Abstract
This paper presents a distributed cognitive Medium Access Control (MAC) protocol designed for multi-hop Mobile Ad Hoc Networks (MANETs) using single-transceiver hardware. By integrating the 802.11 Power Saving Mode (PSM) for time-field separation and introducing a mobility-aware channel maintenance time metric, the protocol addresses challenges in spectrum sensing, link stability, and channel fairness. Key performance indicators include channel vacate time (CVT) and channel opening time (COT).
Introduction
Cognitive Radio (CR) technology is essential for addressing spectrum scarcity in wireless networks. However, most theoretical CR protocols assume nodes possess multiple transceivers—one for continuous spectrum monitoring and another for data transmission. In practical MANET deployments, such as handheld radios or small unmanned aerial vehicles, nodes are typically limited to a single transceiver. Furthermore, existing protocols often struggle with node mobility, multi-hop relaying, and unfair channel allocation. This research presents a distributed cognitive MAC protocol specifically designed for these hardware-constrained, dynamic environments.
Proposed Methodology
The core of the proposed protocol is the use of Time-Field Separation via the IEEE 802.11 Power Saving Mode (PSM). This mechanism divides the beacon interval into two distinct phases:
- ATIM Window (Control Phase): All nodes remain active to perform spectrum sensing, exchange channel state information, and negotiate channel reservations.
- DATA Window (Transmission Phase): Nodes transition to selected data channels for transmission or enter a low-power "doze" mode if no data is scheduled.
Key Innovations
- Channel Maintenance Time ($T_m$): To handle mobility, the protocol estimates the duration a link will remain stable based on the rate of change in received signal power. This allows the system to prioritize channels with higher predicted stability.
- Gray-Coded Quantization: To reduce control overhead, raw signal strength data is quantized into 3-bit Gray codes. This approach reduces overhead by approximately 91% while maintaining sufficient precision for $T_m$ calculations.
- Three-Step Reservation Handshake: Within the ATIM window, nodes execute a Request/Response/Release (RREQ, RRSP, CHREL) handshake to prevent collisions during the subsequent DATA window.
Simulation and Evaluation
The protocol was evaluated using custom simulations in a 5km × 5km environment. Nodes followed a "Random Walk" mobility model at 2m/s. Traffic was modeled as Constant Bit Rate (CBR) at 11Mbps over a 16-channel spectrum using the Dynamic Source Routing (DSR) protocol.
Key Results
- Throughput and Scalability: Throughput increased significantly with the addition of more channels, reaching a saturation point as network density increased.
- Spectrum Fairness: Unlike traditional models that often congest primary channels, the proposed protocol successfully distributed traffic across the available spectrum, enhancing overall utilization.
- Interference Mitigation: Metrics for Channel Vacate Time (CVT) and Channel Opening Time (COT) were maintained within acceptable limits, ensuring that secondary MANET traffic did not disrupt Primary Users.
Conclusion
The proposed distributed cognitive MAC protocol effectively enables cognitive radio capabilities on single-transceiver MANET nodes. By leveraging PSM for synchronization and introducing mobility-aware metrics, the system achieves higher throughput and better fairness. Future work will focus on addressing potential selfish node behavior, improving sensing accuracy, and mitigating the multi-channel hidden terminal problem.
Summary updated