In single-user MIMO communication, the first-order throughput scaling is determined by the smallest of the number of transmit and receive antennas. This typically renders terminals the constraining bottleneck. In a multiuser downlink, this bottleneck can be bypassed by having the base station communicate with multiple terminals simultaneously, in which case the receive antennas at those termina...

We propose a framework which ports Dirichlet Gaussian mixture model (DPGMM) based labels to deep neural network (DNN). The DNN is used to extract a low-dimensional unsupervised speech representation, named as unsupervised bottleneck feature (uBNF), which captures considerable information for sound cluster discrimination. We investigate the performance of uBNF in query-by-example spoken term det...

Among the major challenges of transitioning to exascale in HPC is the ubiquitous I/O bottleneck. For analysis and visualization applications in particular, this bottleneck is exacerbated by the write-onceread-many property of most scientific datasets combined with typically complex access patterns. One promising way to alleviate this problem is to recognize the application’s access patterns and...

This work presents a novel hardware/software codesign approach to blob detection. Camera images are preprocessed on a field-programmable gate array that separates foreground pixels from the background. The foreground pixels are transferred to the PC over a 100 Mbit/s Ethernet interface by using a custom protocol. The PC reconstructs the images from received foreground pixels and extracts center...

We present distributed randomized leader election protocols for multi-hop radio networks that elect a leader in almost the same time TBC required for broadcasting a message. For the setting without collision detection, our algorithm runs with high probability in O(D log n D + log n) · min{log logn, log n D } rounds on any n-node network with diameter D. Since TBC = Θ(D log n D + log n) is a low...