This paper examines the use of Event-Driven Architecture (EDA) patterns to improve the optimization of financial risk evaluation in a distributed cloud-based system of finance.Today's financial system is characterized by a number of difficulties in processing high-speed data feeds in a timely manner, ensuring sub-millisecond latency and high availability.This paper proposes a decoupled system utilizing distributed event brokers and stream processors to identify market anomalies and credit risks in a timely fashion.This research utilizes a risk data set of 404 unique risk scenarios, including high-frequency trading (HFT) simulation data and credit transaction data, to measure system efficiency.The system environment utilizes Apache Kafka for event streaming, Kubernetes for cloud orchestration, and Prometheus for monitoring.The results show that event-driven architecture can improve system efficiency by eliminating traditional requestresponse processing bottlenecks.Furthermore, by utilizing distributed ledgers and serverless architecture, financial organizations can improve their risk profile granularity.The results show that by utilizing reactive programming, financial organizations can improve their risk management approach by shifting their traditional reactive approach to a proactive approach.
Across a wide range of modern process-aware systems, behavioral knowledge is captured and used in multiple shapes. The same system may be described at design time through process modeling, observed at runtime through execution data, and acted upon through decision-making components that adapt to changing operational conditions. As a result, process representations must remain consistent across heterogeneous tasks, supporting verification, diagnostics, and optimization, without the need to redefine them from scratch every time the viewpoint changes. Meeting these requirements calls for representations that are expressive enough to capture complex behavioral relations, while remaining susceptible to analytical reasoning, quantitative assessment, and integration with data-driven techniques. These challenges arise in a variety of domains, including business processes, service- oriented architectures, and autonomous and learning-based systems. Among these landscapes, Business Process Management (BPM) provides a particularly illustrative domain in which the value of behavioral process knowledge becomes both operational and measurable. Business information systems generate large amounts of execution data that records how processes unfold in practice. This has fostered the development of process mining techniques, which aim to discover process representations from observed execution traces in the form of event logs and to assess whether observed behavior matches expected behavior. Consequently, process representations should not be regarded as mere design-time artifacts. Instead, they enable organizations to obtain a complete view of their operations, identify inefficiencies and potential deviations, and systematically improve the way work is performed. Indeed, behavioral insights directly translate into measurable performance gains, affecting key performance indicators (KPIs), namely throughput time, resource utilization, regulation compliance, and operational costs. Business process modeling (hereafter process modeling) has traditionally relied on imperative (or procedural) representations (e.g., Workflow net or BPMN) that explicitly define the allowed execution paths of a system. While these models provide precise operational semantics and support automated execution, their explicit nature may restrict analysis and diagnosis in settings characterized by variability, flexibility, or partial observability. Conversely, declarative specifications have received increasing interest, as they suggest a significantly different way of representing processes, describing behavior in terms of constraints that executions must satisfy, offering a more flexible representation that is particularly suited to highly dynamic environments. These constraints are typically expressed as temporal rules grounded in logics such as Linear Temporal Logic on Finite Traces (LTLf), stating what must hold in a process rather than prescribing how it is achieved. As a result, they can serve as an internal behavioral layer that can be reused across multiple analysis and reasoning tasks. Despite their complementary strengths, imperative and declarative paradigms are often treated as separate approaches, adjusting their adoption across different analytical and reasoning tasks. Similar process representation challenges arise in service-based and cyber-physical settings, as well as in autonomous and learning-based systems, where agents must continuously adapt their behavior while still respecting safety and compliance constraints. In these scenarios, behavioral knowledge configures as a basis for monitoring and assessing whether observed or learned policies remain consistent with expected temporal and causal relations. This thesis investigates the use of declarative temporal specifications as a unifying abstraction span the full lifecycle of behavioral reasoning in process-aware systems. The main research objective guiding the work is stated as follows: RO: Declarative Specifications as a Unifying Behavioral Abstraction Investigate the use of declarative temporal specifications as a unifying abstraction for representing, verifying, andexploiting behavioral knowledge in process-aware systems. With this goal in mind, we set three research questions across process representations, verification, and decision making. RQ1: Process Representation Paradigms How can a declarative process specification be systematically derived from an imperative process model, preserving behavioral equivalence between the two representations? First, we discuss the relationship between imperative and declarative process representations in the business processes domain, showing how a declarative specification can be systematically derived from an imperative model while preserving behavioral equivalence. To this end, we introduce a systematic method to synthesize declarative specifications from safe and sound Workflow nets, together with formal guarantees ensuring that the two representations capture exactly the same behavior. This result enables principled transitions between paradigms, allowing analysts to select the most suitable representation depending on the task at hand. RQ2: Conformance Measurement How can the satisfaction of declarative process specifications be quantitatively assessed over observed executions? Once a process has been modeled, the growing availability of execution data makes it natural to quantify how closely observed traces adhere to the normative representation. This is the goal of conformance checking, which measures the