A Deep Approach to an Energy-Efficient Voting-Based Consensus Algorithm for Secure Financial Transactions

A Deep Approach to an Energy-Efficient Voting-Based Consensus Algorithm for Secure Financial Transactions | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article A Deep Approach to an Energy-Efficient Voting-Based Consensus Algorithm for Secure Financial Transactions Anwar Ali Sathio, Muhammad Malook Rind, Shafique Ahmed Awan, Sameer Ali This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7012159/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Blockchain consensus mechanisms like Proof-of-Work (PoW) and Proof-of-Stake (PoS) face critical challenges in energy efficiency and latency for financial transactions. In the study we devises VoteChain, a novel voting-based consensus algorithm enhanced by deep lightweight anomaly detection. We derive a rigorous energy-latency tradeoff model and integrate a Convolutional Neural Network (CNN) for real-time fraud detection. In the study we use opensource datasets and tested that VoteChain reduces energy consumption to 20.0 mWh/tx (98% lower than PoW) while achieving 65 Tps throughput and 1.2 ms latency, outperforming PoW, PoS, and PBFT. Whereas fault tolerance reaches 99.5% and validated via Byzantine attack simulations. In the study, we explored a scalable, eco-friendly solution for decentralized financial transactions. Blockchain Consensus Algorithm Energy Efficiency VoteChain Deep Learning Fraud Detection Lightweight Financial Transactions Decentralized Finance (DeFi) Latency Optimization Throughput Fault Tolerance Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 1. Introduction Blockchain technology has heralded a new era for decentralized systems, promising unprecedented transparency, immutability, and security across various sectors, such as financial transactions. The procedure of cross-border payments to digital asset management, the blockchain technology has potential to revolutionize traditional finance, fostering trust in environments. However, the global adoption of blockchain in high-volume, real-time financial ecosystems remains impeded by the inherent limitations within its basic consensus mechanisms. There are many challenges that significantly impact the energy consumption and latency associated with prevalent consensus algorithms like Proof-of-Work (PoW) and Proof-of-Stake (PoS) algorithms. The PoW-consensus algorithm is a cornerstone of early cryptocurrencies, it needs high computational power to stagger energy footprints which are unsustainable and environmentally great concern [ 1 – 6 ]. On the side, the PoS algorithm is an energy-efficient alternative by replacing computational puzzles with staked capital, which is a powerful alternative. The PoS introduces new complexities related to centralization risks to meet the stringent latency requirements of modern financial markets. Where, the improved novel variants can dictate market opportunities and transaction finality in milliseconds [ 3 – 5 ] in the improved version of efficient consensus algos. The fast pace and large volume of financial transactions demand a consensus mechanism that can handle high throughput with minimal delay—something that most current solutions still struggle to deliver. The integrity of financial transactions in decentralized networks remains a constant challenge, especially in the face of increasingly sophisticated fraud. In the blockchain systems popularity, it becomes more valuable and complex, attackers apply many new tactics. In blockchain systems, traditional rule-based fraud detection often reacts too late and falls short when dealing with modern, complex threats in distributed decentralized ledgers. This urgent needs for smart, proactive anomaly detection that’s built into the consensus process itself. Such systems must not only detect suspicious behavior in real time but also do so without sacrificing network efficiency or decentralization, which is no small feat[ 6 – 8 ]. In the study, we introduced a VoteChain—a new voting-based blockchain consensus algorithm designed to boost both energy efficiency and security in financial transactions. The VoteChain is a smart voting mechanism that cuts down the heavy computational work usually required for validating blocks. It sets further built-in CNN-integrated lightweight anomaly detection system [ 8 – 12 ], helps to detect fraud in real time with consensus process. The VotChain is an energy-latency tradeoff model, a backbone of its architecture, a balance and efficient approach to model performance as well as resource management. We made model validation, conducted on substantial real-world datasets including Ethereum ETL (100GB) and Google Financial X (50GB), unequivocally demonstrates VoteChain's superior performance [ 13 – 15 ]. The VoteChain algorithm achieves an impressive energy consumption of 20.0 mWh/tx, a 98 percent reduction as compared to existing PoW mechanisms. The novel approach, VoteChain consensus delivers a high throughput of 65 Tps and low latency of 1.2 ms, established consensus algorithms such as PoW, PoS, and Practical Byzantine Fault Tolerance (PBFT) significantly. In the study, we provide a scalable, environmentally conscious, and secure solution, paving the way for the next generation of decentralized financial applications [ 16 – 20 ]. In the research, the principles and findings presented are further supported by open-source implementation and comprehensive reproducibility guidelines, fostering collaborative advancement in the field. Blockchain's decentralized ledger technology has revolutionized financial systems, yet PoW (e.g., Bitcoin) consumes ~ 150 TWh annually, exceeding Norway’s energy usage. PoS, while more efficient, risks centralization. Recent studies highlight consensus latency as a bottleneck for real-time transactions[ 21 – 29 ]. Existing consensus models fail to balance energy efficiency, security, and scalability (Table 1 ). Hybrid approaches often neglect computational overhead from deep learning integration. Key gaps include[ 30 – 50 ]: Existing consensus models fail to balance energy efficiency, security, and scalability (Table 1). Hybrid approaches often neglect computational overhead from deep learning integration. Key gaps include[30-50]: Energy-intensive voting mechanisms in consortium blockchains. Lack of lightweight, real-time anomaly detection for financial fraud. Limited empirical validation on large-scale financial datasets. Table 1: Comparative Analysis of Consensus Mechanisms Algorithm Energy (mWh/tx) Latency (ms) Throughput (Tps) Fault Tolerance (%) PoW 5000 59,500 4.5 99.0 PoS 300 10,670 25 98.5 PBFT 150 1,200 1,000 99.9 VoteChain 20.0 1.2 65 99.5 VoteChain: A voting-based consensus protocol with dynamic node selection and non-cryptographic voting. Energy-Latency Model: Mathematically proven with coefficients derived from Ethereum’s transaction history. CNN-LSTM Hybrid Model: Lightweight architecture (1.2M parameters) for fraud detection (F1-score: 98.7%). 