Energy-aware network coding circuit and system design
Author(s)Angelopoulos, Georgios, Ph. D. Massachusetts Institute of Technology
Massachusetts Institute of Technology. Dept. of Electrical Engineering and Computer Science.
Muriel Médard and Anantha P. Chandrakasan.
MetadataShow full item record
Network Coding (NC) has been shown to provide several advantages in communication networks in terms of throughput, data robustness and security. However, its applicability to networks with resource constrained nodes, like Body Area Networks (BANs), has been questioned due to its complexity requirements. Proposed NC implementations are based on high-end CPUs and GPUs, consuming hundreds of Watts, without providing enough insight about its energy requirements. As more and more mobile devices, sensors and other low power systems are used in modern communication protocols, a highly efficient and optimized implementation of NC is required. In this work, an effort is made to bridge NC theory with ultra low power applications. For this reason, an energy-scalable, low power accelerator is designed in order to explore the minimum energy requirements of NC. Based on post-layout simulation results using a TSMC 65nm process, the proposed encoder consumes 22.15 uW at 0.4V, achieving a processing throughput of 80 MB/s. These numbers reveal that NC can indeed be incorporated into resource constrained networks with battery-operated or even energy scavenging nodes. Apart from the hardware design, a new partial packet recovery mechanism based on NC, called PPRNC, is proposed. PPRNC exploits information contained in partial packets, similarly to existing Hybrid-ARQ schemes, but with a PHY-agnostic approach. Minimization of the number of retransmitted packets saves transmission energy and results in higher total network throughput, making PPRNC an attractive candidate for energy constrained networks, such as BANs, as well as modern, high-speed wireless mesh networks. The proposed mechanism is analyzed and implemented using commercial development boards, validating its ability to extract information contained from partial packets.
Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2011.Cataloged from PDF version of thesis.Includes bibliographical references (p. 73-78).
DepartmentMassachusetts Institute of Technology. Dept. of Electrical Engineering and Computer Science.
Massachusetts Institute of Technology
Electrical Engineering and Computer Science.