Blockchain Papers

Sami Rashid Mohammed Shibah

No abstract is available for this record.

Carlos Nú Nez-Gómez, Víctor Garcia-Font, Helena Rifà-Pous, Muhammad Asad

The submitted work contains the following highlights: • We detail NxGenT, a decentralized reputation system for B5G and 6G networks. • We propose a three-phase reputation mechanism based on smart contracts. • We implement NxGenT and release its source code as open-source software. • We analyze NxGenT’s design and resilience against relevant trust attacks. • We evaluate NxGenT’s functionality and performance through experimental analysis. The evolution towards Beyond 5G (B5G) and 6G networks presents highly heterogeneous and dynamic scenarios in which numerous entities, including network operators, service providers and end users, interact in environments of mutual trust. However, the open nature of these networks poses significant challenges regarding security and trust, as traditional centralized mechanisms may prove inadequate or insufficient in such scenarios. In this context, decentralized trust systems are positioned as a promising solution to assess the reliability of entities participating in B5G and 6G networks, thereby enhancing decision-making processes and resilience of these environments. This paper introduces NxGenT , a decentralized reputation system based on blockchain and smart contracts for B5G/6G networks that guarantees the immutability and transparency of the collected evidence on entities’ performance and behavior, while decentralizing and automating the reputation mechanism. NxGenT is a decentralized, trustless system in which entities establish and verify compliance with Service Level Agreements (SLA) and provide feedback or subjective opinions about the entities they interact with in order to compute and assign reputation scores. To evaluate the proposal, we implement a local B5G testbed that deploys the primary components of this type of network, along with a second cloud-based testbed to analyze scalability in networks of different sizes. Finally, we contextualize NxGenT within the 6GENABLERS project as a representative use case of the proposed trust system, thus demonstrating its applicability in real scenarios.

Chonghe Zhao, Deen Ma, Xicheng Zhang, Taotao Wang · 5 authors

Ethereum is a leading blockchain system whose security and performance depend critically on efficient transaction propagation. Currently, to curb bandwidth, Ethereum employs a hybrid protocol wherein nodes eager-push full transactions to a random peer subset and announce hashes to the remainder for on-demand lazy-pull. Yet, our MainNet measurements reveal that this strategy suffers from “blind” forwarding—indiscriminately transmitting full transactions without evaluating receiver acceptance. This causes peers to receive multiple redundant copies of non-viable transactions that are ultimately discarded by local pools. Also, randomized peer selection results in substantial redundant transmission, excessive hops, and a high lazy-pull rate (the proportion of transactions acquired via the slower request-response mechanism). These two inefficiencies collectively prolong propagation latency and limit further bandwidth savings. To address these limitations, we propose BELLTP, aBandwidth-Efficient andLow-LatencyTransactionPropagation protocol that incorporates two key awareness modules.Transaction Awarenessclassifies transactions into “urgent” and “leisure” using a lightweight predictor trained on Mainnet data to forecast receiver acceptance. Based on this, leisure transactions (with low predicted acceptance) are propagated via full lazy-pull to ensure that each node receives the full payload once, while urgent transactions utilize strategically targeted eager-push guided by node awareness to rectify random selection inefficiencies.Node Awarenessidentifies hub and common nodes for urgent transaction propagation by combining a theoretical network model, enabling hub nodes to eagerly push to all peers (shortening hop count while maintaining low redundant transmission) and common nodes to leverage a peer need score to eagerly push only to high-need peers (lowering lazy-pull rate). Extensive experiments on Ethereum MainNet and in a controlled local testbed demonstrate that BELLTP significantly outperforms the current protocol in both bandwidth consumption and propagation latency.

Richa Golash, Shahnawaz Ahmad, Bhawana, Naween Kumar

No abstract is available for this record.

