CO-MAP: Improving Mobile Multiple Access Efficiency With Location Input

Based on plenty of sensors on smartphones and tablets, their location information becomes more accurate and easy to access. Many position-based applications have thus been developed. As these mobile devices have become the main access approach of 802.11-based wireless local area networks (WLAN), location information also provides opportunities to improve the performance of underlying wireless communication. This paper presents CO-MAP (Co-Occurrence MAP), which leverages location information to handle exposed and hidden terminal problems, which are still a major cause of throughput degradation in mobile WLANs. With the positions of nodes, CO-MAP rapidly builds a co-occurrence map showing which two links can occur concurrently. Meanwhile, to avoid potential collisions, it selects the best settings of frame transmissions based on a novel analytic network model when hidden terminals are distinguished. CO-MAP improves the goodputs of both downlinks and uplinks in an instant and distributed manner. Our implementation on a hardware testbed demonstrates that CO-MAP can accurately detect potential interferers, and provide significant gain of goodput for both exposed and hidden terminal scenarios. The simulation results also suggest that imperfect position hints can still bring substantial improvement on the multiple access efficiency.