No abstract

Power consumption and performance are the two main concerns in designing pulse‐triggered level‐converting flip‐flops (LCFF). In this letter, through the research on the level conversion circuits, an explicit‐pulsed dual‐edge triggered LCFF based on the carbon nanotubes field‐effect transistors (CNTFETs) is proposed by using the multi‐source voltage technology. The proposed flip‐flop increases the speed by reducing the number of transistors in the charge/discharge path. The characteristics of the CNTFETs are used to obtain transistors with different threshold voltages by changing the chiral vector to optimize the circuit structure and further reduce the power consumption. In addition, the power consumption is further reduced due to the small on‐current of the CNTFETs. The proposed novel structure is simulated in HSPICE using the Stanford model. Simulation results indicate that our proposed flip‐flop reduces 62.9% to 72.4% of power consumption and 73.6% to 89.1% of delay in comparison with other flip‐flops proposed in the references at 50% data switching activity.

No abstract

In recent years, location big data (LBD) has become an important resource for intelligent transportation systems. LBD mining faces many serious questions, in order to tackle these questions, we introduce deep learning algorithm (DLA) into LBD mining. First, we explore the current research on deep neural networks (DNN) and DLA. Then, we design a framework that uses DLA in LBD mining from the view of probability. Finally, we provide some research trends of DLA after analysis.