degree of correspondence between a model (or specification) and recorded executions. Traditional conformance checking verification techniques for declarative specifications often assess individual rules in isolation, overlooking their mutual interplay and offering primarily binary satisfaction outcomes. We address this gap by proposing a probabilistic framework to quantify the satisfaction of declarative process specifications over event logs. We introduce interestingness measures that characterize the degree to which LTLf declarative constraints are supported by observed behavior. These measures enable behavioral diagnostics and provide a foundation for detecting deviations, changes, and emerging patterns in process executions. RQ3: Decentralized Data Sources How can declarative specifications be checked against execution data that is distributed across multiple independent sources? Capturing specification-level insights becomes even more relevant when execution data are distributed across independent sources, each party observing only a partial view of the process, while a global assessment must be derived from distributed evidence. In these inter-organizational settings, traditional conformance checking approaches are hindered by confidentiality and governance constraints that prevent centralizing raw event logs. To overcome this limitation, we propose CONFINE, a secrecy-preserving framework that enables process mining and declarative conformance checking without disclosing sensitive execution data between collaborating organizations. CONFINE executes mining and checking algorithms inside trusted execution environments (TEEs), where event logs are transmitted and processed under attested confidentiality guarantees. The architecture supports a symmetric collaboration model in which each party can provide execution data and perform joint analyses, enabling behavioral insight to cross organizational boundaries while preserving data sovereignty.
Blockchain technologies are making it possible to develop crypto-currencies and programmable smart contracts that can work in worldwide trustless and decentralized environments. Decentralized autonomous organizations (DAOs) that can coordinate the works of crowds of users, developers, and researchers can be built using smart contracts on blockchains. We contribute a decentralized autonomous software organization model and an Ethereum blockchain-based smart contract named AutonomousSoftwareOrg that provides a continuously operating virtual organization for open-source software development communities and users. AutonomousSoftwareOrg provides a project funding mechanism based on crypto-currencies, a decision-making mechanism based on voting, and recordkeeping for software usage citations and executions. Furthermore, software executions, along with their input and output data files, can also be transactionally recorded in AutonomousSoftwareOrg. This enables software execution graphs to be constructed for analysis. An AND/OR graph model of input/output data and software executions is presented, along with analysis algorithms for execution traceability and reproducibility assessment. AutonomousSoftwareOrg is deployed and tested on the Ethereum-based Bloxberg blockchain network which is operated by academic and research institutions, demonstrating its practical viability for sustainable open-source software development.
Engineering Next-Gen Financial Systems: Designing Cloud-Native Middleware for Scalable Trading Architectures Objectives:1. To explain the principles of cloud-native architecture and middleware design for buildingscalable, resilient, and high-performance financial trading systems.2. To provide a comprehensive understanding of microservices, containerization, orchestration,messaging systems, and distributed data management in modern financial platforms.3. To demonstrate how advanced technologies such as Apache Kafka, Kubernetes, Redis, gRPC,and AI-driven systems can improve real-time trading operations and financial servicedelivery.4. To guide software engineers, architects, and financial technology professionals in designingsecure, fault-tolerant, and low-latency trading infrastructures with strong observability andcompliance practices.5. To prepare readers for the future of financial technology by exploring emerging trendsincluding serverless computing, WebAssembly, AI/ML integration, decentralized finance(DeFi), and next-generation cloud-native trading ecosystems. Table of Contents CHAPTER 1 The Great Migration: From Monoliths to MicroservicesCHAPTER 2 Foundations of Cloud-Native DevelopmentCHAPTER 3 Financial Middleware: The Digital Nervous SystemCHAPTER 4 High-Performance Messaging PatternsCHAPTER 5 Data Persistence and Caching StrategiesCHAPTER 6 Service Communication and API ManagementCHAPTER 7 Architectural Deep Dive: Market Data SystemsCHAPTER 8 Architectural Deep Dive: Order Management Systems (OMS)CHAPTER 9 Architectural Deep Dive: Algorithmic Trading PlatformsCHAPTER 10 Building for Failure: Resilience and ObservabilityCHAPTER 11 Security, Compliance, and Emerging TechnologiesCHAPTER 12 Case Study: A Cloud-Native Equity Trading Platform
Apr 22, 2026·2026 International Conference on Recent Advances in Electrical, Electronics, Ubiquitous Communication, and Computational Intelligence (RAEEUCCI)
Most blockchain anomaly detection research is model-centric, focusing on either proposing new models or comparing benchmarks. Very little work addresses how such methods behave when actually deployed. Data ingestion, feature updates, and inference stability are among the practical concerns that usually get ignored. This work targets this: design, implementation, and evaluation of an AutoML-based platform for anomaly detection targeting Ethereum wallets. The emphasis is on operational behavior, rather than algorithmic novelty. The system is an AutoGluon ensemble that is trained offline from historical, labeled data. Each wallet behavior is represented as a fixed-length feature vector. During inference, live blockchain data is fetched through the blockchain API. This data is transformed into features required by the trained model. The predictor then outputs probabilistic risk scores along with feature-level explanations. Feature computation and model inference are treated as separate processes. This separation allows repeated inference without online learning or continuous retraining. The experimental evaluation discusses several deployment-relevant factors, including class imbalance during training and the contribution of different feature groups. It also examines the stability of the risk score under repeated feature recomputation. Results show that imbalance-aware training improves the reliability of detection. They also indicate that anomaly detection depends on the combined effect of multiple behavioral feature categories.
