05/21/2019 | News release | Distributed by Public on 05/21/2019 18:00
By Tony Ticknor, Ph.D. on May 21, 2019 | Leave a Comment
Once in a while one needs to crawl out on a limb and see what can be shaken loose. It is always fun to graft a few different branches of technology lingo onto the tree and see what fruit may fall. Today's menu: drones; lasers; control, and the bonus word, data.
That list is not a purely random selection. I can recall more than once over the years hearing a question along the lines of 'what about laser-controlled drones?' Now let us try to consider some answers. To begin, I attempt to imagine what that might be. Are we going to point a laser at a robot aircraft and send flight commands along the laser beam to tell it how to fly, where to go, what to look for, when to shoot, etc.? Certainly that could be done, and there may even be a few reasons you might want to do that. But excluding the sheer entertainment, those reasons would probably be substantially outweighed by the reasons you might not want to do that. Plus, if that is the kind of thing you are inclined to do, you have probably already 'freaked-out' your neighbors enough this year.
So, should we decide: 'nothing to see here, move on to something else'? That is not how innovation happens. When presented with an unexpected observation or idea, no matter how off base it seems at first, you should not be limited to recounting why that must be wrong. Instead try to find an explanation of how that might be right. This is the most important behavioral trait taught to me through example by any of my college professors (thank you Dr. Harrison Barrett).
So, if laser-controlled drones are to someday arrive, and we are going to care about them, what then are we talking about? A key use of lasers is in communications networks. In a communications network, control is not about managing the mechanical motion and behavior of physical objects. Network control is about managing the flow of data within the network, which may incidentally include moving parts. (Disclaimer: don't text while driving.) The term 'drone' evokes different concepts among different audiences, but we can loosely consider a drone to be an unmanned aircraft, and even more loosely, also small spacecraft. So maybe we should be considering a communications network employing interconnected drones, with the network control (and what the heck, data too) being transmitted by lasers along optical links. Abracadabra: laser-controlled drones.
OK, so there is an application; but is it a useful application? Well, it turns out that some of the most prominent technology companies have been thinking that it might be. Alphabet subsidiary Loon pursues providing internet connectivity to a broader range of the world with networks of high-altitude balloons (Project Loon). Facebook would like to get more people online through unmanned winged aircraft loitering around also in the stratosphere (Facebook UAV). And SpaceX, not letting rockets be left out of the conversation, wants to up the 'world-wide' factor of the web with a network of low-Earth-orbit satellites run by Seattleites (SpaceX Starlink). (Laser data not to be confused with blinking saddle lights.) Additionally, Verizon is evaluating drones (ALO) as network nodes for rapid repair and/or upgrading of networks impaired by disasters or other events (not yet with lasers, but such evolution would be inevitable). And the list goes on from there. Regardless of the esoteric nature of these platforms, they are still intrinsically communications networks. They still need to connect the data to and from a broader network (backhaul), and move the data around within the specialty network (fronthaul). And now, these network nodes are moving around at tens or hundreds of kilometers per hour, or even thousands of km/h (for the satellites). These then must stay connected to users who may be stationary on Earth, or may be moving in their own directions at tens or even hundreds of kilometers per hour.
If such a network is to be successful, it needs to reliably support a great number of users to send and receive a great deal of personalized data. The backhaul / fronthaul / crosshaul between the 'drones' inevitably must be optical, and in most of the above cited cases that is already planned. Due to spectrum limits, security issues, size, power, and other factors, the links between the drones and the ground stations must also ultimately include optical channels. Those optical channels obviously cannot be over optical fiber and will instead beam the laser signals through the atmosphere using free-space optics technologies. But just as a network with unusual stations is still intrinsically a network, transceivers with unusual optical channels will still be intrinsically transceivers.
However, the characteristics and challenges of free-space optical links unfold differently than the characteristics and challenges of fiber-optic links. A key challenge for transmitting high-speed optical data through more than 1 or 2km of the lower atmosphere (troposphere) is an atmospheric effect called 'scintillation'. You may think that on a clear day you can see forever, but even on the clearest days, to a laser beam every flight is a turbulent flight. The laser beam gets bounced around and split apart, held back and surged forward. The laser signal reaching the detector carries varying intensity and phase that can easily overwhelm the underlying data signal, substantially reducing data speeds or even destroying the optical link. Conventionally, compensating for the scintillation requires bulky opto-mechanical systems such as adaptive optics or cumbersome spatial and temporal-modulation corrections known as phase conjugation. Recently however, new optical communication capabilities for correcting impairments have emerged in the form of coherent optical communications technology. With coherent technology, many of the signal impairments that in the past could only be compensated with bulky physical devices can now be corrected or reduced in software by a digital signal processor (DSP) when coherent optics is used. Some of those same principles can be applied where the impairments are coming from scintillation and other free-space degradations. These degradations are not seen in fiber-optic systems, but they can likewise be suppressed by coherent optics and DSP software configured for free-space use. A coherent transceiver using software phase recovery and error correction routines can in essence emulate physical phase conjugation to correct atmospheric scintillation. It is by no means a perfect replacement, but as with fiber-optic applications, a coherent transceiver is sufficiently better than a traditional transceiver that it can provide significantly improved performance while at the same time enabling the simplification or even elimination of bulky physical compensating devices. For the drone-containing networks cited above, this trade-off is not just an economic equation, it is an enabling capability. Small drones and tiny satellites can have the space and resources to carry coherent transceivers and free-space optics, but they simply will not have the room and capabilities to implement full-scale physical phase conjugation.
Large areas of the world's population remain underserved for access to the global network. Reasons for this are highly varied, but some common factors are the various challenges of establishing permanent terrestrial infrastructure. And network infrastructure that is established can experience widespread significant disruption and failure as a result of events like storms, earthquakes, and wildfires, and at least two of those are expected to significantly increase in the coming years. More and more, vital information and even lives depend on working networks. Being able to quickly restore or enhance network capabilities is often a public-safety issue. By anyone's measure, demands on network capacity will continue to increase at breakneck speeds. 5G architectures and 'internet of things' concepts move even more data around among more segments of the networks. In fact, for the drone networks cited above, the drones may not be only network nodes, but they may also be internetworked 'things.' They may be providing their own data such as high-definition video, LIDAR imaging, intensive environmental monitoring, etc. And to do that, they will also need to consume ever more command data in order to control what data they must produce. Upon all such considerations and more, it should be no surprise at all that there is much ongoing interest in non-terrestrial high-bandwidth networks.
Furthermore for such networks, the network resources are by design out in the open. Protecting the security and integrity of the data and even the network itself will take on more elaborate aspects. For these matters, the lasers and optics can further chip in to immunize the network against conventional RF interference, surreptitious listening, and network skyjacking.
Absent unforeseen revelations, data networks using drones (as liberally defined hereinabove) are inevitable. That they will use lasers is inevitable. Since the network nodes will be moving along their own complex trajectories, since energy will be at an even greater premium on these remote platforms, andto enable enhanced security, the laser beams will be tightly collimated and hence must be actively pointed at their targets for the greatest efficiency. The network to the users will include 5G (or later) cellular technologies, requiring further beam-steering capabilities between the in-flight nodes and the potentially moving ground targets. It is also therefore inevitable that the 'control' data needed to keep the 'work' in such a network must increase many fold. So yes, our future robot overlords will include laser-controlled drones overhead, shooting directed-radiation beams at each other, and at us (leaving us no recourse but to have laser beams attached to our heads).