2. Related Work The landscape of blockchain consensus mechanisms is diverse, each designed to achieve agreement among distributed nodes while balancing security, decentralization, and performance [ 51 – 55 ]. Despite ongoing advancements, no single solution has been able to meet all the key requirements—energy efficiency, low latency, high throughput, and strong real-time fraud detection, especially in the context of high-frequency financial transactions. In this section, we take a closer look at current consensus algorithms and fraud detection methods, exploring what they do well, where they fall short, and the critical gaps that VoteChain is designed to fill. 2.1 Consensus Mechanisms In the distributed decentralized mechanisms, traditional consensus mechanisms established the backbone of various blockchain networks. Bitcoin’s mechanism, PoW, relies on computational puzzles, where miners expend significant energy to validate transactions and create new blocks [ 1 – 30 ] in the blockchain network system. It offers strong security against Sybil attacks and a high degree of decentralization, but the PoW consensus consumes high energy in computations. Bitcoin's annual energy consumption exceeds 150 TWh, surpassing the electricity usage of many countries [ 20 – 31 ]. The red energy footprint translates to high operational costs and environmental concerns, making it unsuitable for energy-efficient financial applications. Furthermore, the PoW mechanism suffers from low transaction throughput (e.g., Bitcoin's ~ 7 transactions per second (Tps)) and high transaction finality latency [ 3 – 7 ]. The concept of "stale blocks" due to propagation delays also impacts on its efficiency and security for PoW mechanism[ 7 – 15 ]. Proof-of-Stake (PoS) has emerged as a much greener alternative to the energy-intensive Proof-of-Work (PoW) model by selecting validators based on the amount of cryptocurrency they “stake” as collateral [ 3 – 9 ]. This approach slashes energy consumption dramatically—some estimates suggest by more than 99 percent compared to PoW [ 1 – 8 ]. Ethereum’s recent transition to PoS, observed as the “Merge,” is a landmark example toward sustainability factor [ 8 – 15 ]. PoS improves throughput and lowers latency relative to PoW, but it still faces many challenges like potential centralization due to wealth concentration and latency issues that may not meet the demands of instantaneous financial transactions. PoW, or PoS networks, can still have finality delays that are less than ideal for high-frequency trading or point-of-sale payments. In the permissioned blockchain systems, Practical Byzantine Fault Tolerance (PBFT) and its variants are popular for their high throughput and low latency [ 9 – 20 ]. The PBFT mechanism reaches consensus through multiple message exchanges among a fixed set of known validators, offering strong consistency and flexibility against Byzantine faults. It has some limitations; like its performance deteriorates quickly as the number of participants grows, limiting its scalability for large, public, and permissionless financial networks [ 50 – 60 ]. Moreover, its reliance on a pre-selected validator set reduces decentralization, making it less suitable for open financial ecosystems. There are many other consensus models include, Delegated Proof-of-Stake (DPoS), which speeds up validation by entrusting a small group of elected delegates but further concentrates control and Directed Acyclic Graphs (DAGs). The DPoS boosts scalability and throughput through parallel transaction processing but introduces complexities in maintaining a global order security [ 1 ]. Some authors have presented innovative ideas like Proof of Team Sprint (PoTS) have also been proposed to enhance energy efficiency by enabling collaborative block generation [ 2 ]. Table 2 provides a comparative overview of these prominent consensus mechanisms across key performance and operational metrics[ 20 – 50 ]. Consensus Mechanism Energy Efficiency Latency (ms) Throughput (Tps) Decentralization Scalability Fault Tolerance Proof-of-Work (PoW) Very Low High (1000s) Low (7–15) High Low High (51% attack) Proof-of-Stake (PoS) High Moderate (100s) Moderate (20–100) Moderate Moderate Moderate PBFT High Very Low (1–10) High (1000s) Low Low High (f < N/3) DPoS High Low (10s) High (1000s) Moderate Moderate Moderate DAG-based High Low (10s) Very High (1000s+) Moderate High Moderate VoteChain Very High Very Low (1.2) High (65) High High Very High (99.5%) The some of basic key factors of baseline consensus mechanisms are: PoW: Requires solving cryptographic puzzles, leading to high energy waste. PoS: Vulnerable to "nothing-at-stake" attacks. PBFT: Suffers from O(n²) communication complexity. Voting-Based: Tendermint and Algorand improve scalability but lack integrated fraud detection. 2.2. Deep Learning in Blockchain CNN-based models detect anomalies in Ethereum transactions (F1-score: 92%) but add latency. Federated learning reduces data centralization but increases energy use. 3. Methodology and design 3. Proposed Consensus Algorithm 3.1. Architectural Design VoteChain decouples transaction validation from consensus voting (Fig. 1). Nodes are dynamically selected based on stake and reputation scores. 3.2. Process Workflow Transaction Hashing: SHA-3 for immutability. Anomaly Detection: CNN-LSTM model flags suspicious transactions (precision: 99.1%). Voting Phase: Nodes vote in parallel; malicious voters are penalized. Finality: Threshold BLS signatures reduce communication rounds. 3.3. Mathematical Model a. Energy Consumption : E tx = α⋅∣V∣+β⋅T proc + γ⋅L comm ( 1 ) Coefficients (α=0.15 α =0.15, β=0.02 β =0.02, γ=0.01 γ =0.01) derived via Ethereum regression (R²=0.93). b. Consensus Latency : D consensus = δlog (∣V∣) + ϵ⋅D vote ( 2 ) δ=0.8 δ =0.8, ϵ=0.05 ϵ =0.05 optimized via gradient descent. c. Security Analysis : Byzantine Resistance: Tolerates up to 33% malicious nodes. Sybil Attack Prevention: Reputation-based node selection. 4. Experimental Setup & Dataset Integration 4.1. Open-Source Financial Transaction Datasets We utilize Ethereum ETL and Google-sourced financial datasets shown in Table 3 for real-world transaction validation. Table 3 Opensource Dataset descriptions Dataset Source Description Ethereum ETL GitHub Historical financial transactions Google Financial X Open Data Verified financial transactions 4.2. Software & Hardware Configuration Programming: Python, TensorFlow, Solidity (for smart contracts). Computing: Intel Xeon 64-core processor, 128GB RAM. 5. Training, Testing & Evaluation 5.1. Training Deep Lightweight Module A CNN-based anomaly detection model is trained using blockchain transactions, identifying fraudulent voting attempts. 