A Batch Private Information Retrieval (batch-PIR) scheme allows a client to retrieve multiple data items from a database without revealing them to the storage server(s). Most existing approaches for batch - Pirare based on batch codes, in particular, probabilistic batch codes (PBC) (Angel et al. S&P'18), which incur large storage overheads. In this work, we show that zero storage overhead is achievable for tree-shaped databases. In particular, we develop TreePIR, a novel approach tailored made for private retrieval of the set of nodes along an arbitrary root-to-leaf path in a Merkle tree with no storage redundancy. This type of tree has been widely implemented in many real-world systems such as Amazon DynamoDB, Google's Certificate Transparency, and blockchains. Tree nodes along a root-to-leaf path forms the well-known Merkle proof. TreePIR, which employs a novel tree coloring, outperforms PBC, a fundamental component in state-of-the-art batch-PIR schemes (Angel et al. S&P'18, Mughees-Ren S&P'23, Liu et al. S&P'24), in all metrics, achieving 3 ×lower total storage and 1.5-3 ×lower computation and communication costs. Most notably, TreePIR has 8-160× lower setup time and its polylog-complexity indexing algorithm is 19–160 ×faster than PBC for trees of 210_224leaves.

For autonomous vehicles (AVs) to navigate safely and reliably in unpredictable environments, ensuring the robustness of their perception systems is critical. This contrasts with conventional performance evaluations, which typically focus on accuracy (i.e., how ‘good’ a system is with known input). Existing robustness metrics, however, are fundamentally limited by their heavy reliance on large, fully annotated datasets, making them costly and impractical for scalable, real-time safety monitoring. This study addresses this limitation by introducing a novel robustness measure that can be applied effectively even when ground-truth annotations are very limited. The significance of our approach lies in its self-referential evaluation: instead of relying entirely on ground truth, it primarily assesses a perception model by comparing the output of a degraded input against a baseline output generated by the same model on the original clean input. This measure is calibrated using a correction factor derived from a small, available labeled subset, ensuring consistency with established metrics while removing the need for continuous, large-scale annotation. To facilitate this evaluation, we develop a novel testing framework that systematically introduces realistic input degradations—such as variations in lighting, camera artifacts, and adversarial perturbations—across six AV perception models, including two camera or LiDAR-only (i.e. uni-modal), and two multi-sensor data fusion (i.e. multi-modal) systems. Our results demonstrate that our limited-data measure highly correlates with full-data ground-truth metrics, confirming its reliability. We further discuss the insights gained from applying the robustness measure to these deep learning models against different degradation categories. To our best knowledge, the proposed measure is the first in the relevant literature to provide insights into the robustness of the AV perception systems across complex environments without a continuous reliance on the ground-truth annotations. The testing framework and robustness measure developed in this paper provide a practical tool for the continuous safety monitoring and improvement of AVs in their real-time operation.

Distributed computing, in which a resource-intensive task is divided into subtasks and distributed among different machines, plays a key role in solving large-scale problems.Coded computingis a recently emerging paradigm where redundancy for distributed computing is introduced to alleviate the impact of slow machines (stragglers) on the completion time. We investigate coded computing solutions over elastic resources, where the set of available machines may change in the middle of the computation. This is motivated by recently available services in the cloud computing industry (e.g., EC2 Spot, Azure Batch) where low-priority virtual machines are offered at a fraction of the price of the on- demand instances but can be preempted on short notice. Our contributions are three-fold. We first introduce a new concept calledtransition wastethat quantifies the number of tasks existing machines must abandon or take over when a machine joins/leaves. We then develop an efficient method to minimize the transition waste for the cyclic task allocation scheme recently proposed in the literature (Yang et al. ISIT’19). Finally, we establish a novel solution based on finite geometry achievingzerotransition wastes given that the number of active machines varies within a fixed range.

Motivated by recently available services in the cloud computing industry, e.g., EC2 Spot or Azure Batch, where spare/low-priority virtual machines are offered at a fraction of the price of the on-demand instances but can be preempted on short notice, we investigate coded computing solutions over elastic resources, where the set of available machines may change in the middle of the computation. Our contributions are two-fold: We first propose an efficient method to minimize the transition waste, a newly introduced concept quantifying the total number of tasks that existing machines have to abandon or take on anew when a machine joins or leaves, for the cyclic elastic task allocation scheme recently proposed in the literature (Yang et al. ISIT'19). We then proceed to generalize such a scheme and introduce new task allocation schemes based on finite geometry that achieve zero transition wastes as long as the number of active machines varies within a fixed range. The proposed solutions can be applied on top of existing coded computing schemes tolerating stragglers.