Donghan Chen, Zhihui Lu, Chenchi Luo, J I N Y I Lin · 7 authors
The growth of decentralized finance (DeFi) has been accompanied by an increase in rug pull scams, in which developers misappropriate investors’ funds, rendering the associated tokens worthless. Existing detection methods struggle to capture dynamic on-chain information and provide interpretable risk assessments. This paper presents RugKeeper, a multi-agent framework leveraging large language models for rug pull detection. RugKeeper constructs comprehensive token contexts via a two-step question-driven process and performs multi-path collaborative reasoning, with a Judger Agent validating results to reduce model hallucinations. Evaluations on historical datasets demonstrate that RugKeeper outperforms state-of-the-art methods, achieving 93.55% accuracy, 95.92% F1-score and robust generalization across model backbones. In a real-world sampled dataset from the BNB Chain, 638 previously undetected rug pull tokens were identified. These results highlight the effectiveness of RugKeeper in enhancing DeFi security and supporting risk mitigation.
Digital enterprises operating across multiple regions require an architecture that ensures high availability, low latency, and seamless multi-currency support. In this paper, we propose a cloud-native distributed system design that leverages microservices, geo-replication, and fault-tolerant patterns to meet these requirements. We detail the system architecture - including a multi-region deployment, microservices for currency conversion and transaction processing, and a replicated ledger - and present our methodology for performance evaluation. Our experiments compare the proposed design to a traditional monolithic baseline, showing significant improvements: for example, currency conversion latency falls from ~220 ms to ~50 ms and throughput increases sixfold under load (p<0.01). We also demonstrate 99.99% availability via automated failover and load balancing across regions. Key contributions include a detailed description of the architecture (with figures of component interactions and data flow), an analytical model of system performance, and statistical validation of results. We conclude by discussing limitations, strengths, and directions for future work. The results validate that our design substantially enhances availability and performance for global multi-currency platforms.
Abstract Non-Fungible Tokens (NFTs) have be- come a significant component of the dig- ital asset ecosystem, yet the visualiza- tion of rarity—a primary determinant of NFT value—remains fragmented, in- consistent, and methodologically opaque across web platforms. This paper pro- poses a comprehensive framework for NFT rarity visualization in web systems, integrating on-chain metadata extrac- tion, information-theoretic rarity scor- ing, and interactive, multi-layered vi- sualization components. The frame- work addresses three core challenges: (1) the absence of a unified, transpar- ent rarity computation standard; (2) the lack of effective visual encodings tai- lored to trait-level and collection-level rarity distributions; and (3) the scala- bility constraints inherent in real-time web-based rendering of large NFT col- lections. We describe the architecture in detail—from blockchain data ingestion pipelines through normalization, scoring, and front-end rendering—and present a prototype implementation using modern web technologies. A comparative eval- uation against existing commercial rar- ity tools demonstrates improvements in computational transparency, visual inter- pretability, and user task performance. The contributions of this work are three- fold: a formalized rarity scoring model grounded in information content theory, a modular visualization architecture for web deployment, and empirical evidence of the framework’s effectiveness for collec- tors, analysts, and marketplace designers.
