A review article in the journal National Science Review examines how this complex electromagnetic environment, or CEME, is straining traditional radio propagation and interference models. The authors, led by researchers at the Beijing Institute of Technology with collaborators from Nanyang Technological University and Aristotle University of Thessaloniki, describe how static modeling methods cannot keep pace with the rapid growth of communication nodes, from low Earth orbit satellites to urban internet of things devices.

The team proposes a move from passive monitoring of signals to active electromagnetic situation awareness, in which the network continuously builds and updates a picture of spectrum use and interference patterns. To enable this, they outline a three in one modeling strategy that combines digital twins, agent based modeling, and artificial intelligence into a unified framework.

In this scheme, digital twins provide high fidelity virtual replicas of physical communication environments so that system behavior can be simulated and tested before or alongside real world operations. Agent based modeling represents individual nodes, such as unmanned aerial vehicles or satellites, as autonomous agents whose interactions can be explored in large scale simulations to reveal emergent interference and coordination patterns.

Artificial intelligence algorithms then operate on the data streams produced by these models and by live networks to detect anomalies, predict interference, and recommend spectrum management actions. The authors describe this combined framework as analogous to a nervous system for SAGIN architectures, allowing the system to not only observe signal activity but also infer the intent and behavior behind transmissions.

The review stresses that electromagnetic situation awareness must be tailored to the distinct conditions found in different domains of a space air ground system. In space, the framework must account for high speed orbital dynamics, line of sight propagation over long distances, and background cosmic noise that affects signal reception.

In the air domain, the modeling must capture the motion of aircraft and drones, rapid topology changes in ad hoc aerial networks, and intermittent connectivity as platforms move in and out of coverage. On the ground, models must address dense urban deployments, multipath fading from buildings and terrain, and the coexistence of many systems and standards in limited frequency bands.

Looking ahead, the authors identify the integration of advanced AI techniques as critical to making electromagnetic situation awareness practical at 6G scale. One direction is the use of generative AI to produce synthetic data that represent rare or data scarce conditions in orbital and airborne environments, improving model training where real measurements are limited or costly.

Another direction is the development of semantic aware cognitive networks that interpret not only the raw signal properties but also the meaning and context of transmitted information. Such networks could adjust spectrum allocation policies and routing strategies based on application needs, user priorities, and security requirements inferred from traffic patterns.

The review positions this three in one modeling roadmap as a basis for engineering future SAGIN deployments that can treat the complex electromagnetic environment as a managed resource rather than an uncontrollable source of interference. By combining digital twins, agent based modeling, and AI driven analysis, the authors argue that 6G era networks can maintain reliable and secure connectivity across space, air, and ground domains under crowded spectrum conditions.

Research Report:Electromagnetic Situation Awareness and Modeling for Space-Air-Ground Integrated Networks

Related Links
Science China Press
The latest information about the Commercial Satellite Industry