Design And Implementation Of A Corruption-Aware Retransmission Protocol For Low-Power Wireless Networks

With the development of low-power technology, wireless sensor networks are becoming widely used. Wireless sensor network are composed of large number of low-power, resource-limited sensor node. Typical application is enviroment protection, bridge health monitoring, etc. Communications among nodes are unreliable in wireless sensor network,especially for low-power network like 802.15.4 Pakcet corruption happens with interference, fading and noise.In this thesis, we investigate the feature of packet corruption with interference in COTS low-power devices. One step further, we propose a RSSI-based corruption detection algorithm to detect the corruption level of a corrupted packet. Based on this corruption level, we design an adaptive coding scheme. This coding scheme is based on Reed solomon codes(RS code). With careful parameter selection, this scheme is targeted to optimize the network performance.The contributions of this paper are summarized as follows:· We propose a novel corruption detection algorithm considering the transitional region of wireless packet reception, resulting in significantly better accuracy compared with state-of-the-art algorithm on COTS devices.· We design an adaptive coding scheme based on Reed Solomon (RS) codes. We carefully select the coding parameters to optimize the network performance on resource-constrained low-power devices.· We design and implement the CARE retransmission protocol. We conduct extensive ex-periments based on COTS low-power devices. Results show that CARE improves the per-formance significantly for highly interfered links while incurring no additional overhead for good links.