Shuhan Qi, Qinglin Zhao, Zijie Liu, Mengchu Zhou · 7 authors

Ethereum 2.0 (ETH2) marks a pivotal shift in blockchain technology, transitioning from a Proof-of-Work (PoW) to a Proof-of-Stake (PoS) consensus mechanism, with Gasper at its core. While this evolution promises enhanced scalability and energy efficiency, the performance of its block proposal stage is highly sensitive to network latency and system parameters, such as slot length. This sensitivity introduces a critical trade-off between throughput and security, measured by the probability of blockchain forking. This paper reveals that network latency is not just a passive risk but an exploitable attack surface. We introduce the "adaptive latency-driven equivocation attack", a novel adversarial strategy where an attacker deliberately creates forks while mimicking the behavior of a high-latency node, thus achieving plausible deniability. To formally analyze and quantify the impact of this threat, we develop a comprehensive theoretical model by using Markov chains to analyze the fork probability and throughput of the Gasper's block proposal mechanism under both honest and adversarial conditions. Through extensive simulations, we validate the accuracy of our model in both normal and bursty traffic conditions. Our findings provide a systematic methodology for optimizing system parameters to achieve a robust balance between performance and security, offering a foundational guide for configuring ETH2 networks against sophisticated, latency-based threats.

Adya Pratapchandran

No abstract is available for this record.

Steven Paul Nohr

<b><i>Zombie Validator Resurrection</i></b> is a core consensus-layer exploit in Proof-of-Stake (PoS) and alternative Layer-1 networks where inactive, slashed, or economically abandoned validators regain influence without restoring proportional economic security. Through protocol gaps, state resets, or weak liveness enforcement, validators that should be neutralized re?-enter consensus, undermining safety assumptions and enabling stealth attacks. This paper formalizes the structural conditions enabling zombie validators, analyzes common resurrection mechanisms, and examines systemic risks to consensus integrity. We propose mitigation strategies to enforce validator lifecycle accountability and safeguard decentralized networks against stealth reactivation attacks.

Ashray Dubey, Thatikonda Supraja, Priyanka Chawla

No abstract is available for this record.

Baptiste Beltzer, Emmanuel Conchon, Sylvain Giroux

No abstract is available for this record.

Hugo Hedlund

A majority of well-known blockchains supports Non-Fungible Tokens (NFTs) operations via smartcontracts. Smart contracts often allow users to store arbitrary logic and code for a small programon the blockchain, and run it on the validator nodes when a monetary transaction is made to thecontract’s account. The use of smart contracts creates extra complexity in the blockchain protocolwhen allowing concurrent smart contract execution between shards alongside monetary transactions.To reduce the complexity of the blockchain protocol, native tokens and operations could be supportedby the blockchain, albeit restricting developers to a predefined set of functions and operations. Nativetoken operations can serve as a secure complement to the versatility of Turing-complete smartcontracts, providing robust, predefined primitives that could reduce the surface area for smart contractvulnerabilities. This project introduces a native token model that is centered on unique, non-fungible tokens thatsupport three primary use cases: ownership, rights and authenticity. By requiring bilateral transferconsent, and supporting optional clawback and freeze mechanisms, for revocable rights and non-transferable tokens respectively, the model bridges the gap between blockchains and legal frameworks.Additionally, an indirect transaction cost mechanism is implemented to protect the blockchain networkagainst Denial-of-Service attack during non-monetary operations. The project is supported by an emulator implementing monetary and token operations, alongsidea implementation design of how the token protocol could be integrated within ScaleGraph. The emulator’s throughput displays a slightly degraded performance of token transfer operations of ~150TX/s on average, compared to monetary transfers in the current implementation. Despite this, theemulator maintains a high average throughput of ~1300 TX/s with load generators of both monetaryand token transfer operations, when tested on a VPS with 2 vCPUs and 4GB of ram.If the token operation properties and performance compared to monetary transaction are transferableto ScaleGraph, native tokens could be a preferred way of implementation of tokens independentlywether or not with implementation of smart contracts.

