arXiv subjects

Arindam Khanda

Publications and source records attributed to Arindam Khanda.

6 recordsLinked to original sources

arXiv2026

ESCHER: Efficient and Scalable Hypergraph Evolution Representation with Application to Triad Counting

Higher-order interactions beyond pairwise relationships in large complex networks are often modeled as hypergraphs. Analyzing hypergraph properties such as triad counts is essential, as hypergraphs can reveal intricate group interaction patterns that conventional graphs fail to capture. In real-world scenarios, these networks are often large and dynamic, introducing significant computational challenges. Due to the absence of specialized software packages and data structures, the analysis of large dynamic hypergraphs remains largely unexplored. Motivated by this gap, we propose ESCHER, a GPU-centric parallel data structure for Efficient and Scalable Hypergraph Evolution Representation, designed to manage large scale hypergraph dynamics efficiently. We also design a hypergraph triad-count update framework that minimizes redundant computation while fully leveraging the capabilities of ESCHER for dynamic operations. We validate the efficacy of our approach across multiple categories of hypergraph triad counting, including hyperedge-based, incident-vertex-based, and temporal triads. Empirical results on both large real-world and synthetic datasets demonstrate that our proposed method outperforms existing state-of-the-art methods, achieving speedups of up to 104.5x, 473.7x, and 112.5x for hyperedge-based, incident-vertex-based, and temporal triad types, respectively.

cs.DC↗

arXiv2026

DynLP: Parallel Dynamic Batch Update for Label Propagation in Semi-Supervised Learning

Semi-supervised learning aims to infer class labels using only a small fraction of labeled data. In graph-based semi-supervised learning, this is typically achieved through label propagation to predict labels of unlabeled nodes. However, in real-world applications, data often arrive incrementally in batches. Each time a new batch appears, reapplying the traditional label propagation algorithm to recompute all labels is redundant, computationally intensive, and inefficient. To address the absence of an efficient label propagation update method, we propose DynLP, a novel GPU-centric Dynamic Batched Parallel Label Propagation algorithm that performs only the necessary updates, propagating changes to the relevant subgraph without requiring full recalculation. By exploiting GPU architectural optimizations, our algorithm achieves on average 13x and upto 102x speedup on large-scale datasets compared to state-of-the-art approaches.

cs.DC↗

arXiv2026

ReACT-TTC: Capacity-Aware Top Trading Cycles for Post-Choice Reassignment in Shared CPS

Cyber-physical systems (CPS) increasingly manage shared physical resources in the presence of human decision-making, where system-assigned actions must be executed by users or agents in the physical world. A fundamental challenge in such settings is user non-compliance: individuals may deviate from assigned resources due to personal preferences or local information, degrading system efficiency and requiring light-weight reassignment schemes. This paper proposes a post-deviation reassignment framework for shared-resource CPS that operates on top of any initial allocation algorithm and is invoked only when users diverge from prescribed assignments. We advance the Top-Trading-Cycle (TTC) mechanism to enable voluntary, preference-driven exchanges after deviation events, and extend it to handle many-to-one resource capacities and unassigned resource conditions that are not supported by the classical TTC. We formalize these structural cases, introduce capacity-aware cycle-detection rules, and prove termination along with the preservation of Pareto efficiency, individual rationality, and strategy-proofness. A Prospect-Theoretic (PT) preference model is further incorporated to capture realistic user satisfaction behavior. We demonstrate the applicability of this framework on an electric-vehicle (EV) charging case study using real-world data, where it increases user satisfaction and effective assignment quality under non-compliant behavior.

cs.GT↗

arXiv2025

CARGO: A Co-Optimization Framework for EV Charging and Routing in Goods Delivery Logistics

With growing interest in sustainable logistics, electric vehicle (EV)-based deliveries offer a promising alternative for urban distribution. However, EVs face challenges due to their limited battery capacity, requiring careful planning for recharging. This depends on factors such as the charging point (CP) availability, cost, proximity, and vehicles' state of charge (SoC). We propose CARGO, a framework addressing the EV-based delivery route planning problem (EDRP), which jointly optimizes route planning and charging for deliveries within time windows. After proving the problem's NP-hardness, we propose a mixed integer linear programming (MILP)-based exact solution and a computationally efficient heuristic method. Using real-world datasets, we evaluate our methods by comparing the heuristic to the MILP solution, and benchmarking it against baseline strategies, Earliest Deadline First (EDF) and Nearest Delivery First (NDF). The results show up to 39% and 22% reductions in the charging cost over EDF and NDF, respectively, while completing comparable deliveries.

cs.AI↗

arXiv2025

A Parallel Algorithm for Finding Robust Spanners in Large Social Networks

Social networks, characterized by community structures, often rely on nodes called structural hole spanners to facilitate inter-community information dissemination. However, the dynamic nature of these networks, where spanner nodes may be removed, necessitates resilient methods to maintain inter-community communication. To this end, we introduce robust spanners (RS) as nodes uniquely equipped to bridge communities despite disruptions, such as node or edge removals. We propose a novel scoring technique to identify RS nodes and present a parallel algorithm with a CUDA implementation for efficient RS detection in large networks. Empirical analysis of real-world social networks reveals that high-scoring nodes exhibit a spanning capacity comparable to those identified by benchmark spanner detection algorithms while offering superior robustness. Our implementation on Nvidia GPUs achieves an average speedup of 244X over traditional spanner detection techniques, demonstrating its efficacy to identify RS in large social networks.

cs.SI↗

arXiv2024

SMEVCA: Stable Matching-based EV Charging Assignment in Subscription-Based Models

The rapid shift from internal combustion engine vehicles to battery-powered electric vehicles (EVs) presents considerable challenges, such as limited charging points (CPs), unpredictable wait times, and difficulty selecting appropriate CPs. To address these challenges, we propose a novel end-to-end framework called Stable Matching EV Charging Assignment (SMEVCA) that efficiently assigns charge-seeking EVs to CPs with assistance from roadside units (RSUs). The proposed framework operates within a subscription-based model, ensuring that the subscribed EVs complete their charging within a predefined time limit enforced by a service level agreement (SLA). The framework SMEVCA employs a stable, fast, and efficient EV-CP assignment formulated as a one-to-many matching game with preferences. The matching process identifies the preferred coalition (a subset of EVs assigned to the CPs) using two strategies: (1) Preferred Coalition Greedy (PCG) that offers an efficient, locally optimal heuristic solution and (2) Preferred Coalition Dynamic (PCD) that is more computation-intensive but delivers a globally optimal coalition. Extensive simulations reveal that PCG and PCD achieve a gain of 14.6% and 20.8% over random elimination for in-network charge transferred with only 3% and 0.1% EVs unserved within the RSUs vicinity.

cs.GT↗