Decentralized applications (DApps) represent a paradigm shift in software architecture, leveraging blockchain technology and distributed consensus mechanisms to eliminate single points of failure and centralized control. As the adoption of DApps accelerates across sectors such as finance, supply chain, healthcare, and governance, ensuring their functional correctness and behavioral reliability has become a critical engineering challenge. Unlike traditional software, DApps operate in adversarial, permissionless environments where smart contracts execute autonomously and immutably on distributed nodes, making post-deployment correction extremely costly or impossible. This review systematically examines the landscape of functional testing methodologies tailored for decentralized applications, analyzing their suitability, limitations, and practical applicability in modern DApp development workflows. We survey research spanning smart contract verification, consensus protocol testing, oracle interaction validation, cross-chain interoperability testing, and user-layer functional testing of Web3 interfaces. The review identifies four dominant testing paradigms: (1) unit testing of smart contract functions, (2) integration testing of DApp components, (3) property-based testing using formal specifications, and (4) end-to-end simulation on testnets. Through comparative analysis across 13 seminal studies, we evaluate each approach along dimensions of automation feasibility, coverage depth, gas efficiency awareness, and scalability to complex DApp ecosystems. Our findings indicate that while static analysis and symbolic execution tools such as Mythril, Slither, and Manticore offer strong vulnerability detection, they address security properties more than functional correctness. Conversely, framework-based testing tools like Hardhat, Truffle, and Foundry provide adequate unit-level coverage but struggle with cross-contract orchestration and event-driven logic verification. A critical gap exists in testing oracle-dependent and DAO governance workflows. This review concludes with a synthesis of best practices, open research challenges, and a directional roadmap for developing holistic functional testing frameworks suited to the evolving complexity of decentralized systems.
Jiahao Qi, Dian Ding, Jie Li, Jiannong Cao · 7 authors
Account migration in sharded blockchains presents a critical trade-off between optimization effectiveness and system availability. While dynamically reallocating accounts across shards can significantly reduce cross-shard transaction overhead, existing migration mechanisms cause service disruptions that intensify as state data volumes grow. To address this challenge, we propose BIND, a batch-wise account migration protocol that eliminates service interruptions by enabling continuous transaction processing throughout migration. BIND introduces a dual transaction pool architecture that isolates transactions involving migrating accounts while allowing non-migrating accounts to operate uninterrupted. To optimize migration efficiency, we design a reverse greedy heuristic algorithm that partitions accounts into batches based on community cohesion, maximizing intra-batch connectivity to front-load cross-shard communication reduction. We evaluate BIND using real Ethereum transactions, demonstrating superior performance over existing mechanisms. BIND achieves 12% higher overall throughput, reduces migration time to 23.6%-39.3% of the one-shot baseline (across 1-10Gbps bandwidth), and lowers cross-shard transaction rates by 24.1% compared to random batching. These results confirm BIND as a practical solution for large-scale, non-disruptive account migration in production sharded blockchains.
Insider threats pose a persistent and evolving challenge to contemporary software ecosystems, where privileged users can exploit access for malicious purposes, often evading traditional perimeter-based defences. This paper introduces a novel hybrid framework that synergistically integrates zero-knowledge proofs (ZKPs) and behavioural analytics to detect and mitigate such threats with enhanced privacy and precision. ZKPs enable secure authentication and data verification without revealing sensitive information, ensuring compliance with privacy regulations like GDPR while thwarting unauthorized access. Complementarily, our behavioural analytics engine employs advanced machine learning models, including graph neural networks and unsupervised anomaly detection (e.g., isolation forests), to profile user behaviours across software pipelines, identifying deviations indicative of insider malice. The proposed architecture is deployed in a microservices-based ecosystem, demonstrating scalability via containerized components on Kubernetes. Extensive evaluations on benchmark datasets (e.g., CERT Insider Threat) and simulated enterprise environments yield a 95% detection accuracy, with 40% fewer false positives than state-of-the-art methods like UEBA systems. Latency remains under 50ms for real-time operations, preserving performance in high-throughput scenarios. Our framework outperforms baselines by 25% in F1-score, validated through rigorous ablation studies. By bridging cryptographic privacy with AI-driven intelligence, this work advances proactive security for modern software, offering deployable solutions against sophisticated insiders. Future extensions explore quantum-resistant ZKPs for post-quantum resilience.
As blockchain technology evolves from specialized financial tools to foundational infrastructure for Web3, the necessity for rigorous performance validation becomes paramount. Stress testing—defined as the evaluation of system stability under extreme workloads—is critical for identifying bottlenecks in consensus mechanisms and peer-to-peer communication. This survey provides an exhaustive analysis of web-based stress testing frameworks. Unlike traditional CLI-based tools, web-based frameworks provide real-time telemetry and distributed orchestration capabilities essential for modern decentralized applications. We categorize existing literature into three generations of benchmarking, evaluate ten prominent frameworks based on a multi-dimensional rubric, and identify significant research gaps including the lack of standardized cross-chain stress protocols and AI-integrated anomaly detection. This work aims to provide a roadmap for researchers and DevOps engineers to select and implement robust testing environments for enterprise-grade blockchain deployments.
