(Note: all the information contained here is publicly available in Aviation Week magazine. Some personal insights come from my brief stint as acting DARPA program manager for agile operations long ago. I’ll link to Wikipedia for some background.)
I am thinking out loud here, trying to understand current military strategy. This is for war of the usual kind apart from so-called terrorism and cyberattacks. The planning I am trying to understand is what to do if we go to war with an organized military like China, Iran, Korea or Russia.
For strategic action, the Air Force plans a new long range strike bomber to be lower cost than the current, scarce B1 and B2. The bomber will be unarmed (except for the payload), optionally manned and fly with accompanying manned and unmanned aircraft. In terms of existing aircraft these UAVs are the still secret, but flying RQ-180 and the more secret, new joint RAQ-25.
The General Strategy
Okay. I understand this strategy. The Air Force wants to keep men in cockpits and rely on large craft and remote bases (to keep their excuse for existence). The new bomber will be more of a bus than a battlestar in order to be funded. If it were flying today, the F-22 will have the job of keeping the airways clear of other planes; the F-35 is to destroy countermeasure facilities; a combination of standoff missiles launched from safe locations will help.
The drone RQ-180 stays high and stays long; all it does is hide, watch and report. F-35 is not well armed and the services won’t get many of them, so their approach is to plug the all holes created by these weaknesses with the unmanned RAQ-25, which is the new item in the mix, derived from the Phantom Ray but smaller.
That’s a lot of holes to fill. The new UAV will be a stealthy, unmanned subsonic vehicle, optionally armed and with some autonomy. It is supposed to also find targets (with help) and destroy them while providing safety for itself and others. It will have periods where it goes dark, meaning that no one can guide or advise it.
Oh and it has to come from far away too, with refueling unless the Navy’s version can be enlisted.
But the Navy has decided to have its version be unarmed, apparently because it desperately needs a replacement of the F/A-18 (which could be bested in modern combat). US carriers, especially the new model, require a lot of support and protection as they are pretty vulnerable to a modern military. There isn't much room on carriers for manned mission aircraft and there isn’t room in the budget for both a sixth generation manned fighter and a fully armed unmanned fighter.
The Navy’s version of the F-35 is a problem and they need something to fly alongside and jam the bad guys. So the Navy will use the projected new unarmed drone to protect its manned aircraft and at the same time starve itself of attack capability so as to get a new aircraft funded.
But jamming is a hard job, in part because it makes it easy to see and kill you. What that means is that the folks designing this new craft have to give it a few qualities that are creative: it has to be aerodynamically agile as a matter of survivability; it has to be flexible in what it can do and that means it has to have a wide range of reasoning abilities in different contexts. This thing will have the ability to collaborate, see, kill, report, jam and evade but it will not be able to do any of these except in concert with others, most of which are not listed here.
In short, it needs to be individually dumb and instinctively nimble, but part of a shared brain that is genius-level. Let’s see how one might approach this problem.
Agility and Swarming
Why is this problem remotely interesting to me?
Well, there is a lot of money and other opportunity cost involved and for several reasons, no one in charge seems to be doing it right. This is my money and my democracy. Though I disagree with most major military decisions that have been made in my adult life, I do expect we will be atwar with a significant power before I die. As disturbing as it sounds, we could lose with disastrous consequences.
Also, I think that the problem if suitably abstracted is amenable to research funding that can be used for a wide variety of non-military purposes, including my biomedical research interest. Some of these I hold dear and would like to see advanced.
It is not necessary to sell your soul to misdirected powers to be interested in the class of solutions.
It Is All About Form
We can think of the forms of this situation in several ways.
- The craft and its companions have to get somewhere and back. The path is often complex and changes based on what is happening and possibly how the mission adapts.
- The physical shape of each craft matters a lot to us. The shape has to allow a number of aerodynamic abilities; it has to hold all the stuff it carries (and needs to otherwise work); it has to be shaped in a way that supports stealth; and it has to be coated with tiny transmit-receive modules. The skin is also the antenna.
- The arrangement of yourself with others is a very big deal. To different degrees of certainty, you need to know where everyone is and where they will shortly be. This has to be worked out with the form of both the above and below bullets. The physical assembly will include craft of many different kinds with specialized jobs, and where they are in the constellation may affects their ability to do that job.
