| Long-distance wireless mesh network is a new type of wireless access networks, which adopts the off-the-shelf IEEE 802.11 wireless cards and high-gain directional antennas to achieve long-distance links. IEEE 802.11 long-distance wireless mesh networks can provide a low-cost connectivity solution for rural areas, especially in sparse-popularity and low-income regions. In such networks, links are point-to-point, thus MAC protocol is type of TDMA instead of CSMA/CA. Link quality is time-variant suffering external WiFi interference Energy efficiency is also critical since the nodes are usually powered by solar batteries. In view of the above issues, adaptive transmission rate adaptation considering energy efficientcy is a key problem in long-distance wireless mesh networks.A joint transmission rate and power adaptation algorithm is proposed in this paper, which operates on top of a TDMA-like MAC protocol. In this algorithm, the FDR(Frame Delivery Ratio)-RSSI(Received Signal Strength Indicator) mapping is measured by sending probing frames with different rates and power. The transmission rate and power are chosen jointly considering the transmit power in the last slot and RSSI values predicted by the sender. However, the FDR-RSSI mapping is affected by external WiFi interference, therefore an algorithm based on CUSUM (CUmulative SUM) is proposed to detect interference intensity leveraging beacon loss ratio statistics. When interference intensity exceeds a threshold, the receiver informs the sender to re-estimate the FDR-RSSI mapping.The transmission rate and power adaptation algorithm is implemented and evualated on our long-distance wireless mesh network test-bed, which consists of several long-distance links up to 20km. The experiment results show that it significantly outperforms fixed transmission rate and power schemes in terms of energy efficiency and link throughput. |