Search for: All records

We characterize the asymptotic performance of nonparametric goodness of fit testing. The exponential decay rate of the type-II error probability is used as the asymptotic performance metric, and a test is optimal if it achieves the maximum rate subject to a constant level constraint on the type-I error probability. We show that two classes of Maximum Mean Discrepancy (MMD) based tests attain this optimality on Rd, while the quadratictime Kernel Stein Discrepancy (KSD) based tests achieve the maximum exponential decay rate under a relaxed level constraint. Under the same performance metric, we proceed to show that the quadratic-time MMD based two-sample tests are also optimal for general two-sample problems, provided that kernels are bounded continuous and characteristic. Key to our approach are Sanov’s theorem from large deviation theory and the weak metrizable properties of the MMD and KSD. 

In this paper, we propose a secure lightweight and thing-centered IoT communication system based on MQTT, SecT, in which a device/thing authenticates users. Compared with a server-centered IoT system in which a cloud server authenticates users, a thing-centered system preserves user privacy since the cloud server is primarily a relay between things and users and does not store or see user data in plaintext. The contributions of this work are three-fold. First, we explicitly identify critical functionalities in bootstrapping a thing and design secure pairing and binding strategies. Second, we design a strategy of end-to-end encrypted communication between users and things for the sake of user privacy and even the server cannot see the communication content in plaintext. Third, we design a strong authentication system that can defeat known device scanning attack, brute force attack and device spoofing attack against IoT. We implemented a prototype of SecT on a $10 Raspberry Pi Zero W and performed extensive experiments to validate its performance. The experiment results show that SecT is both cost-effective and practical. Although we design SecT for the smart home application, it can be easily extended to other IoT application domains. 

We investigate the nonparametric, composite hypothesis testing problem for arbitrary unknown distributions in the asymptotic regime where both the sample size and the number of hypothesis grow exponentially large. Such asymptotic analysis is important in many practical problems, where the number of variations that can exist within a family of distributions can be countably infinite. We introduce the notion of discrimination capacity , which captures the largest exponential growth rate of the number of hypothesis relative to the sample size so that there exists a test with asymptotically vanishing probability of error. Our approach is based on various distributional distance metrics in order to incorporate the generative model of the data. We provide analyses of the error exponent using the maximum mean discrepancy and Kolmogorov–Smirnov distance and characterize the corresponding discrimination rates, i.e., lower bounds on the discrimination capacity, for these tests. Finally, an upper bound on the discrimination capacity based on Fano's inequality is developed. Numerical results are presented to validate the theoretical results. 

Microelectrodes are widely used as a peripheral nerve interface (PNI) to connect the peripheral nerve to a computer for restoration of sensorimotor function and bionic device motion control. Materials used for implantable microelectrode are still facing the challenges from biocompatibility and bio‐fidelity in neural signal recording and nerve stimulating. In this study, we report that carbon multi‐electrode arrays (cMEAs) can be fabricated using carbon ink, micro resin dimethylsiloxane and 3D printing technology and ink for PNI. In vitro cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) demonstrated that the cMEAs have higher charge storage capacity (CSC) and less impedance than conventional platinum (Pt) electrode. In vivo studies using an animal model demonstrated that cMEAs are more effective in stimulating the nerve to elicit muscle contraction and recording compound muscle action potentials (CMAPs) than the Pt electrode. The cMEAs has lower stimulating threshold to elicit muscle activity, higher signal‐to‐noise (SNR) in CMAP. Our studies demonstrate that cMEAs can be an advanced healthcare materials in nerve signal nerve stimulation for PNI and muscle bioelectrical signal recording for peripheral muscle interface (PMI).