In addition to these physical forms, all of them dynamic in different ways, we have two virtual forms:
- We have the universe of possible forms, for each of these (including the shape of individual craft) and in combination. This is a quantum universe in some respects, and will be amenable to so-called functional geometric logics. For example, each one needs to know where to go among the many possible paths and that has to be decided in the context of many possible paths of the others.
- Electronic Warfare is a collection of effects. We won’t list them here, except for one. The transmit-receive modules that coat much of the surface of modern craft are capable of seeing and sending a wide variety of signals. One thing they can do is sense radar coming in and send back an erasing signal. Far more clever is that ability to send back a bounce that indicates that yes, there is something here, but it is looks different and is in a different place.
The ability to effectively do this depends on how far apart the modules are, so obviously it would be better if modules on many aircraft acted as if they were one. This is currently limited by the communication lags that would be required to coordinate things. The probing signals are nimbly elaborate.
We will come back to this notion of a spatially computed, dynamically changing virtual arrangement of aircraft shortly.
This is the second extreme statement I have made in as many blog posts. Earlier, I said that cyber warfare was as significant a threat as climate change. This is based on two opinions. The first is that from some inside perspective, shared with many, I see profound unfixable vulnerabilities. Just profound, some of them created by NSA. Some of these will not be obvious or even believed. It is just too easy for even small groups to wreak havoc.
The second notion is that we cannot escape the fact that we run everything on information. I am not so much focused on the computer hosting and conveyance but the information itself, whether computerized or not. What happens when facts, communication, even language cannot be trusted? What happens when even narratives get broken?
Climate change is a big deal and is going to stress world affairs more abruptly than we expect. Some consequences will be water and food shortages, epidemics and incidentally the near collapse of governments. What happens when things start to go wrong? War. War will hit us first.
Probably we will see some ‘small’ wars, like the ones that have already crippled us (the US), with continued conflict in the Muslim worlds, and combat with Korea. But what I am talking about here is big war, war with China, or India over sea lanes. Or Israel allied with an unexpected partner over Arctic resources.
I would like to be wrong about the stress on governments from climate. And wrong that this stress will trigger conflict. And wrong that our bad procurement decisions will disadvantage us. I hope so.
The Air Force/DARPA Vision
We know that DARPA’s program for futures in this regard follows long established traditions in the aerospace business: you can design a system of several components but each one has a distinct mission, separate funding (and contractors). This comes from a reductionist, adversarial method of buying things. For example, even though a modern fighter aircraft is basically a few large computers and communication/sensor systems that move in unison, each of those is designed and contracted separately and integrated using the crudest of metrics.
DARPA and the services have other programs for Unmanned Aerial Vehicle (UAV) swarming. They are as limited in vision as what we describe here. The focus here are the candidates for the so-called Long Range Strike Family of Systems. (The Navy has a parallel Multi-Axis Protection of Surface Ships program with a similar vision. The Army, always the late-comer, names its program the Mobility Infantry program and focuses on unmanned ground vehicles. Army research programs often seem dumb, and planners usually have a more reasonable parallel program. In this case it is the UAV-supported Strategic Mobility program to move fighters and material quickly.
You should note that the current generations of UAVs are actually piloted; it is just that the pilots aren’t in the planes. What we are talking about here are autonomous UAVs.
The DARPA/AF idea is that you will have a primary strike aircraft, either the F-35 or the proposed manned/unmanned Long Range Strike Bomber. (The F-35 isn’t designed for this; a next generation general attack aircraft is being prepared under the classified Aerospace Innovation Initiative for the probable collapse of F-35 funding.) The Multi-Domain Unmanned Systems program is supposed to come up with ideas to have solutions that are flexible across services and missions.
You can get a feel here for how the folks in charge segment the problem so that no one can innovate at the systems level.
In the (current) pure vision, the attack aircraft will be accompanied not by the usual protective manned assets, but by a manned aircraft (or two) that serves as controller for the unmanned aircraft in the family. These will be expensive and ideally will be dropped from and recovered to a mother ship, using the Gremlin system. The program to make them generally worthy has the South Park like name: Technology for Enriching Manned-Unmanned Systems (TEAM-US).
So, this vision has it that a manned aircraft controls a master drone, whose job is to manage communications. In the battle airspace, you’ll have three kinds of UAVs and each kind could have several versions:
- Strike. These will carry lethal weapons, as lethal as they come. Some will be to do the main job, and others to take out physical countermeasures, like missile and radar facilities.
