Future Operating Environment
SwarmNet – Heterogeneous Autonomous Multi-Domain Swarms for Network ResilienceBy Daniel Lee May 30, 2019
If the science fiction movies Battle Los Angeles, Battleship and Independence Day have a lesson for modern militaries, it is that technologically advanced forces can come undone if they rely on communications networks that lack redundancy (like a mother ship which acts as a single point of failure), are easily detected (like a gigantic high-power emitter which is unable to be concealed) and are unable to defend themselves, physically or otherwise. Any army looking to realise the potential that a networked force offers should also aim to preserve the resilience of its network, lest it present itself as an Achilles heel.
The Chief of Army’s Strategic Guidance 2019 states that “future conflict is likely to be across domains where networks and integration are the key to generating military power”. Indeed, as The Army continues to develop as a Joint Force integrator and credible coalition partner, with the ability to target ships, aircraft, vehicles and military installations over the horizon, our own forces will become more network-reliant as well as network-enabled. Simultaneously, potential competitors are improving their ability to disrupt, degrade and deny friendly networks. China continues to raise and train Electronic Warfare units through large-scale exercises to improve future readiness for modern warfare. Russian Electronic Warfare forces have already demonstrated their current readiness by disrupting recent NATO exercises and pose a potential threat to Australia’s allies on current operations.
As such, a 'SwarmNet', or Heterogeneous Autonomous Multi-Domain Swarm, may be a type of swarming technology that can give Army an advantage over its enemies in fighting future wars by enhancing network resilience through various means. Firstly, they may improve Network redundancy through meshing, range extension and layering. Secondly, they could achieve deception, hiding important nodes within the network and using expendable assets to emulate vital network assets. Thirdly, they could improve the ability of the network to defend itself through physical or non-physical means.
What is a swarm?
Correl and Rus define a swarm as “a networked robot system... comprised of multiple robots in which robots actively communicate with each other, sensors, or other computational agents using some form of wireless communication”. Swarming technology is inspired by nature; much like insect swarms, a desired collective behaviour emerges from the interaction between the robots and each other as well as the environment. They can be used for various purposes, such as replacing humans in situations of high danger, where large geographical areas must be covered, when a large population effect must be achieved, and where redundancy is required. These are all relevant in the context of Army’s future networks.
A Heterogeneous Autonomous Multi-Domain Swarm (HAMS) in support of a future network may consist of uninhabited vehicles (UVs) of varying function. These may act as transmitters, receivers, range extenders, deceptive emulators, or as a combination. Additionally, the primary role of these UVs may include Intelligence, Surveillance and Reconnaissance (ISR), Electronic Warfare (EW), Strike, Logistics Support or Counter Uninhabited Aerial Vehicle (UAV), with network support as an additional, ad hoc, opportunistic task. The one swarm may consist of UVs of varying size, endurance and persistence, although with a common function in supporting the network. For example, a HAMS could be comprised of high altitude long-endurance UAVs that primarily act as ISR on long-range targets, but could also double as long-range line of sight repeaters for ground forces transiting to the area of operations to provide redundancy for satellite communications for a period of days or weeks. The same HAMS may also be made up of small Uninhabited Ground Vehicles (UGVs) that extend and mesh the same network, but whose primary task is short duration logistics resupply to combat forces during warfighting operations. The HAMS need not be constrained to the physical domains; in urban operations, ‘virtual’ robots may be used to exploit local telecommunications infrastructure such as mobile phone towers and radio stations in order to transmit or relay friendly force communications without the use of overtly military means, thus reducing the signature that the enemy may detect. A swarm of this nature could provide the future Army network with resilience through redundancy, deception and defence.
Unlike traditional meshing and range extension methods that involved manned retransmission vehicles or stations, an Autonomous Swarm provides the ability to cover a large area and provide high-density redundancy without the need to significantly increase the personnel or vehicle size of a contingent. As previously described, a virtual swarm within the cyber domain may have the ability to provide connectivity in urban environments by exploiting existing infrastructure, such as mobile phone towers, public Wi-Fi hotspots and other transmitters. The swarm’s autonomy would allow the network to self-heal at machine speed without human intervention in the event range extension nodes are destroyed, and would self-determine how to support the manoeuvre commander’s operational plan without prompting. In other words, the HAMS-enabled network would be an enabler rather than a limitation on warfighting operations. Self-determined dynamic connections in the absence of human intervention, providing agile, on-demand bandwidth, provisioned with the precise connection needed depending on the type of data in transit, is a critical component of such an autonomous network swarm that provides decision superiority which remains assured at machine speed. The uninhabited nature of the nodes within the swarm have the added benefit of reducing the cost of human lives for the purpose of maintain the network, unlike vulnerable manned retransmission detachments which are generally hard to defend and easy to detect.
The advanced state of modern electronic warfare means that hiding electromagnetic signatures of friendly forces is practically impossible, short of not eschewing the use of all electronic devices and communications. Since avoiding detection is impossible, deception may more likely be achieved through the generation of false or misleading electromagnetic emissions – a modern version of Imitative Communications Deception, where friendly forces imitate the communications of the enemy to deceive them, as demonstrated during Operation CORONA during the Second World War. Artificial Intelligence (AI), specifically Generative Adversarial Networks, could be employed to emulate signals that appear as something else, such as a lower-value target, an adversary emitter, or multiple platforms. In a future where other nations aim to become world leaders in the use of AI, autonomous swarms may be necessary to be able to generate deception not only of the required quality, quantity and disposition, but also at the speed required to maintain decision superiority against technologically advanced adversaries. Even if an autonomous swarm could deceive an enemy as to the disposition, magnitude and type of friendly network capabilities within an area of operations, the effect may only be temporal. Consequently, thought must also be given as to how to defend the network once it is found.
The use of Cognitive Electronic Warfare for electronic attack and defence has been explored by the US’ Defense Advanced Research Projects Agency’s (DARPA’s) Behavioral Learning for Adaptive Electronic Warfare (BLADE) and the US Army’s Rapid Capabilities Office’s Cognitive EW competition (of which an Australian team came second place). While the endeavours examined the use of AI to execute EW through software defined radios and analytical tools at machine speed, the next logical step is to extend that function into the physical domains. An autonomous networked swarm could then actively defend itself through pre-emptively positioning emitter and range extension nodes to mitigate anticipate communications degradation, or cue kinetic offensive action against enemy jammers and intercept stations either from another capability system such as precision long-range fires, or from one of the multi-role platforms within the swarm itself. This will increase not only the difficulty but also the potential danger for adversaries when targeting Army’s future networks.
Army needs to be ready for a future where networks are key to generating military power across domains, and where adversaries constantly seek to disrupt or exploit those networks. ‘SwarmNets’, or Heterogeneous Autonomous Multi-Domain Swarms have the potential to provide network resilience though redundancy, deception and defence at machine speed without the requirement for human intervention, thereby enabling decision-making superiority.