Breaking News

Faster Network Decision Making with AI/ML – Signal Magazine

The availability of highly capable, small footprint radio systems designed for mobility has allowed military forces to deploy tactical communications networks virtually anywhere. As the military shifts its focus from counterinsurgency wars to peer conflicts, the opportunities for expanding networking reach in the operational theatre are becoming a reality. Advances in mobile technology means the network will generally expand in scale and complexity. To fully leverage and deploy mobile networks, militaries need to have technology which can help operators to extend, bridge and protect the network.Expanding networks will add significant communications capability, but also increase risk and technical sustainment challenges. Due to closer proximity to near-peer threat environments, the network will be more susceptible to jamming and interference. Forces on the move at the tactical edge usually do not have the technical resources of a battalion or brigade level command post for network troubleshooting. Mobile networks at the Lower Tactical Tier (LTT) will need to integrate with fixed and semi-fixed line-of-sight (LOS) networks at the upper tactical tier (UTT) since the mobile network is only as good as its ability to connect to the backhaul network.  

In the simplest of terms, tactical networks will be much more complex to manage and setting up the initial network will present only part of the challenge. Threats and network issues can appear out of nowhere and human operators, as well as the systems themselves, need to be able to respond quickly with solutions. In near-peer conflicts, tactical networks will be attacked by intelligent jammers leveraging state-of-the-art artificial intelligence (AI)/machine learning (ML) technologies in multiple bands. New jamming platforms on UxVs or loitering munitions will also increase the threat to the network.

Making intelligent use of a diversity of frequency bands, network topologies, channels, routes and waveforms is a complex endeavour even in noncombat environments. AI/ML-enabled cognitive radio and network automation will be critical for future tactical communication systems and networks to operate in highly congested and contested environments. Cognitive enhancements will be needed for performance optimization, ease-of-use, flexibility and resilience to advanced jammers and general interference. The future fight for electromagnetic spectrum dominance will involve intelligent machines capable of split-second decisions. The concept of self-healing, adaptive networks is nearing reality due to advances in AI/ML technology that can sense and evolve without human intervention.  

Not all network troubleshooting can be handled by AI/ML of course. A broken piece of equipment will always need to be physically replaced by a human. However, when it comes to network management software applications, the technology can be highly useful to rapidly apply interventions, which the operator would do 100% of the time, as an example. These types of interventions can be setup to intervene automatically and even remotely. AI/ML can also be employed as a decision-assist tool by providing operators with choices and recommended actions.  In summary, there are many ways AI/ML can help in an automated, semi-automated and human-in-the-loop capacity with the goal of making faster networking decisions.

In 2019, Ultra created a collaborative research institute with several universities and research centers. The Resilient Machine Learning Institute (ReMI) has since been developing cognitive capabilities that mission critical networks will require to adapt rapidly or autonomously to changing field conditions. This research is being developed for integration in current and future Ultra ORION radios, as well as future third-party systems used at the tactical edge to favor a unified and open network architecture approach. As a result, Ultra is well, if not uniquely positioned to provide real, verifiable, and tangible AI/ML features for tactical communications applications.  Visit https://www.ultra.group/gb/our-business-units/intelligence-communications/communications/cognitive-tactical-network/ for more information about Ultra’s Cognitive Tactical Network capabilities.
Source: https://www.afcea.org/signal-media/faster-network-decision-making-ai/ml