Cognitive Radio (CR)

The foundational awareness mechanism. A CR must reliably detect underutilized spectrum holes (white spaces) and, more critically, the reappearance of a Primary User (PU) to avoid harmful interference.

• Techniques: Energy detection, matched filter detection, and cyclostationary feature detection.
• Key Challenge: The Sensing-Throughput Tradeoff, where longer sensing times improve detection accuracy but reduce transmission windows.
• Advanced Approach: Cooperative sensing fuses data from multiple nodes to overcome hidden node problems and multipath fading.

The intelligent core that selects the optimal frequency, power, and modulation scheme. This is typically modeled as a Markov Decision Process (MDP) or a Partially Observable MDP (POMDP) due to sensing uncertainty.

• Reinforcement Learning (RL): Algorithms like Deep Q-Networks (DQN) and Proximal Policy Optimization (PPO) learn optimal access policies by balancing the Exploration-Exploitation Trade-off.
• Multi-Armed Bandit (MAB): A simplified model for channel selection when state transitions are not explicitly modeled.

The ability to seamlessly vacate a channel when a Primary User (PU) reclaims it and transition to a backup band without dropping the connection. This is a non-negotiable requirement for Secondary Users (SU).

• Proactive Handoff: Pre-selecting target channels based on predicted vacancy duration to minimize latency.
• Protocol Stack Adaptation: The handoff requires cross-layer coordination to adjust MAC and network layer parameters to the new channel's characteristics instantly.

The physical realization of the cognitive engine's decisions. A CR uses Software-Defined Radio (SDR) hardware to dynamically alter its operating parameters without physical intervention.

• Reconfigurable Parameters: Carrier frequency, transmit power, modulation scheme (e.g., QPSK to 64-QAM), and channel coding rate.
• Goal: Maintain Quality of Service (QoS) while operating on the best available spectrum, often using Digital Pre-Distortion (DPD) to maintain linearity across varying power levels.

The capability that elevates a radio from adaptive to cognitive. The system learns from past decisions and environmental history to improve future performance.

• Experience Replay: Stores past state-action-reward tuples to break temporal correlations during training.
• Anti-Jamming RL: Learns adaptive frequency hopping patterns to evade malicious jammers without pre-programmed scripts.
• Safe RL: Incorporates constraints to ensure the agent never explores actions that cause interference to incumbents.

The foundational awareness mechanism of the cognitive cycle. Spectrum sensing is the process by which a cognitive radio monitors the electromagnetic environment to detect the presence or absence of primary user signals. Key techniques include:

• Matched filter detection: Optimal when signal characteristics are known.
• Energy detection: A blind method with low computational complexity.
• Cyclostationary feature detection: Exploits periodic signal statistics for robust classification in low SNR conditions.

The dominant machine learning paradigm for building the cognitive engine's decision-making policy. RL is a framework where an agent learns an optimal policy by interacting with an environment and receiving scalar rewards or penalties. In a CR context, the agent learns to select frequencies and adjust parameters to maximize throughput while avoiding collisions, without requiring explicit supervision or a pre-programmed model of the RF environment.

An integrated spatial-spectral database that provides a cognitive radio with comprehensive situational awareness beyond its local sensing capability. A REM aggregates multi-domain information—including spectrum occupancy, terrain features, and transmitter locations—to enable informed spectrum decisions. It serves as a long-term memory for the cognitive engine, allowing it to predict coverage holes and interference patterns.

The hierarchical relationship that defines all spectrum access logic. A Primary User (PU) is the licensed incumbent with exclusive statutory rights to a frequency band and must be protected from harmful interference. A Secondary User (SU) is an unlicensed or lower-priority device that opportunistically accesses spectrum holes on a non-interfering basis, vacating the channel immediately upon detection of a PU transmission.

The capability of a cognitive radio to seamlessly vacate its current operating frequency and transition to an alternative vacant band when a primary user reclaims the channel. This spectrum handoff process must maintain uninterrupted communication by executing a predefined channel selection policy that minimizes latency and prevents service disruption, making it a critical component of the cognitive cycle's action phase.