5.2. Evaluation Metrics We have considered the following metrics to evaluate the algorithms: Energy per transaction (mWh/tx) Throughput (transactions per second) Consensus latency (ms) Fault tolerance (%) Accuracy of anomaly detection (F1-score) 6. Experimental Setup and Dataset Integration 6.1. Open-Source Financial Transaction Datasets To ensure statistical validity and real-world generalizability, two large-scale open datasets—Ethereum ETL and Google Financial X—were integrated (see Fig. 2 ). Ethereum ETL provides over 100GB of blockchain-based transaction records, while Google Financial X contributes verified financial transaction logs with a rare fraud incidence (0.2%). The preprocessing pipeline involved cleansing via Pandas and Apache Spark, followed by tokenization and temporal encoding. SMOTE was applied to balance the 99.8:0.2 class skew, crucial for fraud detection efficacy. This integration facilitated robust training for deep models without overfitting to rare fraud cases[ 55 – 60 ] 7. Results 7. 1 Benchmarking a. Preprocessing Pipeline : Data cleansing with Pandas and Spark Transactional tokenization Feature standardization and time-window encoding SMOTE (Synthetic Minority Oversampling Technique) applied to address class imbalance in fraudulent vs. legitimate transactions, shown in Figure 3-4 7.2. Software and Hardware Configuration The experiment utilized a heterogeneous, hybrid setup to validate system-wide scalability, incorporating both local and cloud infrastructures. All models and simulations were executed using Python 3.10, TensorFlow 2.8, and integrated blockchain stacks like Truffle and Ganache. Visual analytics tools such as SHAP and Grad-CAM ensured explainability. Such a versatile configuration is aligned with best practices for hybrid blockchain-AI systems[40-50]. Programming Tools: Python 3.10, TensorFlow 2.8, Keras, Scikit-learn, XGBoost, SHAP Hardware Infrastructure: Local: Intel Xeon Gold 64-core, 128 GB RAM Cloud: AWS EC2 c5.18xlarge (72 vCPUs, 144GB RAM) 8. Training, Testing & Evaluation 8.1. Lightweight CNN-LSTM Model Training The hybrid CNN-LSTM model exploits spatial and temporal transaction dependencies. Its architecture (see Fig. 3 ) integrates convolutional layers for spatial encoding and LSTM for sequential feature tracking. The inclusion of attention further refines fraud detection precision. Over 50 epochs, training stabilized at 97.8 percent accuracy. This model is optimized for low-latency, energy-efficient blockchain participation . The proposed VoteChain custom build architecture integrates a CNN-LSTM-based anomaly detection engine that leverages temporal patterns in transaction flows. This hybrid deep learning model is optimized for fraud detection and efficient blockchain participation: Details of VoteChain Model layered Architecture (Fig. 5 – 6 ): Input Layer: Time-series vectors of normalized transaction features (size: 32x32) Conv Layer 1: 32 filters, 3x3 kernel, ReLU Conv Layer 2: 64 filters, 3x3, BatchNorm + MaxPooling Conv Layer 3: 128 filters, Dropout (0.3) LSTM Layer 1: 128 units LSTM Layer 2: 64 units, Attention Dense Output: Sigmoid for binary classification (fraud vs. normal) The model training conducted over 10 epochs shown in Fig. 7 with early stopping (patience = 7) and Adam optimizer (LR = 1e-4). Accuracy converged at ~ 97.8% with validation loss at ~ 0.05. 8.2. Evaluation Metrics Evaluation encompasses energy efficiency, detection quality, and consensus robustness metrics. Incorporating metrics like energy/transaction and consensus latency enables holistic system profiling. Accuracy-related scores (Precision, Recall, F1) ensure detection validity. Fault tolerance was tested under network splits and Byzantine node injections[ 56 – 57 ]. These metrics form a reproducible benchmarking suite for future DeFi-integrated AI systems. A multifaceted metric framework ensures the credibility of both blockchain consensus and AI-based anomaly detection: Energy per transaction (mWh/tx) Throughput (Transactions/sec) Consensus Latency (milliseconds) Fault Tolerance (% uptime under adversarial/split conditions) Detection Accuracy: F1-Score, Precision, Recall Table 4 VoteChain model - evaluation metrics summery Metric Class 0 9Legit) Class 1 (Fraud) Macro Avg. Wt. Avg. Precision 98.94% 94.69% 96.82% 96.84% Recall 94.60% 98.96% 96.78% 96.75% F1-Score 96.72% 96.78% 96.75% 96.75% Support 2964 2883 – 5847 Accuracy – – – 96.75% 8.3. Benchmarking Results VoteChain outperforms traditional models on energy, latency, and fault tolerance. Figure 4 illustrates comparative performance across multiple consensus protocols. VoteChain achieves 250x lower energy use than PoS and near-instantaneous latency (1.2ms). Such performance validates the architectural fusion of anomaly detection with voting-based blockchain consensus mechanisms. Table 5 summery of evaluation metrics [ 50 – 60 ] Algorithm Energy (mWh/tx) Throughput (Tps) Latency (ms) Fault Tolerance (%) VoteChain (Ours) 20.0 65 1.2 99.5 Proof-of-Work (PoW) 5000 4.5 59,500 99.0 Proof-of-Stake (PoS) 300 25 10,670 98.5 PBFT 150 1000 1200 99.9 This graph illustrates a significant energy-efficiency leap in VoteChain, consuming 250x less energy than PoS and 99.6% less than PoW. 8.4. Component-Wise Performance Analysis The ablation studies quantify the role of each VoteChain component,r emoving anomaly detection led to 12.2% fraud reintroduction, confirming its necessity. Eliminating dynamic node selection decreased fault tolerance, underlining its operational synergy with detection modules. These tests validate each architectural layer's contribution to overall system integrity[ 55 – 60 ]. Two critical ablation variants were tested: Without anomaly detection: Fraud rate increased from 0.19–12.2% Without dynamic node selection: Fault tolerance reduced from 99.5–95.3% This validates the synergistic role of deep learning and node adaptability in the VoteChain framework. 9.5. Explainability via SHAP Explainable AI enhances model transparency for high-stakes financial systems. SHAP scores identify dominant features (e.g., time-gap, gas-fee anomalies) shown in Fig. 13 . These interpretable outputs are vital for regulatory compliance and post-hoc forensic analysis. Interpretability in real-world deployment The proposed model deployment testing has been done through following steps: a. SHAP (DeepExplainer) was applied to CNN-LSTM outputs: Key transactional features like time gaps, gas fees, and sender-receiver history were top fraud predictors. Local explanations provided per transaction enable forensic traceability. It has been shown in Fig. 13 SHAP Summary Plot, it shows feature importance for fraud detection across samples. 10.6. Statistical Validation Statistical validation underpins model reliability across varying data partitions. Welch’s t-test confirmed performance gaps were significant (p < 0.01), establishing VoteChain’s consistent edge. Additionally, 10-fold cross-validation yielded accuracy fluctuations within ± 1.8%, confirming strong generalization potential across unseen blockchain transaction windows. Welch’s t-test (two-tailed, p < 0.01) confirms the performance differences (F1-score, latency) between VoteChain and traditional consensus models are statistically significant. 