SIG Chair, Prof. Andy Wu
Connected context computing includes the research areas in sensing, connectivity, computing and applications. For sensing, the first goal is to develop high-efficiency, low-power, and low-cost chips and modules. Therefore, SIGGSP (Green Sensing Platform) intends to address the development, design, implementation, and optimization issues for smart sensor nodes (in contrast to the so-called dust nodes) for high-end sensor-network applications. As shown in figure 5, a sensor node consists of four major components: a processing unit (processor and memory element), a sensing unit (sensor and ADC), a transceiver unit (receiver and transmitter), and a power unit. To meet the future demand of M2M, SIGGSP aims to develop intelligent senor nodes (i-Sensor) by adding new features to current sensor node technology: the processing unit will be augmented with video coding and analysis units; the sensing unit will have various integrated sensors; the transceiver unit will be designed for ultra-low-power operations; and the power unit will be capable of capturing and storing energy from external sources in the ambient environment. The following research projects are being conducted:
The aim of this project is to develop coding and analysis subsystems for distributed video sensors that employ distributed video coding techniques to scale with the energy consumption/complexity of the sensor and aggregation nodes. A three- to four-fold improvement in power efficiency is expected for different platforms, including ASIC-based platforms or processor-based platforms.
Sensor node structure. (a) Traditional sensor node；(b) Intelligent sensor node (iSensor).