Tiago Ferreira Cavazin

O presente artigo investiga a otimização do throughput em redes blockchain baseadas em mecanismos de consenso Proof-of-Stake (PoS) a partir de um estudo comparativo de latência entre diferentes arquiteturas, parâmetros de configuração e arranjos de infraestrutura de rede. Partindo do desafio clássico da engenharia de sistemas distribuídos de conciliar escalabilidade, segurança e descentralização – conhecido popularmente como trilema do blockchain –, analisa-se como decisões de projeto, dentre as quais se incluem o desenho do protocolo de consenso, a duração dos intervalos (slots), o tamanho dos blocos, o número de validadores ativos e a topologia da rede P2P, afetam simultaneamente o volume de transações processadas por unidade de tempo e o tempo de confirmação percebido pelos usuários finais. A metodologia adotada combina revisão sistemática da literatura sobre modelos de desempenho em sistemas distribuídos e blockchains PoS, comparação de estudos empíricos que mensuram throughput e latência em plataformas públicas de referência – particularmente Ethereum pós-Merge, Solana e Polkadot – e análise conceitual dos trade-offs identificados. Os resultados obtidos sugerem que técnicas de otimização orientadas a parâmetros, tais como ajuste fino de tempos de slot e políticas de propagação acelerada de blocos, combinadas a inovações em camada de consenso – dentre as quais se destacam mecanismos híbridos com Proof of History e execução paralela de transações –, podem reduzir significativamente a latência de confirmação sem necessariamente comprometer a segurança criptográfica ou a robustez da rede. Entretanto, emergem limites estruturais associados à variabilidade de condições de rede em escala global, à heterogeneidade de capacidade computacional entre nós validadores e às exigências clássicas de tolerância a falhas bizantinas, os quais impõem fronteiras práticas ao ganho de throughput em ambientes permissionless. Conclui-se que uma abordagem sistêmica, que integre ajuste criterioso de parâmetros, desenho criterioso da infraestrutura de nós e mecanismos de monitoramento e benchmarking contínuos, é fundamental para o amadurecimento da infraestrutura Web3 e para o suporte a aplicações com requisitos estritos de qualidade de serviço, entre as quais se incluem protocolos de finanças descentralizadas, mercados de capitais tokenizados e aplicações sensíveis ao tempo.

Steven Paul Nohr

Proof-of-Stake (PoS) networks rely on economic staking guarantees to align validator behavior with network security. However, a growing class of architectures permits validators to reuse, rehypothecate, or synthetically represent the same stake across multiple validation contexts. This paper defines <i>Double-Staking Validator Drift</i> as a structural vulnerability in which a validator’s effective economic exposure diverges from its apparent security contribution. We demonstrate how this drift undermines slashing guarantees, enables risk amplification, and creates false decentralization signals across PoS Layer-1 and restaking ecosystems. Existing mitigations—including slashing, social governance, and restaking policies—are shown to be insufficient due to enforcement blind spots and cross-domain opacity. We propose a logic-layer enforcement model that introduces validator-level stake exclusivity, temporal binding, and deterministic enforcement across consensus domains. This approach restores economic integrity to PoS security models and closes a critical gap in current network designs.

Steven Paul Nohr

Proof-of-Stake (PoS) networks rely on slashing mechanisms to deter validator misbehavior and preserve consensus security. While early designs emphasized cryptographically verifiable conditions, many contemporary PoS systems increasingly incorporate social, governance, or reputation-based enforcement mechanisms to supplement protocol-level slashing. This paper identifies and formalizes a novel exploit class—<b><i>Social Slashing Exploits</i></b>—where subjective reputation signals, off-chain coordination, or governance influence are weaponized to selectively penalize honest validators or shield malicious actors. We analyze how reputation-driven enforcement undermines determinism, enables cartel behavior, and erodes consensus neutrality without requiring protocol violations. The paper argues that reputation-weighted slashing cannot provide reliable security guarantees in adversarial economic environments and proposes a logic-layer enforcement model based on execution-bound, objective misbehavior proofs. This approach restores deterministic accountability, preserves validator neutrality, and improves long-term system survivability.

Микола Маленко

This article examines the limitations of existing hybrid rollup solutions and presents an adaptive L2 architecture model that leverages artificial intelligence mechanisms. It is shown that current approaches to combining optimistic and ZK verification are largely based on static rules or manual mode selection, which prevents them from effectively accounting for load dynamics, risk profiles, and domain-specific properties of applications. Based on an analysis of optimistic, ZK, and hybrid rollups, an adaptive hybrid rollup model with AI-based transaction routing is proposed. This model combines transaction classification, GNN-based decision making, LSTM-based network condition forecasting, a dual-path execution system, and a continuous learning module. The article describes a Predictive Routing Algorithm that performs proactive selection between ZK and optimistic paths, taking into account cost, latency, security, and risk profile, as well as a Dynamic Resources Allocation mechanism that dynamically redistributes resources between the paths. The proposed multi-criteria optimization framework demonstrates the ability to tune objective weights to the specifics of different classes of DeFi and Web3 protocols. It is shown that the implementation of such a model is promising for systems with high transactional intensity, as it enables a shift from manual configurations to automated, data-driven policies for resource and risk management in hybrid rollup architectures.