Cloud computing underpins modern IT infrastructure by delivering scalable, on-demand resource provisioning, yet controlling cloud expenditure remains a pressing challenge. Dynamic pricing structures, unpredictable workloads, and billing pipelines that lack real-time visibility create conditions in which unauthorized consumption and anomalous usage spikes routinely escape timely detection. This paper presents CloudPay, a blockchain-integrated cloud storage billing system that unifies unsupervised machine learning with smart contract execution to deliver verifiable, fine-grained, and fraud-resistant cost governance. The system converts user storage activity into time-series representations and applies the Isolation Forest algorithm to detect abnormal consumption spikes without any labelled training data. Flagged events are routed through an owner confirmation protocol that validates suspicious uploads before billing proceeds, preventing unauthorized charges from entering the settlement pipeline. Smart contracts autonomously compute GB-time-based charges, execute tokenized payments, and anchor every transaction to an immutable SHA-256 blockchain ledger. Experimental results confirm that the system achieves 94.4% anomaly detection accuracy, 99.7% billing precision, and an 18.4% reduction in overall cloud expenditure relative to static allocation baselines. These results demonstrate that integrating unsupervised anomaly detection with cryptographically enforced billing logic is a viable path toward tamper-evident, real-time cost governance in multi-tenant cloud environments.
Este artigo apresenta um esboço estruturado sobre “Análise de Custo de Deploy em Diferentes EVMs.”. O objetivo é analisar os fundamentos técnicos e econômicos do custo de implantação de contratos inteligentes em Ethereum L1, diversas Layer‑2 (rollups) e outras chains EVM‑compatíveis, discutindo implicações para o ecossistema Web3 e tendências de mercado. A metodologia baseia‑se em revisão bibliográfica e análise de casos práticos, com foco na decomposição do custo de deploy em componentes de gas (execução, armazenamento de código, dados de transação) e em como upgrades recentes – como Cancun/Deneb e a introdução de blobs de dados – alteram a estrutura de custos, especialmente para rollups que publicam dados em L1. Estudos mostram que, enquanto o gas é uma unidade abstrata consistente, o custo econômico por byte de código e por transação varia significativamente entre L1 (onde picos históricos chegaram a dezenas de dólares por transação) e L2s, onde taxas médias frequentemente ficam abaixo de centavos, especialmente após a redução em até 94% do custo por byte de dados com blobs. Ao mesmo tempo, análises de mercado indicam que L2 fees são estruturalmente compostas por uma parcela L1 (custo de dados e liquidação) mais uma parcela L2 (execução local), de modo que mudanças na economia de gas da L1 impactam indiretamente o custo de deploy e operação nas L2s. Conclui‑se que decisões de arquitetura e de escolha de EVM para deploy devem considerar não apenas o custo imediato de gas, mas também a herança de segurança, a volatilidade das taxas e a dependência em upgrades de protocolo que alteram a economia de dados e execução.<br>
The correct and resilient operation of distributed systems—spanning global financial ledgers, decentralized autonomous organizations, and peer-to-peer energy microgrids—depends fundamentally on the correctness of their underlying consensus mechanisms. These protocols must guarantee agreement on shared state among a collection of potentially faulty or adversarial nodes, upholding the dual properties of safety and liveness even under hostile conditions. Despite the growing diversity of consensus algorithms, from classical crash-fault tolerant approaches such as Paxos and Raft to modern Byzantine fault-tolerant (BFT) variants and Directed Acyclic Graph (DAG)-based structures, the research community lacks a unified architecture for their systematic, cross-platform evaluation. This paper proposes a modular, extensible framework called the Consensus Evaluation and Resilience Framework (CERF) to fill this gap. CERF integrates five core components: a high-fidelity network emulation layer, a pluggable system-under-test (SUT) adapter, a fault injection engine (FIE) capable of simulating both benign and Byzantine failure modes, a multi-dimensional performance monitoring unit (PMU), and a formal consistency checker. By decoupling the testing infrastructure from protocol implementation, CERF enables fair comparisons of throughput, latency, energy efficiency, and scalability. The proposed architecture draws on insights from seminal benchmarking works including BlockBench, Jepsen, and ByzzBench, and extends evaluation criteria to address modern challenges in IoT environments, post-quantum security, and DAG-based ledgers. Graph-theoretical principles are incorporated at the topology design stage to model network vulnerabilities and guide adversarial scenario construction.