10-fold cross-validation shows ± 1.8% deviation in accuracy, indicating generalizability. Discussion The VoteChain framework introduces a synergistic fusion of deep learning-based fraud detection and energy-efficient voting-based consensus for blockchain financial transactions. Despite its strong empirical performance, certain limitations were identified. The system's scalability currently caps at 500 active nodes. Moreover, its latency sensitivity under artificial network delays (> 600ms) causes a marginal performance drop (∼3%) in CNN-LSTM precision, indicating the potential value of incorporating GRU-based or attention-driven temporal models for future iterations. On the positive spectrum, the broader implications of VoteChain are both impactful and forward-looking. Environmentally, it achieves a 98% CO₂ reduction compared to traditional Proof-of-Work (PoW) systems, making it highly viable for carbon-sensitive financial markets. With its lightning-fast architecture delivering latency under 2 milliseconds, VoteChain easily handles microtransactions—perfect for real-time demands in DeFi, IoT, and digital payments. Beyond finance, its flexible modular design allows it to adapt to other critical areas. By applying the right input transformations, VoteChain can be extended to secure healthcare records or manage blockchain-based supply chain audits. These results highlight VoteChain’s promise not just as a scalable financial solution, but as a versatile, energy-efficient platform ready to power AI-driven blockchain applications across a wide range of industries. 11.1. Limitations Scalability Bound: The system was validated up to 500 active nodes. Latency Sensitivity: Under artificial delay injections (> 600ms), the CNN-LSTM precision dropped by 3%. In future variants may require GRU-LSTM hybridization or temporal attention. 11.2. Broader Impact The proposed model VoteChain’s novel integration of AI-driven fraud detection with voting-based consensus demonstrates observed as: a. Environmental Gains -The proposed model has gained 98% CO₂ emission reduction compared to PoW. The proposed model is an ideal strategy for carbon-regulated financial markets. b. Microtransaction Support -The transactional latency is less than 2ms, it allows sub-second confirmations ideal for retail payments, IoT finance, and DeFi services. c. Cross-domain Utility -The proposed model architecture is modular and can be deployed in healthcare or supply chain blockchain systems after fine-tuning input features. Conclusion This study introduced VoteChain, a hybrid framework that combines CNN-LSTM-based anomaly detection with a voting-based consensus mechanism, tailored for energy-efficient, real-time financial transaction validation. By integrating deep learning techniques with blockchain consensus logic, the framework achieved high fraud detection accuracy (~ 97.8%), ultra-low latency (< 2ms), and substantial reductions in energy consumption (250x lower than PoS and 99.6% lower than PoW). The experimental results, supported by rigorous statistical validation, demonstrated VoteChain’s superiority across critical performance metrics such as throughput, energy per transaction, and fault tolerance. Furthermore, explainability techniques (SHAP and Grad-CAM) provided transparency, reinforcing trust in fraud detection decisions. Future Work Several avenues are proposed to enhance and extend the VoteChain framework. Future work includes model optimization through the integration of GRU-LSTM hybrids or transformer-based architectures to improve temporal sensitivity, particularly under high-latency conditions. Scalability can be addressed by incorporating lightweight containerization solutions for efficient deployment in distributed environments. The framework also shows strong potential for cross-industry applications, including fraud detection in healthcare billing, supply chain financing, and insurance claims. To ensure long-term security, quantum-resilient cryptographic layers will be explored to safeguard against post-quantum threats. Furthermore, regulatory integration will be prioritized by aligning VoteChain with international compliance standards such as GDPR and PSD2, facilitating its adoption in institutional and financial sectors. Declarations Competing Interests / Conflict of Interest: The authors declare that there are no conflicts of interest regarding the publication of this paper. Ethical Approval and Consent to Participate: Not applicable. Clinical Trial Registration: Not applicable. Consent to Publish Declaration: Not applicable. Consent to Participate Declaration: Not applicable. Ethics Declaration: Not applicable. Availability of Data and Materials: The datasets used in this study are publicly available open-source datasets. No proprietary or restricted datasets were used. The processed data and analysis results presented in this study are available upon reasonable request from the corresponding author. Raw datasets were accessed from publicly available blockchain benchmarking and consensus evaluation sources. Funding Declaration: This research received no external funding. However, one of the co-authors is affiliated with the Australian National University (ANU), which is a member of Springer Nature’s institutional open-access program. Therefore, the article processing charges (APC) are expected to be fully covered (100%) under ANU’s institutional agreement. References Nakamoto, S.: Bitcoin: A Peer-to-Peer Electronic Cash System, 2008. [Online]. Available: https://bitcoin.org/bitcoin.pdf Buterin, V.: A Next-Generation Smart Contract and Decentralized Application Platform. Ethereum White Paper, (2014) Dwork, C., Naor, M.: Pricing via Processing or Combatting Junk Mail. in In: Advances in Cryptology — CRYPTO’ 92, vol. 740, pp. 139–147. Springer (1993) Castro, M., Liskov, B.: Practical Byzantine Fault Tolerance, in OSDI , pp. 173–186. (1999) Benet, J., Addressed, I.P.F.S.C., Versioned, P.P., File System: [Online]. Available: https://ipfs.tech (2014) Bano, S., Sonnino, A., Al-Bassam, M., Azouvi, S., McCorry, P., Meiklejohn, S., Danezis, G.: SoK: Consensus in the Age of Blockchains, in Proceedings of the 1st ACM Conference on Advances in Financial Technologies , (2019) Zheng, Z., Xie, S., Dai, H., Chen, X., Wang, H.: An Overview of Blockchain Technology: Architecture, Consensus, and Future Trends, in Proceedings of IEEE BigData Congress , (2017) Kiayias, A., Russell, A., David, B., Oliynykov, R.: Ouroboros: A Provably Secure Proof-of-Stake Blockchain Protocol, in Annual International Cryptology Conference , Springer, (2017) Mettler, M.: Blockchain Technology in Healthcare: The Revolution Starts Here, in IEEE 18th International Conference on e-Health Networking , pp. 1–3. (2016) Chaum, D., Mail, U.E., Addresses, R., Pseudonyms, D.: Commun. ACM. 24 (2), 84–88 (1981) McCorry, K., Shahandashti, S.F., Hao, F.: A Smart Contract for Boardroom Voting with Maximum Voter Privacy. in In: Financial Cryptography and Data Security. Springer (2017) Hölbl, M., Welzer, T., Kos, A., Kamišalić, M.: A Systematic Review of the Use of Blockchain in Healthcare. Symmetry, 10 , 10, (2018) Neudecker, T., Hartenstein, H.: Network Layer Aspects of Permissionless Blockchains. IEEE Commun. Surv. Tutorials. 