By LTJG Louis A. Catalano, USN

LTJG Louis Catalano is a Cryptologic Warfare Officer attending the Cryptologic Warfare Officer Basic Course. He received his BS in Aeronautical Engineering from Rensselaer Polytechnic Institute in 2020. His current research interest is in aerospace information systems with a focus in unmanned aerial systems, automation, artificial intelligence, GIS, and remote sensing.

In October 2016, the Pentagon’s Strategic Capabilities Office released a video showing three F/A-18 Super Hornets flying in formation. One of the aviators onboard can be heard counting down from five before announcing release. Except, it was not a bomb or missile being dropped from the winged arsenal of these fighter jets; instead, it was a far more terrifying weapon: a drone swarm. The video demonstrated the deployment of 103 swarm drones from specially-designed pods attached to the wings of these F/A-18s. Eventually, the video transitions to a screen capture of a digital map and several tiny plane-like icons scattered and moving around sporadically. Suddenly, the icons begin moving in what appears to be a coordinated effort to a single point on the map (U.S. Department of Defense). It is jarring to see what is evidently each drone flying in such an organized fashion and it is only natural to wonder what this technology is truly capable of and how much it has developed since 2016. Have adversaries of the U.S. been developing this technology as well and if so, what are the countermeasures? And is the naval cryptologic community prepared to defend the fleet from drone swarm attacks especially as conflict with near-peer rivals such as China nears? The purpose of this paper is not to answer these questions but rather, contribute to a discussion that is long overdue.

Drone swarming is fundamentally based on a naturally occurring phenomenon seen in many animals to confuse predators or overwhelm prey. Aerial drone swarming applies this concept to groups of unmanned aerial vehicles (UAVs). There are many civilian applications for aerial drone swarms such as mapping/surveying, agriculture, search & rescue, and even light shows. Though, it is no surprise that the DoD has invested considerable time and resources into developing drone swarm technology for military applications. The Navy was one of the first to develop a drone swarm program which was first announced in 2015 and called the Low-Cost Unmanned Swarming Technology (LOCUST) program specifically intended to launch a swarm from a ship (Smalley). The following year, the Pentagon released the video of its drone swarm being launched from the F/A-18s. Clearly, the DoD has a vested interest in drone swarms but there are several other nations developing swarms as well. Most notably, the Israelis were the first to employ a drone swarm in combat in May 2021. Israel Defense Forces launched an autonomous drone swarm to successfully detect Hamas rocket sites in the Gaza Strait in support of mortar and missile strikes (Kallenborn). The Israelis set a precedent that was seen by the entire world but China in particular was likely taking close notes.

China has been investing heavily in new technology for the People’s Liberation Army (PLA). In August 2014, Xi Jingping announced the new Revolution in Military Affairs which outlined an overhaul of China’s military complex with a focus on developing new technology through military-civil fusion (Chen 2). Unsurprisingly, this has included research into drone swarm technology. In November 2016, China Electronics Technology Group Corporation (CETC) launched 67 swarm drones from a truck with launch tubes (17). A year later, in November 2017, CETC launched 200 drones in a swarm from a truck and helicopter, demonstrating cross-platform deployment of drones in the same swarm. In February 2019, the Chinese tech company, Ziyan, showed off their smart swarm at an exhibition which is a swarm controlled by Artificial Intelligence (AI) (26). Recently, in June 2022, Southern Marine Science and Engineering Guangdong Laboratory announced the Intelligent Mobile Ocean Stereo Observing System which is an AI autonomously-controlled ship capable of deploying several aerial drones and unmanned underwater vehicles in a swarm. Officially, this “mothership” is claimed to be used for oceanographic research but it is likely a dual-use platform (The Marine Executive). It is reasonable to assess that the PLA will use ship-launched drone swarms as a cheap supplement to China’s limited naval aviation capabilities. There are numerous other examples but in less than a decade, China has accelerated its research and development of drone swarm technology to a point of noteworthy lethality. But, what kind of threats do drone swarms pose exactly?

Swarming is used to overwhelm an enemy in a tactic referred to as operational shock (Sander 23). Although by itself drone swarms already pose a significant threat, the technology is best employed within a multi-layered attack (Johnson 28). For example, in a situation where China invades Taiwan, drone swarms may be deployed to overwhelm defense systems either in advance of or in combination with close air support, naval gunfire support, artillery support, and an amphibious assault. In effect, drone swarms are organic force-multipliers that can be used to thicken the fog of war (28). The most commonly considered capability of the drone swarm is its potential for kinetic effects, also known as kamikaze or suicide drones. However, there are several non-kinetic capabilities offered by drone swarming that would likely be applied in combat. Jamming, for example, can be more effective when multiple drones can jam electronic warfare or communications systems from a multitude of shifting vectors. Additionally, drone swarms possess the potential to spoof radar either electro-magnetically or physically by combining radar cross-sections to imitate larger aircraft. Drone swarms could also be employed for ISR and military deception (32). Feinting drone swarm attacks could deceive defenses and allow for the main effort, such as another drone swarm, to deliver the kinetic effect. These are just a few potential applications for drone swarms but as the technology continues to develop and new capabilities are revealed, it is likely that drone swarms will find more applications on the battlefield. So, it beckons the question, how can drone swarms be countered?