The article presents an analysis of concurrent execution issues and delivered performance in large-scale distributed applications deployed in cloud-native environments. The relevance of this direction is driven by the accelerated diffusion of the microservice paradigm and container-orchestration practices, within which classical synchronization and coordination approaches often become the dominant factor behind throughput degradation and latency growth. The text identifies baseline patterns of state management and state processing and then examines–at a detailed level–the causes and enabling conditions of data races in asynchronous execution loops. A separate emphasis is placed on the specificity of Kubernetes operators and on the requirement of idempotent reconciliation cycles as a key prerequisite for predictable system behavior under repeated triggers, partial failures, and mismatches between the observed and desired state. The research goal is formulated as the development of recommendations aimed at reducing latency and increasing reliability under concurrent access to shared resources and shared entities. To achieve this goal, methods of systems analysis are applied, architectural-pattern modeling is performed, and retrospective reflection on recurring failure patterns observed in production systems. The theoretical foundation relies on works devoted to distributed ledgers, while the applied part is supported by operational guidelines and engineering practices for running NoSQL solutions. The outcome is a description of the distinctive properties of a model for handling concurrent requests, designed to improve the resilience and controllability of distributed-component behavior. The findings presented in this work are expected to be of practical interest to system architects, DevOps engineers, and researchers working in the field of distributed computing.
Blockchain technology has emerged as a secure and decentralized solution for data management across various domains. However, existing consensus mechanisms face challenges related to security, scalability, and energy efficiency, while blockchains remain vulnerable to sophisticated attacks such as double spending, selfish mining, and Sybil attacks. This paper proposes a novel hybrid blockchain security framework that integrates a Hybrid Consensus Algorithm (HCA) combining Proof of Stake (PoS) and Practical Byzantine Fault Tolerance (PBFT) with Machine Learning based attack detection. The hybrid consensus improves transaction finality and reduces energy consumption, while the ML module detects anomalous behaviors in real time. Experimental evaluation using a private Ethereum based blockchain demonstrates that the proposed approach improves attack detection accuracy up to 96.8 %, reduces consensus latency by 34 %, and enhances throughput by 27 % compared to traditional PoW based systems. The results confirm that integrating hybrid consensus with intelligent security mechanisms significantly strengthens blockchain resilience.
Blockchain technology has been widely used in the field of Internet of Things, providing effective support for solving security challenges in Internet of Things systems. However, due to the immature development language and deployment platform, smart contracts are prone to various vulnerabilities. Considering the immutability of smart contracts, efficient vulnerability detection before deployment is particularly critical. The existing detection methods have two main limitations: they can only identify a limited number of specific vulnerabilities, resulting in low coverage; the implicit correlation information between vulnerability types is ignored. In order to solve these problems, this paper proposes a smart contract multi-vulnerability detection model CorrelaScan (correlation-aware smart contract analyzer) that integrates vulnerability type correlation awareness. The model is based on a multi-task learning architecture, including a shared layer and a specific task layer. The shared layer uses BERT to extract shared features, while the specific task layer uses BiGRU to learn specific task features for vulnerability detection and type classification. In addition, a vulnerability type embedding module is integrated in the task-specific layer. The module mines potential associations by calculating the similarity between smart contract opcodes and vulnerability types, thereby enhancing detection guidance and improving model performance. Experimental verification on public datasets shows that the model can simultaneously detect 10 types of vulnerabilities such as integer overflow or underflow, reentrancy and timestamp dependence, with an average F1 value of 85.22%. Its detection performance exceeds the current state-of-the-art methods.
Ethereum Inscriptions (Ethscriptions) repurpose Ethereum calldata into a persistent inscription channel by embedding \texttt{data:}~URI payloads. These transactions typically target externally owned accounts, allowing the payload to bypass EVM execution while remaining permanently replicated across full nodes. Although calldata was originally designed for compact smart-contract parameters, this repurposing enables structured data embedding with long-term storage consequences. We present the first large-scale empirical study of Ethscriptions, treating them as a distinct \emph{calldata-resident workload} rather than merely a subset of general calldata usage. Our analysis focuses on the \textit{Ethscription} operational subset, which consists of payloads that decode to JSON and conform to a token-operation grammar (e.g., \texttt{p}, \texttt{op}, \texttt{tick}, \texttt{amt}). From $6.27$ million Ethscription candidates (\Uone), we extract $4.75$ million Ethscription operations (\Utwo, $75.8\%$ of \Uone). This result shows that structured token-like activity dominates the ecosystem. Our measurements further reveal (i) a complete workload lifecycle compressed into nine months (bootstrap, expansion, saturation), (ii) proliferation of $30$+ competing protocols without convergence toward a dominant standard, (iii) a lifecycle funnel exhibiting $201\times$ deploy-to-mint amplification and a $57.6{:}1$ mint-to-transfer collapse indicative of speculative minting, (iv) extreme participation inequality (Gini~$0.86$), and (v) a measurable permanent data footprint imposed on the Ethereum network.