Xi Lin, Junfeng Tian

Blockchain consensus mechanisms are fundamental to the security and decentralization of distributed ledgers. In Proof-of-Stake (PoS) systems, which are lauded for their energy efficiency, the fair and unpredictable selection of block proposers is paramount and relies heavily on secure random number generation. The RANDAO random number generation mechanism in the Gasper protocol is susceptible to hash collision attack, which can introduce adversarial bias in the block proposer selection process. From the perspective of resisting adversarial bias attacks, this paper examines the optimization of the Gasper consensus protocol, focusing on security issues such as vulnerabilities to hash collisions in RANDAO and high latency in asynchronous network environments. By analyzing the spatial–temporal distribution of historical block hashes, we propose a dual-round random number verification mechanism that enhances reliability through multiple validation models. We develop a dynamic game-theoretic model under incomplete information to analyze node strategy selection and interaction dynamics. Our experimental results demonstrate that the improved protocol (RABA-Gasper) offers superior resistance to attacks, fairness, and efficiency compared to conventional protocols. RABA-Gasper outperforms conventional ones, achieving a 6.8% attack success rate (vs. 32.7% for RANDAO and 18.2% for Two Look-Back) with 94.3% hash collision detection, a proposer Gini coefficient below 0.23, 2.3x higher throughput retention than RANDAO in asynchronous networks, and a slightly increased random number generation latency of 125 ms. Supported by a game-theoretic model, it guarantees security when honest nodes account for ≥2/3 of the total.

Abdullah Aljumah, Tariq Ahamed Ahanger, Imdad Ullah

Unmanned Aerial Vehicles (UAVs) are increasingly deployed across diverse domains such as surveillance, logistics, and disaster management. However, ensuring the safety, security, and trustworthiness of UAV operations remains a significant challenge, primarily due to vulnerabilities in centralized data processing architectures. Traditional UAV systems rely on remote cloud servers to perform machine learning (ML)-based analytics, which introduces issues such as data exposure, latency, scalability bottlenecks, and susceptibility to cyberattacks during data transmission and storage. These challenges underscore the urgent need for a decentralized, verifiable, and privacy-preser ving learning mechanism that can support collaborative UAV intelligence without centralized control. To address these limitations, this study proposes a blockchain-enabled distributed ML framework that facilitates secure, peer-to-peer collaboration among UAV nodes. The framework integrates blockchain’s immutable ledger and smart contracts with decentralized ML models, enabling UAVs to share and validate trained models rather than raw data. This ensures data confidentiality, integrity, and transparency throughout the learning process. A stacking-based ensemble mechanism is employed to enhance predictive performance through collaborative knowledge aggregation. The proposed system is experimentally validated using a collaborative intrusion detection (ID) scenario using the KDD99 network attack data set and real-world implementation. The results demonstrate significant improvements in detection accuracy, latency and F1-score compared to conventional centralized ML methods, achieving an average accuracy of 97.9%, latency 198ms, and F1-score exceeding 97%. These outcomes confirm that the integration of blockchain and decentralized ML effectively mitigates cybersecurity risks while enabling scalable, trustworthy UAV intelligence.

Nektarios Deligiannakis, Vassilis Papataxiarhis, Michalis Loukeris, Stathes Hadjiefthymiades · 18 authors

Recently, the need for unified orchestration frameworks that can manage extremely heterogeneous, distributed, and resource-constrained environments has emerged due to the rapid development of cloud, edge, and IoT computing. Kubernetes and other traditional cloud-native orchestration systems are not built to facilitate autonomous, decentralized decision-making across the computing continuum or to seamlessly integrate non-container-native devices. This paper presents the Distributed Adaptive Cloud Continuum Architecture (DACCA), a Kubernetes-native architecture that extends orchestration beyond the data center to encompass edge and Internet of Things infrastructures. Decentralized self-awareness and swarm formation are supported for adaptive and resilient operation, a resource and application abstraction layer is established for uniform resource representation, and a Distributed and Adaptive Resource Optimization (DARO) framework based on multi-agent reinforcement learning is integrated for intelligent scheduling in the proposed architecture. Verifiable identity, access control, and tamper-proof data exchange across heterogeneous domains are further ensured by a zero-trust security framework based on distributed ledger technology. When combined, these elements enable increasingly autonomous workload orchestration, trading centralized control for adaptive, decentralized operation with enhanced interoperability, scalability, and trust. Thus, the proposed architecture enables self-managing and context-aware orchestration systems that support next-generation AI-driven distributed applications across the entire computing continuum.

