We investigate a binary classification problem in an edge intelligence system where false negatives are more costly than false positives. The system features a compact, locally deployed model, supplemented by a larger, remote model that is accessible via the network, albeit at an offloading cost. For each sample, our system first uses the locally deployed model for inference. Based on the output of the local model, the sample may be offloaded to the remote model. This work aims to understand the fundamental trade-off between classification accuracy and the offloading costs within such a hierarchical inference (HI) system. To optimise this system, we propose an online learning framework that continuously adapts a pair of thresholds on the local model's confidence scores. These thresholds determine the prediction of the local model and whether a sample is classified locally or offloaded to the remote model. We present a closed-form solution for the setting where the local model is calibrated. For the more general case of uncalibrated models, we introduce H2T2, an online two-threshold hierarchical inference policy, and prove it achieves sublinear regret. H2T2 is model-agnostic, requires no training, and learns during the inference phase using limited feedback. Simulations on real-world datasets show that H2T2 consistently outperforms naive and single-threshold HI policies, sometimes even surpassing single-threshold offline optima. The policy also demonstrates robustness to distribution shifts and adapts effectively to mismatched classifiers.

REST APIs serve as the core interface of modern distributed systems, where operations are highly interconnected and depend on one another. Many severe vulnerabilities arise from hidden dependencies across multiple API operations, making accurate dependency modeling essential for effective testing. Existing approaches rely on basic producer-consumer relationships or simple parameter matching, which introduce false dependencies, cause inefficient exploration, or miss implicit relationships.

In the energy sector, IoT manifests in the form of next-generation power grids that provide enhanced electrical stability, efficient power distribution, and utilization. The primary feature of a Smart Grid is the presence of an advanced bi-directional communication network between the Smart meters at the consumer end and the servers at the Utility Operators. Smart meters are broadly vulnerable to attacks on communication and physical systems. We propose a secure and operationally asymmetric mutual authentication and key-exchange protocol for secure communication. Our protocol balances security and efficiency, delegates complex cryptographic operations to the resource-equipped servers, and carefully manages the workload on the resource-constrained Smart meter nodes using unconventional lightweight primitives such as Physically Unclonable Functions. We prove the security of the protocol using well-established cryptographic assumptions. We implement the proposed scheme end-to-end in a Smart meter prototype using commercial-off-the-shelf products, a Utility server, and a credential generator as the trusted third party. Additionally, we demonstrate a physics-based attack named load modification attack on the Smart meter to demonstrate that merely securing the communication channel using authentication does not secure the meter, but requires further protections to ensure the correctness of the reported consumption. Hence, we propose a countermeasure to such an attack that goes side-by-side with our protocol implementation.

Formal Concept Analysis (FCA) finds applications in several areas including data mining, artificial intelligence, and software engineering. FCA algorithms are computationally expensive and their recursion tree has an irregular structure. Several parallel algorithms have been implemented to manage the computational complexity of FCA. Most of them assume a shared memory environment where they maintain a shared queue of computational tasks and the workers store and retrieve tasks from that queue. Although the shared queue approach addresses the computation skew by fine grained sharing, it causes communication bottlenecks in a distributed memory environment. In this work, we propose static and dynamic load balancing strategies that are applicable in distributed memory environment. We parallelize the FCA algorithm called Linear time Closed itemset Miner and show that the proposed load balancing strategies effectively deal with the computation skew. They not only distribute the load evenly among the workers but also minimize the communication overhead.