Lightweight Subspace Coding (LSC) for Wireless Sensor Networks

点击翻译按钮获取中文摘要

Canadian wildfires have grown in quantity and severity over the past few decades. In 2015 Canada experienced a 40% increase in the total number of wildfires (5,830 fires) and an increase of 23% in burned area (3,893,861 Hectares), compared to 2014. It is known that Wireless Sensor Networks (WSN) offer early detection of wildfires, however, the deployment of WSN in the wild remains economically prohibitive. Alternatively, small scale deployments to high-risk areas are only feasible if a reliable and flexible solution can be found to overcome the challenges of light weight communications. In the past few years development of network architectures have been characterized by a consistent increase in complexity and sophistication. Since modern protocols must ultimately operate at acceptable speeds in production networks-and alongside one another-the demand for faster and higher capacity hardware has grown considerably. Alternatively, WSN environments require minimum hardware capabilities, minimum protocol structure, and minimum battery consumption; yet, it requires rapid integration with the increasingly elaborate infrastructure built for notably smarter devices. Our industrial partner is focused on studying light weight custom data planes for WSN in order to guarantee resource isolation that enable the utilization of parallel data planes. This project is focused on improving the capacity of generic sensors with less than a linear increase in resource utilization, so the design can scale as the field-programmable gate array (FPGA) capacity continues to increase. This is achievable if the routing layer is replaced by a Lightweight Subspace Coding (LSC) layer. The proposed solution will allow our partner to develop WSN for wildfire monitoring leading to increased client-base in Canada and internationally.