In the past few years, there has been significant development of counter-UAV technology; however, these devices are typically designed to take out individual UAVs, not swarms of small UAVs. As an example, NAVSEA is developing a laser to be placed on the decks of ships in order to take out drones. But it is estimated that the system takes about 15 seconds to kill a drone (Sanders 28). Will this be fast enough to take down potentially hundreds of drones in a swarm? Considering other options, the Navy already has kinetic weapon systems on destroyers that may be effective against swarms such as the Phalanx CIWS and the 5-inch cannon with airburst rounds. Nevertheless, kinetic weapons are typically last-resort options and so non-kinetic countermeasures must be considered as well. Jamming or spoofing is an obvious choice when countering drones because if the signal between the swarm and ground control station (GCS) can be disrupted or deceived, this could potentially disarm the swarm. However, with increasing research in AI-autonomously controlled swarms, the necessity for a connection to a GCS decreases, meaning that jamming may not be as dependable (Johnson 33). Sonic countermeasures, on the other hand, are being considered by researchers to stop swarms in air. An example of a sonic countermeasure already onboard Navy vessels is the Long-Range Acoustic Device (LRAD) which is essentially a powerful speaker capable of sending high-intensity sound in a narrow direction. The LRAD could be tuned to the resonant frequency of drone components, like the gyroscope, to degrade coordinated flight (Sanders 473). Additionally, there are some suggestions that acoustic intelligence can assist in detecting and countering drone swarms. Swarms of drones are relatively loud and can sometimes be detected by sound before detection by methods of electronic intelligence collection. “UAS (in any formation – especially Swarms) present detectable acoustic signatures that can be collected in an IFF sound libraries and like fingerprints or DNA they are unique to the make, model and origin manufacturer” (471). Thus, Cryptologic Technicians can be trained to listen and identify swarms by their acoustic signatures. Another countermeasure, and possibly the most viable, resides in the cyber domain. If a drone swarm has an uplink/downlink with a GCS then that signal is potentially exploitable. By acquiring the encryption keys of the data transmitted on these signals through computer network operations, the drone swarm could be hijacked and even told to self-destruct or change targets. But this capability is dependent on the presence of such signals which, as previously stated, may not always be the case. Nevertheless, even if the drone swarm is completely autonomous and has no uplink/downlink, cyber can still play a critical role. If the networks that store the training models of the AI targeting systems can be infiltrated and manipulated, it can potentially cripple the navigation and targeting of AI-controlled drone swarms before they even take off. Cyber is a major countermeasure against drone swarms but may not be sustainable for long periods of conflict since it is only effective for as long as it remains undetected. If the adversary discovers that their networks have been compromised, they may take actions that would prevent future access such as updating operating systems and changing or elevating security measures (Nichols 74). One final countermeasure being seriously considered by the DoD and other nations is to fight fire with fire i.e., using drone swarms to fight other drone swarms. It has been predicted that the future of warfare between near peer adversaries will eventually lead to a battle of AIs where computer algorithms are fighting to outsmart other algorithms. In fact, China is reportedly researching an AI that is specifically designed to destroy other AIs (Chen 534). In an automated world, battles may be completely waged and won with little to no human intervention. So, it is safe to say that whichever nation has the more advanced drone swarm will have the edge on the battlefield.

New drone swarm technology is being released all the time and a simple online search will reveal hundreds of research articles proposing new methods for designing drone swarms. But the most advanced swarms are those being designed to take advantage of AI technology. AI-enabled drones benefit from more advanced autonomy such as automatic targeting. In the case of image recognition software, an AI is trained using thousands of sample pictures of its intended targets so that the algorithm can identify objects based on its own derivations. Visual targeting recognition is already being developed for missiles to distinguish between warships and civilian vessels (16). This is important for the cryptologic community because now ships may need to not only mask their electronic, acoustic, and thermal signatures but visual as well. Some methods could include using smoke but when this technology is integrated into swarms, the threat increases significantly. Consider a swarm of a few hundred drones where a portion of the drones are using thermal targeting, some are using electromagnetic targeting, and the rest are using visual targeting. In most swarms, all it takes is one drone to acquire a target and the rest will follow, meaning ships will have to go to exhaustive lengths to remain hidden from drone swarms. Is the U.S. Navy of today prepared for this level of readiness?