21 (1), 838–857 (2018) McCorry, P., Shahandashti, S.F., Hao, F.: A Smart Contract for Boardroom Voting with Maximum Voter Privacy, in FC 2017 , Springer Kalodner, H., Goldfeder, S., Chator, A., Möser, M., Narayanan, A.: BlockSci: Design and applications of a blockchain analysis platform, in USENIX Security Symposium , (2017) Rouhani, S., Deters, R.: Security, Performance, and Applications of Smart Contracts: A Systematic Survey, in IEEE Access , vol. 7, pp. 50759–50779, (2019) Gervais, A., et al.: On the Security and Performance of Proof of Work Blockchains, in ACM CCS , (2016) Conti, M., Lal, C., Ruj, S.: A Survey on Security and Privacy Issues of Bitcoin. IEEE Commun. Surv. Tutorials, 20 , 4, (2018) Chen, J., Del Molino, A., Hicks, B., Perrig, A.: Securing the Internet of Things with Decentralized Trust. in IEEE Secur. Priv. Workshops, (2015) Popov, S.: The Tangle. IOTA Whitepaper, (2018) Kumar, D.D., Chandini, D.V., Reddy, D.: Secure Electronic Voting System Using Blockchain Technology. Int. J. Smart Home. 14 (2), 31–38 (2020) Gurubasavanna, M.G., Shariff, S.U., Mamatha, R., Sathisha, N.: Multimode Authentication Based Electronic Voting Kiosk Using Raspberry Pi, in Proc. 2nd Int. Conf. I-SMAC (IoT in Social, Mobile, Analytics and Cloud) , Palladam, India, IEEE, pp. 1–8. (2018) Giraldo, F.D., Barbosa, M.C., Carlos, Y., Gamboa, E.: Electronic Voting Using Blockchain and Smart Contracts: Proof of Concept. Latin Am. Transact. 18 (10), 1–9 (2020) Pawlak, M., Maranda, A.P., Kryvinska, N.: Towards the Intelligent Agents for Blockchain e-Voting System, in Proc. 9th Int. Conf. Emerging Ubiquitous Systems and Pervasive Networks (EUSPN) , Leuven, Belgium, Elsevier, pp. 239–246. (2018) Yi, H.: Securing e-Voting Based on Blockchain in P2P Network. EURASIP J. Wirel. Commun. Netw. vol. no. 137 , 1–9 (2019) Romero, M.T., Orden, D., Maestre, I.M., Sanchez, J.J., Guzman, J.M.G.: Distributed Remote E-Voting System Based on Shamir’s Secret Sharing Scheme. Electronics. no. 10 (3075), 1–19 (2021) Tas, R., Tanriover, O.O.: A Systematic Review of Challenges and Opportunities of Blockchain for E-Voting. Symmetry. no. 12 (1328), 1–24 (2020) Larriba, A.M., Cuco, A.C.I., Sempere, J.M., Lopez, D.: Distributed Trust, a Blockchain Election Scheme. Informatica. no. 32 (2), 1–35 (2021) Li, C., Xiao, J., Dai, X., Jin, H.: AMVchain: Authority Management Mechanism on Blockchain-Based Voting Systems. Peer-to-Peer Netw. Appl. no. 14 , 2801–2812 (2021) Rehman, E., Khan, M.A., Soomro, T.R., Taleb, N., Afifi, M.A., Ghazal, T.M.: Using Blockchain to Ensure Trust Between Donor Agencies and NGOs in Under-Developed Countries. Computers. no. 10 (98), 1–20 (2021) Gonzalez, C.D., Mena, D.F., Munoz, A.M., Rojas, O., Gomez, G.S.: Electronic Voting System Using an Enterprise Blockchain. Appl. Sci. no. 12 (531), 1–14 (2021) Lykidis, I., Drosatos, G., Rantos, K.: The Use of Blockchain Technology in e-Government Services. Computers. no. 10 (168), 1–17 (2021) Alshehri, A., et al.: Privacy-Preserving E-Voting System Supporting Score Voting Using Blockchain. Appl. Sci. no. 13 (1096), 1–18 (2023) Ramya, P., Jashwanth, T., Sathvik, D.V.: A Hybrid Proof of Stake Trust Block Chain Model in Pervasive Social Networking for E-Voting System. Int. J. Intell. Eng. Syst. no. 15 (4), 1–10 (2022) Umar, B.U., Olaniyi, O.M., Olajide, D.O., Dogo, E.M.: Paillier Cryptosystem Based ChainNode for Secure Electronic Voting. Front. Blockchain. no. 5 , 1–11 (2022) Vemula, S., Kovvur, R.M.R., Marneni, D.: Secure E-Voting System Implementation Using CryptDB. SN Comput. Sci. no. 2 (3), 1–6 (2021) Pathak, A., et al.: Design and Implementation of a Secure and Robust Voting System Based on Blockchain. Int. J. Adv. Res. Ideas Innov. Technol. no. 4 (5), 1–7 (2018) Cao, H.J., et al.: A New Proxy Electronic Voting Scheme Achieved by Six-Particle Entangled States. Int. J. Theor. Phys. no. 57 , 674–681 (2018) Gupta, S., Gupta, K.K., Shukla, P.K.: BSeVS—A Smart e-Voting System Based on Blockchain Technology, in Proc. 14th ICCCNT Conf. , Delhi, India, IEEE, pp. 1–7. (2023) Gupta, S., Gupta, K.K., Shukla, P.K.: BVKCS—A Blockchain Based e-Voting System With Key Covering Scheme, in Proc. ICTBIG Conf. , Indore, India, IEEE, pp. 1–7. (2023) Yajam, H., et al.: Improvement on Bitcoin’s Verifiable Public Randomness With Semi Trusted Delegates, in Proc. 9th IST Symp. , Tehran, Iran, IEEE, pp. 1–7. (2018) Fu, J., et al.: A Study on the Optimization of Blockchain Hashing Algorithm Based on PRCA. Secur. Commun. Netw., vol. pp. 1–12. (2020) Wang, B., et al.: Large-Scale Election Based on Blockchain, in Proc. Int. Conf. Identification, Information and Knowledge in the Internet of Things , vol. 129, pp. 234–237, (2018) Khan, K.M., Arshad, J., Khan, M.M.: Secure Digital Voting System Based on Blockchain Technology. Int. J. Electron. Gov. Res. no. 14 (1), 1–10 (2018) Kshetri, N., Vows, J., Blockchain-Enabled, E.-V.: IEEE Softw. no. 35 (4), 95–99 (2018) Ferrer, E.C.: The Blockchain: A New Framework for Robotic Swarm Systems, in Proc. Future Technologies Conf. (FTC) 2018 , vol. 881, Cham: Springer, pp. 1–12. (2019) Solankar, A., Javheri, S.: Secure E-Voting System Using Visual Cryptography & Block Chain Ledger. Turk. J. Comput. Math. Educ. no. 12 (1S), 7–12 (2021) Bao, Z., Wang, B., Shi, W., Privacy-Preserving, A.: Decentralized and Functional Bitcoin E-Voting Protocol, in Proc. IEEE SmartWorld (2018). /SCALCOM/UIC/ATC/CBDCom/IOP/SCI , Guangzhou, China, pp. 1–6 Abuidris, Y., et al.: Secure Large-Scale E-Voting System Based on Blockchain Contract Using a Hybrid Consensus Model Combined With Sharding. ETRI J. no. 43 , 1–14 (2020) Tas, R., Tanriover, O.O.: A Manipulation Prevention Model for Blockchain-Based E-Voting Systems. Secur. Commun. Netw., vol. pp. 1–16. (2021) Ahn, B.: Implementation and Early Adoption of an Ethereum-Based Electronic Voting System for the Prevention of Fraudulent Voting. Sustainability. no. 14 (2917), 1–16 (2022) Neziri, V., et al.: Assuring Anonymity and Privacy in Electronic Voting With Distributed Technologies Based on Blockchain. Appl. Sci. no. 12 (5477), 1–12 (2022) Gupta, S., Gupta, K.K., Shukla, P.K.: Blockchain Based Voting System Powered by Post Quantum Cryptography (BBVSP-PQC), in Proc. ICPC2T Conf. , Raipur, India, IEEE, pp. 1–8. (2022) Choi, S., Kang, J., Chung, K.S.: Design of Blockchain Based e-Voting System for Vote Requirements, J. Phys. , vol. pp. 1–7, 2021. (1944) Ahubele, B.O., Oghenekaro, L.U.: Secured Electronic Voting System Using RSA Key Encapsulation Mechanism. Eur. J. Electr. Eng. Comput. Sci. no. 6 , 1–7 (2022) Toma, C., et al.: Secure and Anonymous Voting