The Ethereum ecosystem, which secures over $381 billion in assets, fundamentally relies on client APIs as the sole interface between users and the blockchain. However, these critical APIs suffer from widespread implementation inconsistencies, which can lead to financial discrepancies, degraded user experiences, and threats to network reliability. Despite this criticality, existing testing approaches remain manual and incomplete: they require extensive domain expertise, struggle to keep pace with Ethereum’s rapid evolution, and fail to distinguish genuine bugs from acceptable implementation variations. We present APIDiffer , the first specification-guided differential testing framework designed to automatically detect API inconsistencies across Ethereum’s diverse client ecosystem. APIDiffer transforms API specifications into comprehensive test suites through two key innovations: (1) specification-guided test input generation that creates both syntactically valid and invalid requests enriched with real-time blockchain data, and (2) specification-aware false positive filtering that leverages large language models to distinguish genuine bugs from acceptable variations. Our evaluation across all 11 major Ethereum clients reveals the pervasiveness of API bugs in production systems. APIDiffer uncovered 72 bugs, with 90.28% already confirmed or fixed by developers, including one critical error in the official specifications themselves. Beyond these raw numbers, APIDiffer achieves up to 89.67% higher code coverage than existing tools and reduces false positive rates by 37.38%. The Ethereum community’s response validates our impact: developers have integrated our test cases, expressed interest in adopting our methodology, and escalated one bug to the official Ethereum Project Management meeting. By making APIDiffer open-source, we enable continuous validation of Ethereum client API implementations, thereby strengthening the foundational integrity of the entire Ethereum ecosystem.
Byzantine Fault Tolerance (BFT) has developed from a theoretical concept in distributed database reliability to the basic structure of today's decentralized finance and global infrastructure. This paper provides a holistic overview of BFT approach, delineating imperative strides from classical synchronous resolutions to the most recent blockchain protocols. We give a formal treatment of the transition of the architecture from quadratic complexity (O(n2)) in PBFT to linear scalability$(O(n))$in HotStuff and probabilistic guarantees, as compared with Nakamoto Consensus. Moreover, we present a critical review of some recent developments in 2024 and 2025 related to Machine Learning-accelerated adaptive consensus, probabilistic relaxation for high volume trading, lightweight protocols for IoT. By combining theoretical limits with a range of practical issues such as state transfer and cryptographic overhead, we hope that the survey provides a structured roadmap for researchers tackling the scalability-security trade-offs in future distributed systems.
ABSTRACT Deterministic execution models used in distributed systems and generative pipelines share a common weakness, fixed concurrency schedules that fail under skewed workloads, hotspots, or noisy inputs. This paper proposes Adaptive Time stamp Informed Deterministic Flow Control (ATDFC), a unified concurrency control framework that combines time stamp‐based ordering with adaptive deterministic flow execution. ATDFC dynamically adjusts segmentation, initialization, and replay based on conflict intensity or input quality, enabling deterministic correctness while eliminating unnecessary computation. Experimental evaluation shows that ATDFC reduces conflict rates by more than half, lowers bandwidth consumption significantly, and maintains higher throughput and lower latency under extreme skew when compared to leading deterministic flow and concurrency control approaches. The results demonstrate that ATDFC offers a domain‐agnostic concurrency model capable of unifying deterministic computation across distributed ledgers, databases, and multimodal generative systems.
For most distributed financial systems, the constraints imposed by the CAP (Consistency, Availability, Partition Tolerance) theorem must be reconciled against the ordering constraints needed to satisfy regulatory requirements and meet the performance requirements of real-time transaction processing. This paper presents the Temporal Sequence Barrier consistency model for asynchronous high-throughput ledger systems. Combining logical vector clocks with epoch-based orchestration patterns imposes a strict causal ordering of events across multiple geographic regions without sacrificing availability. Its database-centric architecture allows stateful routing and selective replication of entities in order to achieve linearizability of causally related transactions while allowing independent sets of entities to be processed in parallel. We provide a detailed evaluation that shows that we can provide causal consistency at latency bounds equal to or better than existing systems using clever buffering and adaptive timeouts, while also addressing the classic challenges in distributed transaction management and operator complexity.