AI is just the tip of the iceberg when it comes to emerging technology in the realm of drone swarms. Edge computing, for example, benefits from the idea that as computer processors become more advanced over time so too does the ability for devices, such as drones, to process complex algorithms, such as AI-targeting, on the edge or in other words, at the location of the device. Traditionally, algorithms like AI require dedicated servers to process and send data to the user but, with edge computing, drones could be executing processes using the combined processing power of the swarm. Again, this removes the need for communication between the swarm and GCS that would otherwise be exploitable. Similarly, cloud computing allows for data to be stored externally from a device. In the context of drone swarms, there has been research in China to create drone clouds which can store data and transport it as needed (17). A potential use-case could include uploading AI-targeting data to a swarm then sending that swarm to a weapon system that isn’t connected to the network to download the data so that it remains hidden from an adversary. Some other innovations in drone swarms being developed by DARPA and adversaries like China include 5G, human-machine cognition, hypersonics, micro-UAVs, and blockchain technologies (Chen 9; Johnson 32). The blockchain, which is under significant research due to its profitability, is unexpectedly finding applications in drone swarms. In its essence, the blockchain is the decentralized storage and secure transfer of data between devices in a network, not unlike a drone swarm (Johnson 265). Lastly, but most importantly, research into the power of drones is what will ultimately propel UAVs, and subsequently drone swarms, forward in terms of capabilities. Drones can be fitted with several different types of engines but it generally depends on size and weight. For drone swarms, which typically feature smaller sized drones, the propulsion is usually electric so speed and range is dependent on battery quality. But, as electric vehicles increase in popularity, new battery technology will continue to emerge and will undoubtedly spill over into the UAV industry. Therefore, UAVs and drone swarms of the future may have extended ranges and the ability to loiter for days, maybe even weeks, awaiting orders. In June 2022, Airbus in partnership with the U.S. Army, launched a drone called the Zephyr which flew continuously for 64 days by taking advantage of advancements in solar power technology (Buchaniec). These are just a few examples of innovations in drone swarm technology that are making swarms more lethal by the day.

The rate at which new technologies, like drone swarms, are being introduced to the world is rapidly outpacing our ability to deeply understand their impact. On the battlefield, these technologies are even more impactful where the cryptologic community is in the best suited role for addressing these new challenges. If it has not already, the community’s understanding of this role needs to start with discussions that explore the drone swarm capabilities of adversaries including China, the potential threats they pose to the fleet and nation, and the best TTPs to counter this threat. The future of warfare will predominantly be fought in the information domain and it is the U.S. naval cryptologic community that must lead the fight against aerial drone swarms.

Works Cited

Buchaniec, Catherine. “Airbus’ Zephyr Drone Test Unexpectedly Halted after Two Months Aloft.” C4ISRNet, C4ISRNet, 23 Aug. 2022, http://www.c4isrnet.com/newsletters/2022/08/23/airbus-zephyr-drone-test-unexpectedly-halted-after-two-months-aloft/.

Chen, Ling. “Testimony before the U.S.-China Economic and Security Review Commission: State and Business Players in the Evolution of China’s Industrial Development.” SSRN Electronic Journal, 2021, doi:10.2139/ssrn.4146967.

“CSSC Launches World’s First ‘Drone Carrier.’” The Maritime Executive, 19 May 2022, maritime-executive.com/article/cssc-launches-world-s-first-drone-carrier.

“Department of Defense Announces Successful Micro-Drone Demonstration.” U.S. Department of Defense, http://www.defense.gov/News/Releases/Release/Article/1044811/department-of-defense-announces-successful-micro-drone-demonstration/.

Johnson, James. “Artificial Intelligence, Drone Swarming and Escalation Risks in Future Warfare.” The RUSI Journal, vol. 165, no. 2, 2020, pp. 26–36., doi:10.1080/03071847.2020.1752026.

Nichols, Randall K., et al. Unmanned Aircraft Systems in the Cyber Domain – Second Edition. New Prairie Press, 2019.

Sanders, Andrew William. “Drone Swarms.” Thesis / Dissertation ETD, Fort Leavenworth, KS: US Army Command and General Staff College, 2017.

Smalley, David. “Locust: Autonomous, Swarming UAVs Fly into the Future.” Office of Naval Research, 18 Mar. 2022, www.nre.navy.mil/media-center/news-releases/locust-autonomous-swarming-uavs-fly-future.