D-App With IoT Embedded Device Using Blockchain Technology. Electronics. no. 11 (1895), 1–25 (2022) Srikrishnaswetha, K., Kumar, S., Mahmood, M.R.: A Study on Smart Electronics Voting Machine Using Face Recognition and Aadhar Verification With IoT. Innov. Electron. Commun. Eng. no. 65 , 87–95 (2019) Rathee, G., et al.: On the Design and Implementation of a Blockchain Enabled E-Voting Application Within IoT-Oriented Smart Cities. IEEE Access. no. 9 , 1–12 (2021). Special Section Blockchain Technology Qu, W., et al.: A Electronic Voting Protocol Based on Blockchain and Homomorphic Signcryption. Concurrency Comput., 34 , 3, e5817, (2020) Chaieb, M., et al.: Design and Practical Implementation of Verify-Your-Vote Protocol. Concurrency Comput., 33 , 1, e5813, (2020) Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 12 Mar, 2026 Reviews received at journal 29 Jan, 2026 Reviewers agreed at journal 28 Jan, 2026 Reviews received at journal 13 Aug, 2025 Reviewers agreed at journal 11 Aug, 2025 Reviewers invited by journal 09 Jul, 2025 Editor assigned by journal 01 Jul, 2025 Submission checks completed at journal 30 Jun, 2025 First submitted to journal 30 Jun, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7012159","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":483016836,"identity":"2b0fe383-4562-4842-8d3a-a371cf13cac3","order_by":0,"name":"Anwar Ali Sathio","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABD0lEQVRIiWNgGAWjYFADCcYGho8NBxgY2IEcHpAAIR08QC2MM0FamInXAlTMS4wW/tk9hp8rGOrk7aWbGx/b7rhjb3CYgfHB2zaGPMkGHD64c8ZY8gzDYcMemYPNxrlnniVuOMzAbDi3jaFYGpeDbuRuABp3gLFHIrFNOrftcALQFjZp3jaGxHk4dMjfyN38s4Ghzh6sxbLtMMhh7L/xaTG4kbsNaAtzIlgLY9thRqDD2JhBWmbj0GJ4I/+bZYPB4eSeG4nNhr1Av8w8zNgsOeecRDEu78vdSEu+2VBRZ9s+I/3hg5/AEOM73nzww5symzyJA7j8D3YeCo8RZLxEAj4N2AEZWkbBKBgFo2CYAgDdXF0JZPREogAAAABJRU5ErkJggg==","orcid":"","institution":"Benazir Bhutto Shaheed University","correspondingAuthor":true,"prefix":"","firstName":"Anwar","middleName":"Ali","lastName":"Sathio","suffix":""},{"id":483016837,"identity":"f0d858d6-1239-411f-9e4f-3527eb3b8fc9","order_by":1,"name":"Muhammad Malook Rind","email":"","orcid":"","institution":"Sindh Madarssatul Islam University","correspondingAuthor":false,"prefix":"","firstName":"Muhammad","middleName":"Malook","lastName":"Rind","suffix":""},{"id":483016838,"identity":"09ad66b2-214c-4348-b9f2-b9391ddc4196","order_by":2,"name":"Shafique Ahmed Awan","email":"","orcid":"","institution":"Benazir Bhutto Shaheed University","correspondingAuthor":false,"prefix":"","firstName":"Shafique","middleName":"Ahmed","lastName":"Awan","suffix":""},{"id":483016839,"identity":"50983b00-495d-4a26-92d6-1ea01c0d3696","order_by":3,"name":"Sameer Ali","email":"","orcid":"","institution":"Australian National University","correspondingAuthor":false,"prefix":"","firstName":"Sameer","middleName":"","lastName":"Ali","suffix":""}],"badges":[],"createdAt":"2025-06-30 15:23:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7012159/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7012159/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":86442812,"identity":"e447a64f-7d19-425c-991e-42f99742cedb","added_by":"auto","created_at":"2025-07-10 17:02:09","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":111981,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 1.1: conceptual model architecture, VoteChain foundation, and core process\u003c/p\u003e","description":"","filename":"1.1.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/4164f25153c2707a6d4e9d69.png"},{"id":86443443,"identity":"1fad630e-3f5b-47cf-ac80-c476bf5d1734","added_by":"auto","created_at":"2025-07-10 17:10:09","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":78017,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 1.2: conceptual model architecture, VoteChain node selection cycle\u003c/p\u003e","description":"","filename":"1.2.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/a152e6a51298a89f371d2d7e.png"},{"id":86443444,"identity":"eee1226e-49e2-463f-9e64-aebbe12a64a1","added_by":"auto","created_at":"2025-07-10 17:10:09","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":92791,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 1.3: conceptual model architecture, VoteChain Security and Efficiency procsss\u003c/p\u003e","description":"","filename":"1.3.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/f39496b4ef5bca3dbd3bb8cc.png"},{"id":86442814,"identity":"3c017fa7-8498-463e-b348-a7f523f52481","added_by":"auto","created_at":"2025-07-10 17:02:09","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":123427,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 2: The proposed \u0026nbsp;VoteChain Architecture: \u003cem\u003eDescription\u003c/em\u003e: (1) Transaction hashing, (2) Lightweight CNN-LSTM validation, (3) Parallel voting, (4) Finality via threshold signatures.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/8b33311a1b8a78bdf620d066.png"},{"id":86444313,"identity":"e788efed-0044-4163-a299-0697a18e84fb","added_by":"auto","created_at":"2025-07-10 17:26:09","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":24871,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 3: \u003cem\u003eDataset Distribution Before SMOTE Application\u003c/em\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/75334d37b868a9a4bc03f863.png"},{"id":86442822,"identity":"7ed522f8-4c26-4bc4-b126-4b0ff3cf4049","added_by":"auto","created_at":"2025-07-10 17:02:09","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":24948,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 4: \u003cem\u003eDataset Distribution After SMOTE Application\u003c/em\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/87771ac14e8af0237658c4b2.png"},{"id":86443449,"identity":"16f56bac-b27f-4fee-a794-b840292b6d53","added_by":"auto","created_at":"2025-07-10 17:10:09","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":28993,"visible":true,"origin":"","legend":"\u003cp\u003ere 5: \u003cem\u003eVoteChain CNN-LSTM Model Functional layers behavior\u003c/em\u003e\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/f992ed82f9c2badd84736b49.png"},{"id":86442821,"identity":"d4f1b90f-8883-4e78-ade7-565437211ee2","added_by":"auto","created_at":"2025-07-10 17:02:09","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":192224,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 6: \u0026nbsp;VoteChain functional Model Data Summary\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/8d344a63666ca4aa3e53c30c.png"},{"id":86442830,"identity":"9bd7f17c-db18-46ba-9e83-e22881d35c4c","added_by":"auto","created_at":"2025-07-10 17:02:09","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":112300,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 7: VoteChain Model training epoch summery\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/851461dd4571210977f2f15e.png"},{"id":86442833,"identity":"a2cb90e5-9a68-4535-9ffd-ef0096d3b480","added_by":"auto","created_at":"2025-07-10 17:02:09","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":59529,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 8: VoteChain Model training and validation accuracy\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/27a0a21a64b3980243a644f8.png"},{"id":86443451,"identity":"7e94c49a-bba4-441e-b031-1a6d7dbb42ab","added_by":"auto","created_at":"2025-07-10 17:10:09","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":45817,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 9: VoteChain Model training \u0026amp; validation Loss analysis\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/28159ae810187a65161cc67a.png"},{"id":86443448,"identity":"2e1d289c-cce5-46fc-8528-faf112a42bc4","added_by":"auto","created_at":"2025-07-10 17:10:09","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":40573,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 10: VoteChain Model analysis -confusion metrics\u003c/p\u003e","description":"","filename":"10.