We study what model-identifying information leaks through commercial language-model APIs that expose top-k token log probabilities. Building on extreme-value theory predictions for logit order-statistic gaps, we confirm that the normalized third logit gap (δ norm) remains near the Gumbel-class constant ≈0.318 across 6 models from 3 providers (OpenAI, Google Vertex AI, xAI) and 3 independent measurement sessions, demonstrating that output-layer universality persists through API truncation and quantization. We introduce a PPP-residualization transform that removes the dominant tail scale factor and reveals a low-dimensional but stable endpoint-specific geometry in the remaining gap spectrum. Contrary to common assumption, "provider" is not a geometrically coherent label: models do not cluster by corporate origin under these observables, but they do separate by model identity across independent sessions. Using a challenge-response protocol with centroid averaging and per-model thresholds, we demonstrate cross-session endpoint verification with a 0.83% breach rate (119/120 correct identifications across three temporal sessions); per-model thresholds eliminate all breaches on this dataset. We observe a robustness phase transition governed by enrollment depth. Under single-session enrollment, prompt selection is load-bearing: the majority of bootstrapped banks fail to separate the six endpoints. Under two-session enrollment, bank sensitivity collapses on this dataset, and a bank compiler produces small compiled banks that exceed the margin of larger uncompiled banks. A dimensionless robustness parameter SNR(K,S) unifies both axes: prompt count K and enrollment depth S jointly govern the transition from bank-sensitive to bank-robust verification. We discuss operational implications for re-enrollment cadence and template management in production deployments. Addendum (02/26/2026): Post-publication results extend this framework in two directions. A distillation experiment across six training protocols demonstrates that a model's structural fingerprint (weight-geometry regime) is completely invariant to knowledge distillation, while its functional fingerprint (PPP-residual template) converges 31--52% toward the teacher's — enabling forensic detection of distillation provenance through API measurements alone. A conditional impossibility theorem, machine-checked in Coq (41 theorems, 0 Admitted), proves that no standalone model can spoof another's PPP-residual template across independent challenge prompts without exhausting its KL divergence budget, under four explicit trust assumptions. Technical Note: Gap Invariance: Why PPP Measurements Are Domain-Independent by Construction (DOI: 10.5281/zenodo.19275524) The Neural Network Identity Series — Mathematical foundations, empirical validation, and governance frameworks for verifying which model is running Newest addition: Technical Note: The Disappearing Window — AI Logprob Access Withdrawal and the Structural Verifiability of Frontier Model Contracts (DOI: 10.5281/zenodo.20362098) Paper 1: The δ-Gene: Inference-Time Physical Unclonable Functions from Architecture-Invariant Output Geometry (DOI: 10.5281/zenodo.18704275) Paper 2: Template-Based Endpoint Verification via Logprob Order-Statistic Geometry (DOI: 10.5281/zenodo.18776711) Paper 3: The Geometry of Model Theft: Distillation Forensics, Adversarial Erasure, and the Illusion of Spoofing (DOI: 10.5281/zenodo.18818608) Paper 4: Provenance Generalization and Verification Scaling for Neural Network Forensics (DOI: 10.5281/zenodo.18872071) Paper 5: Beneath the Character: The Structural Identity of Neural Networks — Mathematical Evidence for a Non-Narrative Layer of AI Identity (DOI: 10.5281/zenodo.18907292) Paper 6: Which Model Is Running?: Structural Identity as a Prerequisite for Trustworthy Zero-Knowledge Machine Learning (DOI: 10.5281/zenodo.19008116) Paper 7: The Deformation Laws of Neural Identity (DOI: 10.5281/zenodo.19055966) Paper 8: What Counts as Proof? — Admissible Evidence for Neural Network Identity Claims (DOI: 10.5281/zenodo.19058540) Paper 9: Composable Model Identity — Formal Hardening of Structural Attestations in the Enterprise Identity Stack (DOI: 10.5281/zenodo.19099911) Paper 10:Where Identity Comes From: Path Sensitivity and Endpoint Underdetermination in Neural Network Training (DOI: 10.5281/zenodo.19118807) Paper 11: Post-Hoc Disclosure Is Not Runtime Proof: Model Identity at Frontier Scale (DOI: 10.5281/zenodo.19216634) Paper 12: Family-Dependent Response to Reasoning Distillation Across Structural and Functional Identity Layers (DOI: 10.5281/zenodo.19298857) Paper 13: Safety-Alignment Removal as a Model-Identity Failure — Structural Evidence from Published Weight-Level Mutation Checkpoints (DOI: 10.5281/zenodo.19383019) Technical Note: Agent Identity Is Not Model Identity (DOI: 10.5281/zenodo.19240883) Technical Note: Measured Model Substitution Under Valid Agent Credentials (DOI: 10.5281/zenodo.19342848) Technical Note: Artifact Identity Is Not Runtime Identity — Trustfall Lite and the Boundary of File-Level Model Verification (DOI: 10.5281/zenodo.20019127) Formal Verification Stack for Neural Network Structural Identity (IT-PUF Coq Proofs) (DOI: 10.5281/zenodo.18930621) Copyright (c) 2026 Anthony Ray Coslett / Fall Risk AI, LLC. All Rights Reserved. Confidential and Proprietary. Patent Pending (Applications 63/982,893, 63/990,487, 63/996,680, 64/003,244).