Smita Bhore, N. A. Natraj, Giri Hallur

Blockchain networks have revolutionized decentralized applications but remain vulnerable to evolving security threats due to their reliance on static consensus mechanisms that cannot adapt to changing threat landscapes. This paper addresses this critical security gap by proposing Autonomous Defense-Adaptive Consensus Optimisation for Blockchain Networks (ADACON), an original framework for the dynamic adjustment of consensus mechanisms based on Bayesian threat detection. The research investigates how real-time adaptation between multiple consensus protocols can enhance blockchain resilience while maintaining performance. The approach integrates a Bayesian Threat Detector, Consensus Adapter, and Network State monitor in a modular architecture that continuously assesses network conditions and switches between five consensus mechanisms (PoW, PoS, PBFT, PoA, DPoS) as threats emerge. The framework was evaluated through comprehensive simulations involving 1,000 nodes, testing response to six distinct attack vectors, including Sybil, DoS, Byzantine, Eclipse, Majority, and Routing attacks. Results demonstrate that ADACON effectively identifies and responds to varied attacks with a latency of 29.7 ms and throughput of 833 TPS). Statistical validation across five independent simulation runs (seeds 5-9) confirmed framework reliability with consistent performance metrics (CV < 7.1% for latency, 5.4% for throughput). Delegated Proof of Stake emerged as the most frequently selected mechanism (23.2%) due to its balanced performance across multiple security dimensions. Significantly, the system exhibited greater adaptability and attack coverage than existing hybrid approaches. The previous high switching frequency was reduced by using hysteresis, i.e., by providing dwell time and an improved threshold that avoids unnecessary switching. The study concludes that dynamic consensus adaptation offers substantial security advantages for blockchain networks, particularly in high-security environments like financial systems and critical infrastructure. However, further research must focus on optimizing switching frequency and developing secure transition protocols to maximize effectiveness. ADACON represents an incremental extension tested toward more resilient blockchain systems that can autonomously respond to emerging threats while balancing security, performance, and resource utilization.

Vikash Singh, Little, Barrett, Phil Hayes, Fang, Max · 7 authors

Verifying the private liquidity state of Lightning Network (LN) channels is desirable for auditors, service providers, and network participants who need assurance of financial capacity. Current methods often lack robustness against a malicious or compromised node operator. This paper introduces a methodology for the verification of LN channel balances. The core contribution is a framework that combines Trusted Execution Environments (TEEs) with Zero-Knowledge Transport Layer Security (zkTLS) to provide strong, hardware-backed guarantees. In our proposed method, the node's balance-reporting software runs within a TEE, which generates a remote attestation quote proving the software's integrity. This attestation is then served via an Application Programming Interface (API), and zkTLS is used to prove the authenticity of its delivery. We also analyze an alternative variant where the TEE signs the report directly without zkTLS, discussing the trade-offs between transport-layer verification and direct enclave signing. We further refine this by distinguishing between "Hot Proofs" (verifiable claims via TEEs) and "Cold Proofs" (on-chain settlement), and discuss critical security considerations including hardware vulnerabilities, privacy leakage to third-party APIs, and the performance overhead of enclaved operations.

Yanjie Cai, Junchao Zhang, Yaohui Zhong, Jiahui Huang · 10 authors

Proof-of-Stake (PoS) has become a widely adopted low-energy consensus paradigm, yet its incentive mechanism remains vulnerable to strategic deviations under complex temporal and network conditions. Existing analyzes rely heavily on theoretical reasoning or manually crafted scenarios, leading to limited coverage and substantial expert overhead. This paper presents PoSitive, an automated framework for systematically uncovering incentive weaknesses in PoS consensus. PoSitive establishes a closed-loop workflow composed of four cooperative modules: a scenario construction module that generates diverse and controllable adversarial configurations, a scenario execution module that faithfully reproduces validator interactions, an outcome evaluation module that quantifies incentive deviations and consensus instability, and a policy optimization module that employs reinforcement learning to iteratively refine attack strategies and explore a broader strategic space. Experimental results demonstrate that PoSitive can effectively identify incentive-layer security risks. Using this framework, we uncover three previously unknown attack strategies, and comparative experiments further confirm the significant role of the policy optimization module in enhancing both attack quality and success rate.

