Research Article
Reinforcement Learning-Based Routing Protocol to Minimize Channel Switching and Interference for Cognitive Radio Networks
Table 1
Default simulation and structural parameters.
| Parameter | Value |
| Simulation area | 2000 × 2000 m2 | Number of SUs | 100 users/nodes | Number of PUs | 18 maximum | Model for PUs | Uniform model | No. of channels | 10 | Propagation model | Two-way | Round channel bandwidth | 2, 4, 6, 8, 10, 12, 14, 16, 18, 20 MHz | Channel mean packet error rate (PER) | 0.05 packet/ms | Standard deviation of PER (σPER) | 0.025 | Mobility model for SUs | Random waypoint | Mobility speed for SUs | 0–15 m/s (uniform) | Pause time | 0, 60, 120, 180, 240 seconds | Data type | Best effort | Packet payload | 8184 bits | MAC header | 272 bits | Mean arrival rate (MAR) | [0.0, 1.0] | Transmission delay by SU (dtrSU) | 1.0 ms | Processing delay by SU (dprSU) | 1.0 ms | Packets generation at MAR (λSU) | 0.6 packets/ms | Learning rate (α) | 0.1 | Routing weight factor (ω) | 1.2 ms | Transmission delay by PU (dtrPU) | 120 bits | Standard deviation of MAR (σMAR) | 0.0, 0.4, 0.8 | Mobile user traffic | 5 sources | Fixed destinations | 2 static and 3 mobile | Packet size | 15 kb | Transmission rate | 10 packets/sec | Application profile | Low-resolution video |
|
|