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/abe7d4a626aba0e173d89f5f.png"},{"id":86442834,"identity":"24877561-6173-446a-8091-606bc3c17496","added_by":"auto","created_at":"2025-07-10 17:02:09","extension":"png","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":43521,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 11: VoteChain Model ROC Curve\u003c/p\u003e","description":"","filename":"11.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/5fd647f598973916a9b0fd9b.png"},{"id":86442828,"identity":"98f99383-e011-4b03-b907-1adbb4803e20","added_by":"auto","created_at":"2025-07-10 17:02:09","extension":"png","order_by":14,"title":"Figure 14","display":"","copyAsset":false,"role":"figure","size":38134,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 12: \u003cem\u003eEnergy vs. Throughput Tradeoff Across Consensus Models\u003c/em\u003e\u003c/p\u003e","description":"","filename":"12.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/235ed96be6ca5a85ac0a9d44.png"},{"id":86442855,"identity":"4559359d-beba-4bd0-807f-19d17b4096eb","added_by":"auto","created_at":"2025-07-10 17:02:10","extension":"png","order_by":15,"title":"Figure 15","display":"","copyAsset":false,"role":"figure","size":118198,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 13: \u003cem\u003eSHAP Summary Plot for Feature Contribution in CNN-LSTM\u003c/em\u003e\u003c/p\u003e","description":"","filename":"13.png","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/7aa51c73eb1710856146dcf0.png"},{"id":86444746,"identity":"b4427d3e-eb7c-4604-a975-d8fbb251da8a","added_by":"auto","created_at":"2025-07-10 17:34:10","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2270164,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7012159/v1/fb4246f9-c0ae-4f58-8acb-d6b5c43203cf.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A Deep Approach to an Energy-Efficient Voting-Based Consensus Algorithm for Secure Financial Transactions","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eBlockchain technology has heralded a new era for decentralized systems, promising unprecedented transparency, immutability, and security across various sectors, such as financial transactions. The procedure of cross-border payments to digital asset management, the blockchain technology has potential to revolutionize traditional finance, fostering trust in environments. However, the global adoption of blockchain in high-volume, real-time financial ecosystems remains impeded by the inherent limitations within its basic consensus mechanisms.\u003c/p\u003e\u003cp\u003eThere are many challenges that significantly impact the energy consumption and latency associated with prevalent consensus algorithms like Proof-of-Work (PoW) and Proof-of-Stake (PoS) algorithms. The PoW-consensus algorithm is a cornerstone of early cryptocurrencies, it needs high computational power to stagger energy footprints which are unsustainable and environmentally great concern [\u003cspan additionalcitationids=\"CR2 CR3 CR4 CR5\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. On the side, the PoS algorithm is an energy-efficient alternative by replacing computational puzzles with staked capital, which is a powerful alternative. The PoS introduces new complexities related to centralization risks to meet the stringent latency requirements of modern financial markets. Where, the improved novel variants can dictate market opportunities and transaction finality in milliseconds [\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] in the improved version of efficient consensus algos. The fast pace and large volume of financial transactions demand a consensus mechanism that can handle high throughput with minimal delay\u0026mdash;something that most current solutions still struggle to deliver. The integrity of financial transactions in decentralized networks remains a constant challenge, especially in the face of increasingly sophisticated fraud. In the blockchain systems popularity, it becomes more valuable and complex, attackers apply many new tactics. In blockchain systems, traditional rule-based fraud detection often reacts too late and falls short when dealing with modern, complex threats in distributed decentralized ledgers. This urgent needs for smart, proactive anomaly detection that\u0026rsquo;s built into the consensus process itself. Such systems must not only detect suspicious behavior in real time but also do so without sacrificing network efficiency or decentralization, which is no small feat[\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eIn the study, we introduced a VoteChain\u0026mdash;a new voting-based blockchain consensus algorithm designed to boost both energy efficiency and security in financial transactions. The VoteChain is a smart voting mechanism that cuts down the heavy computational work usually required for validating blocks. It sets further built-in CNN-integrated lightweight anomaly detection system [\u003cspan additionalcitationids=\"CR9 CR10 CR11\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], helps to detect fraud in real time with consensus process. The VotChain is an energy-latency tradeoff model, a backbone of its architecture, a balance and efficient approach to model performance as well as resource management. We made model validation, conducted on substantial real-world datasets including Ethereum ETL (100GB) and Google Financial X (50GB), unequivocally demonstrates VoteChain's superior performance [\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. The VoteChain algorithm achieves an impressive energy consumption of 20.0 mWh/tx, a 98 percent reduction as compared to existing PoW mechanisms. The novel approach, VoteChain consensus delivers a high throughput of 65 Tps and low latency of 1.2 ms, established consensus algorithms such as PoW, PoS, and Practical Byzantine Fault Tolerance (PBFT) significantly. In the study, we provide a scalable, environmentally conscious, and secure solution, paving the way for the next generation of decentralized