Effective Scheduling Algorithms for Cross-Interference Mitigation in Heterogeneous Wireless Networks

Wireless networks are making life easier, smarter and more convenient. However, the well-known Carrier-sense multiple access with collision avoidance (CSMA/CA) mechanism is powerless when dealing with Cross-Technology Interference (CTI) between Wi-Fi and Low-Rate Wireless Personal Area Network (LR-WPAN), because of asymmetric transmission power, incompatible Clear Channel Assessment (CCA) and different timing parameters. Plenty of studies have shown that WiFi always has a higher priority to access the wireless medium and even block LR-WPAN transmission in the worst case. Our experiments confirm this point and conclude that Wi-Fi can interrupt LR-WPAN severely even block LR-WPAN traffic, while the interference from LR-WPAN to Wi-Fi is negligible. Different from other studies, this thesis presents a novel centralized scheduling mechanism in the time domain to harmonize coexistence of Wi-Fi and LR-WPAN, also refer to as time-slot based scheduling mechanism. The mechanism is achieved by introducing a new command frame, named Access Notification (AN), into the IEEE802.15.4 Medium Access Control (MAC) layer. Based on this mechanism, a static time-slot based scheduling algorithm is designed and evaluated on both real hardware-based system and NS-3 simulator. The result shows the algorithm improves LR-WPAN Packet Loss Rate (PLR) significantly but at the cost of reducing Wi-Fi throughput. In order to maximize performance, based on slot-based congestion indicator (CI) that is proposed and defined to tell whether an allocated time slot is adequate for data transmission or not, we further design an adaptive time-slot based scheduling algorithm. The evaluation shows that the adaptive algorithm covers the shortage of the static algorithm and offers a distinct improvement on LR-WPAN Packet Transmission Rate (PTR).