Cognitive SDR-RAN adaptation under active adversarial influence

Abstract

Modern SDR-RAN systems require adaptive mechanisms to maintain communication reliability under intelligent adversarial interference. Conventional anti-jamming approaches focus on isolated parameter adaptation, ignoring the interaction between decoding dynamics, topology reconstruction, and distributed control. In practice, adversarial influence becomes an active process that reacts to network behavior, causing signal degradation, higher decoding complexity, and unstable topology reconstruction.
This paper presents a cognitive SDR-RAN adaptation framework for active adversarial environments. It integrates adaptive LDPC decoding control, interference-aware parameter adaptation, topology reconstruction, and minimax optimization within a unified closed-loop architecture. A nonlinear state-space model jointly describes communication parameters, decoding convergence, and interference dynamics, enabling prediction of worst-case scenarios.
The adaptive control algorithm dynamically adjusts transmission parameters and iterative decoding depth based on observed SINR and convergence conditions. A hysteresis-based topology reconstruction mechanism prevents unnecessary reconfigurations and stabilizes routing. Minimax optimization ensures robust adaptation without explicit knowledge of the interference strategy.
Simulation results show that the proposed framework significantly improves reliability and reduces decoding overhead compared to conventional methods. Specifically, it reduces the average number of decoding iterations by 25–30%, improves BER under low SINR, and stabilizes SDR-RAN topology reconstruction under coordinated adversarial influence. These results confirm the effectiveness of joint cognitive control over decoding, parameter adaptation, and network reconfiguration.

Keywords: SDR-RAN, cognitive adaptation, adversarial interference, LDPC decoding, anti-jamming communication, topology reconstruction, minimax optimization, cognitive radio.