Lalan Kumar, S H Manjula

Blockchain systems, such as Bitcoin and Ethereum 2.0, face vulnerabilities under bandwidth-constrained partitions, where throughput collapses and latency increases. In addition, adversaries can exploit inconsistencies to launch double-spending attacks. This study presents a lightweight dual-layer countermeasure that integrates a robust freezing threshold ( ) with multi-signal disconnection proofs to enhance performance and security without altering consensus rules. Controlled simulation experiments on Bitcoin (PoW) and Ethereum 2.0 (PoS) show throughput gains exceeding 1000% in Ethereum and over 100% in Bitcoin, with inconsistency reduced by up to 64% and latency bounded within 5-6 blocks/s. These results confirm that attacker-aware thresholds and multi-signal validation substantially improve blockchain resilience under partitioned network conditions.

Alexandros Antonov, Evangelos Kolyvas, Spyros Voulgaris

Reaching consensus in Proof-of-Stake (PoS) based consensus protocols, requires supermajority agreement among participating validator nodes. Such protocols need significant network resources due to the concurrent voting of a large number of consensus nodes. As a solution, these nodes are divided into committees, with each committee voting individually at a dedicated time slot. In this paper, we introduce CliqueSensus, a protocol that, given a distribution of consensus nodes into committees, lets them self-organize into small, ephemeral clusters structured in clique topologies, to accelerate the voting process, while using only a small fraction of the network resources required by conventional message dissemination methods. Our evaluation demonstrates that our protocol exhibits rapid convergence and operates with minimal network overhead. We focus on the PoS consensus algorithm adopted by Ethereum 2.0. In addition to our protocol, we also analyze and simulate the clustering approach that Ethereum has adopted, showcasing that our protocol can reduce validation message dissemination time by 23% to 70%, while requiring about 190 times fewer validation message forwards.

Rima Grati, Khouloud Boukadi, Safa Elleuch

The Proof of Stake (PoS) consensus mechanism is increasingly used in blockchain systems; however, resource allocation for PoS-based mobile blockchain networks remains underexplored, particularly given the constraints of mobile devices. This work introduces MEC-Chain, a new framework that integrates Mobile Edge Computing (MEC) with mobile blockchain to support efficient validator-node execution under PoS. MEC-Chain formalizes a multi-objective resource-allocation problem that jointly considers latency, reliability, and cost from both the validator and MEC-provider perspectives. To address this challenge, we develop a deep reinforcement learning-based allocation agent using the Proximal Policy Optimization (PPO) algorithm. Experimental results show that PPO achieves a 30–40% reduction in total execution time, 25–35% lower transmission latency, and 10–15% higher reliability compared to A2C (Advantage Actor–Critic) and DQN (Deep Q-Network), while offering comparable cost savings across all methods. These results demonstrate the effectiveness of MEC-Chain in enabling low-latency, reliable, and resource-efficient PoS validation within mobile blockchain environments.

Rafał Skowroński

Blockchain networks face a critical but understudied threat: wave attacks that exploit difficulty adjustment algorithms through strategic mining participation. Adversaries cyclically withdraw and re-enter mining to create oscillations that degrade network liveness and destabilize honest miners’ revenue. We present the first production-ready framework that maintains network responsiveness while enabling robust, post hoc threat detection. The framework employs a statistically rigorous pipeline featuring controller-aligned anomaly detection, transitive collusion grouping via union-find, and Benjamini–Hochberg False Discovery Rate control. We formally prove the economic viability of this architecture: when penalties on unvested rewards are enabled by governance, wave attacks become asymptotically unprofitable for rational adversaries. Evaluated on a 128-node distributed testbed simulating Bitcoin, Ethereum Classic, and Monacoin networks over 30 independent runs, our framework achieves 92.7% F1-score in detecting attacks, significantly outperforming baseline methods (74.7%). This work provides a complete, theoretically-grounded solution for securing proof-of-work blockchains against difficulty manipulation, forming the foundation for the adaptive AI-driven enhancements presented in our companion paper (Part II).