- Electronic warfare. We’ll talk about this later. As a percentage of what the Air Force does, this has grown to be very big and very sophisticated. Basically the idea (in tactical engagement) is to damage or spoof the eyes of the enemy.
- Communication. These special purpose UAVs are not called out in the initial plan, but are expected by the contractors. The basic problem is you have to balance computers, sensors, and radars with communication. No single platform (an aircraft is a platform) can do the job well without collaborating with the others. But there cannot ever be enough bandwidth and sending messages around makes you visible and vulnerable.
I should note that these basic ideas are little different than those of ten years ago, though the names have changed and the funding increased. The ideas were deliberately suppressed to protect the F-35. (It is now projected at $1.4T. Trillion!) What’s changed in terms of technology are better aerodynamic systems and vastly better electronic warfare ideas. Another change is that the need is more desperate as current programs are collapsing under the weight of how systems are designed and procured.
Speaking as an educated and concerned observer, this is a broken vision — dangerous and irresponsible in several ways. It will put my children at risk, and yours as well. We can’t fix DARPA; most of that is corporate/university welfare anyway and no one cares to fix it. But we could redirect, improve the RAQ-25 effort.
A Better Possibility
Boeing’s Phantom Works in St Louis has the contract for the RAQ-25. They are good guys in this equation. Boeing also leads the team (with Lockheed Martin) that is the likely winner of the long range strike bomber.
Unfortunately, Boeing’s primary interest is not combatting our adversaries in war, but their competitors for contract dollars. This is not simple as they team with competitors or supply them on some programs, and compete on others. Defense dollars are many but finite so they have constraints of different types to strategically place themselves competitively, win the contract(s) and only incidentally help win a war.
They have little choice; they have to design what the Pentagon wants to buy.
But there may be a confluence of talents and interests in this case. As I say, the Phantom Works guys are good guys and much of what they do is on the company’s dollar.
Raytheon has an interesting position as well because it has a flying system that can be upgraded, the ADM-160B,. It has some flexibility and a very clever set of designers behind it. It is a miniature autonomous electronic warfare drone. And they have the Next Generation Jammer contract.
The Problem Restated
Thinking of the problem abstractly: You need to have massive amounts of communication. You also will have extremely refined notions of your situation in relation to friends and foes, some knowledge of their future situations and the universe of where those situations could go.
Meanwhile, you need to create a false situation that you want to be visible your foes to see. You have to appear to be hiding this false situation in order to be believed.
This sounds like a problem for situation theory.
And it sounds like it will be perfect for an approach which uses space as a storage mechanism for complex and dynamic information structures.
The Big Idea
Back to the Active Electronically Scanned Arrays. What they can do now is send out targeted radar probes, establish narrow, encrypted communication channels (if they know where to do), send spoofing signals though relatively low energy and cancel signals trying to detect it.
The big idea has three components:
- Improve and field virtual decoy radar returns to create the believable illusion of a swarm of aircraft
- Mature these into complex structures that can contain (lightly encrypted) global knowledge of the system’s situation.
- Simultaneously create the ability for each node to absorb and reason about their role (local situation) in the global situation, and project back the processed information.
That is, each node creates a station structure that does a double job:
- Create a space of virtual; returns that successfully does its job in confusing hostile radar. This need not be a spatial structure that makes sense in terms of being a realistic projection.
- Encode in this structure two kinds of information useful to the platforms:
- What the situation is, including instant spatial information about the swarm, but also most information stored locally in each node. When combined this constitutes global knowledge.
- What the connections are among all the bits, encoding the actual functions used in the system that produces the dynamic reasoning.
In other words, essentially all the code and all the data is suspended in the air, consumed and modified fractionally by each node. Each node works with the relevant transforms as understood by their local situating of that structure in the local system.
Everything is there, everything, hidden in a structure that has a different purpose, to deceive. Why would anyone put the storage of the system intelligence in the air, potentially accessible to a foe? The way the math works out, even if it were not encrypted, even if it were wholly intercepted, no one other that the specifically situated nodes could make sense of it.
The math also shows that the dynamism of such a global situation is asymmetric in the reversed direction than usual. It would take massively more computer power and time to unravel the utility than to place the utility.
But of course it is not there for a foe to get globally. It is in an externally deceptive structure.
The reader should be able to see how this applies to biological systems modeling when dealing with complex disease vectors.