This paper presents Systolic-CNN, an OpenCLdefined scalable, run-time-flexible FPGA accelerator architecture, optimized for performing the low-latency, energy-efficient inference of various convolutional neural networks (CNNs) in the context of multi-tenancy cloud/edge computing. Systolic-CNN adopts a highly pipelined and parallelized 1-D systolic array architecture, which efficiently explores both spatial and temporal parallelism for accelerating CNN inference on FPGAs. SystolicCNN is highly scalable and parameterized, which can be easily adapted by users to achieve 100% utilization of the coarsegrained computation resources (i.e., DSP blocks) for a given FPGA. In addition, Systolic-CNN is run-time-flexible, which can be time-shared, in the context of multi-tenancy cloud or edge computing, to accelerate a variety of CNN models at run time without the need of recompiling the FPGA kernel hardware nor reprogramming the FPGA. The experiment results based on an Intel Arria 10 GX FPGA Development board show that Systolic-CNN, when mapped with a single-precision data format, can achieve 100% utilization of the DSP block resource and an average inference latency of 10ms, 84ms, 1615ms, and 990ms per image for accelerating AlexNet, ResNet-50, RetinaNet, and Light-weight RetinaNet, respectively. The peak computational throughput is measured at 80–170 GFLOPS/s across the acceleration of different CNN models. Codes are available at https://github.com/PSCLab-ASU/SystolicCNN.

Spam is a problem that refuses to go away. An immense amount of time and money is currently devoted to hiding spam, but not enough is devoted to effectively preventing it. CAPTCHAs are a prevalent spam prevention mechanism, but are getting harder for humans to solve and easier for programs to "break". CAPTCHAs also cannot prevent spam from hijacked accounts since they are mostly used during account creation. Proof-of-work approaches are gaining popularity, but current implementations are not effective enough and cannot be used by generic web applications. We present MetaCAPTCHA, an application-agnostic spam prevention service for web applications. MetaCAPTCHA dynamically issues CAPTCHAs and proof-of-work "puzzles" while ensuring that more malicious users solve "harder" puzzles. In order to support its operation, MetaCAPTCHA implements a novel secure protocol -- for authenticating web sites and validating proof-of-work solutions -- that allows for seamless integration across a variety of web applications.

Crowd-sourced mobile embedded systems allow people to contribute sensor data, for critical applications, including transportation, emergency response and eHealth. Data integrity becomes imperative as malicious participants can launch software and Sybil attacks modifying the sensing platform and data. To address these attacks, we develop (1) a Trusted Sensing Peripheral (TSP) enabling collection of high-integrity raw or aggregated data, and participation in applications requiring additional modalities; and (2) a Secure Tasking and Aggregation Protocol (STAP) enabling aggregation of TSP trusted readings by untrusted intermediaries, while efficiently detecting fabricators. Evaluations demonstrate that TSP and STAP are practical and energy-efficient.