last Mile”
Recently a significant step forward was demonstrated in South Africa. The team showed that it is possible to design a fibre that supports a set of patterns that work in both fibre and air, with minimal pattern mixing during transport. To demonstrate the concept, the team performed a proof-of-principle experiment at the University of the Witwatersrand. Each pixel in an image has a particular shade, which in turn can be designated as a unique pattern (see figures 2 and 3). The patterns were transmitted through both air and fibre, so the original image that was sent could be received at the other end of the link. This is a promising technology for addressing the last mile problem, a necessity if patterns of light are to be used in optical communication networks.
Andrew Forbes is a Distinguished Professor in the School of Physics, University of the Witwatersrand (Johannesburg), and leads the Structured Light Laboratory (structuredlight.org) that focuses on optical communication in free space and in fibre. Kristin Klose is Director for Innovation Priorities and Instruments at the Department of Science and Technology (Pretoria), focusing on innovation policy with the aim of creating and sustaining an enabling environment for innovation, technology development, and the commercialisation of publicly funded R&D.
Before Samuel Morse invented the telegraph in 1837, the only way to communicate across distances was to send messages by smoke signal, homing pigeon or by human messenger on foot or horseback. This could take days, weeks or even months, making the information that was shared rather dated. By the 1850s, the telegraph cable was stretched from coast to coast, making it possible to send a message in mere minutes from London to New York. Soon, however – and certainly by today’s standards – the capacity limitations of the telegraph system became clear.
Only relatively short and simple messages could be sent. This gave added momentum to an enduring human obsession: to perfect the transmission of optical and audio signals for the purposes of communication over ever longer distances in space and time. Since the advent of the computer and the Internet, our expectations today are to access more diverse sources of information even faster, which is driving us to obsess further about how to squeeze more and faster means into communicating across boundaries.
Back to pigeon in SA?
Frustration with the national bandwidth shortage and the extremely slow internet transmission times drove a Durban-based IT company to conduct an unusual experiment in 2009. Their aim: to prove that it was faster to transmit data using a homing pigeon rather than the national internet service provider, Telkom. At the time an 11-month-old carrier pigeon, Winston, reportedly took one hour and eight minutes to fly 80 km from the company offices in Pietermaritzburg to Durban, with a data card strapped to his leg. Including download time, the entire data transfer took two hours, six minutes and fifty seven seconds. This was the same time it took for only four per cent of the data to be transferred using a Telkom line. (Source: reuters.com)
This article was originally published in the May 2016 issue of Popular Mechanics magazine.