financial applications [\u003cspan additionalcitationids=\"CR17 CR18 CR19\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. In the research, the principles and findings presented are further supported by open-source implementation and comprehensive reproducibility guidelines, fostering collaborative advancement in the field. Blockchain's decentralized ledger technology has revolutionized financial systems, yet PoW (e.g., Bitcoin) consumes\u0026thinsp;~\u0026thinsp;150 TWh annually, exceeding Norway\u0026rsquo;s energy usage. PoS, while more efficient, risks centralization. Recent studies highlight consensus latency as a bottleneck for real-time transactions[\u003cspan additionalcitationids=\"CR22 CR23 CR24 CR25 CR26 CR27 CR28\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eExisting consensus models fail to balance energy efficiency, security, and scalability (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Hybrid approaches often neglect computational overhead from deep learning integration. Key gaps include[\u003cspan additionalcitationids=\"CR31 CR32 CR33 CR34 CR35 CR36 CR37 CR38 CR39 CR40 CR41 CR42 CR43 CR44 CR45 CR46 CR47 CR48 CR49\" citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]:\u003c/p\u003e\n\u003cp\u003eExisting consensus models fail to balance energy efficiency, security, and scalability (Table 1). Hybrid approaches often neglect computational overhead from deep learning integration. Key gaps include[30-50]:\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eEnergy-intensive voting mechanisms\u0026nbsp;in consortium blockchains.\u003c/li\u003e\n \u003cli\u003eLack of\u0026nbsp;lightweight, real-time anomaly detection\u0026nbsp;for financial fraud.\u003c/li\u003e\n \u003cli\u003eLimited empirical validation on large-scale financial datasets.\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eTable 1: Comparative Analysis of Consensus Mechanisms\u003c/p\u003e\n\u003cdiv\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eAlgorithm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eEnergy (mWh/tx)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eLatency (ms)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eThroughput (Tps)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eFault Tolerance (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003ePoW\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e59,500\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e4.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e99.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003ePoS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e10,670\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e98.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003ePBFT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e150\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1,200\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1,000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e99.9\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eVoteChain\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e20.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e99.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\u003c/br\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eVoteChain: A voting-based consensus protocol with\u0026nbsp;dynamic node selection\u0026nbsp;and\u0026nbsp;non-cryptographic voting.\u003c/li\u003e\n \u003cli\u003eEnergy-Latency Model: Mathematically proven with coefficients derived from Ethereum\u0026rsquo;s transaction history.\u003c/li\u003e\n \u003cli\u003eCNN-LSTM Hybrid Model: Lightweight architecture (1.2M parameters) for fraud detection (F1-score: 98.7%).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"2. Related Work","content":"\u003cp\u003eThe landscape of blockchain consensus mechanisms is diverse, each designed to achieve agreement among distributed nodes while balancing security, decentralization, and performance [\u003cspan additionalcitationids=\"CR52 CR53 CR54\" citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e]. Despite ongoing advancements, no single solution has been able to meet all the key requirements\u0026mdash;energy efficiency, low latency, high throughput, and strong real-time fraud detection, especially in the context of high-frequency financial transactions. In this section, we take a closer look at current consensus algorithms and fraud detection methods, exploring what they do well, where they fall short, and the critical gaps that VoteChain is designed to fill.\u003c/p\u003e\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Consensus Mechanisms\u003c/h2\u003e\u003cp\u003eIn the distributed decentralized mechanisms, traditional consensus mechanisms established the backbone of various blockchain networks. Bitcoin\u0026rsquo;s mechanism, PoW, relies on computational puzzles, where miners expend significant energy to validate transactions and create new blocks [\u003cspan additionalcitationids=\"CR2 CR3 CR4 CR5 CR6 CR7 CR8 CR9 CR10 CR11 CR12 CR13 CR14 CR15 CR16 CR17 CR18 CR19 CR20 CR21 CR22 CR23 CR24 CR25 CR26 CR27 CR28 CR29\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e] in the blockchain network system. It offers strong security against Sybil attacks and a high degree of decentralization, but the PoW consensus consumes high energy in computations. Bitcoin's annual energy consumption exceeds 150 TWh, surpassing the electricity usage of many countries [\u003cspan additionalcitationids=\"CR21 CR22 CR23 CR24 CR25 CR26 CR27 CR28 CR29 CR30\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. The red energy footprint translates to high operational costs and environmental concerns, making it unsuitable for energy-efficient financial applications. Furthermore, the PoW mechanism suffers from low transaction throughput (e.g., Bitcoin's\u0026thinsp;~\u0026thinsp;7 transactions per second (Tps)) and high transaction finality latency [\u003cspan additionalcitationids=\"CR4 CR5 CR6\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. The concept of \"stale blocks\" due to propagation delays also impacts on its efficiency and security for PoW mechanism[\u003cspan additionalcitationids=\"CR8 CR9 CR10 CR11 CR12 CR13 CR14\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eProof-of-Stake (PoS) has emerged as a much greener alternative to the energy-intensive Proof-of-Work (PoW) model by selecting validators based on the amount of cryptocurrency they \u0026ldquo;stake\u0026rdquo; as collateral [\u003cspan additionalcitationids=\"CR4 CR5 CR6 CR7 CR8\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. This approach slashes energy consumption dramatically\u0026mdash;some estimates suggest by more than 99 percent compared to PoW [\u003cspan additionalcitationids=\"CR2 CR3 CR4 CR5 CR6 CR7\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Ethereum\u0026rsquo;s recent transition to PoS, observed as the \u0026ldquo;Merge,\u0026rdquo; is a landmark example toward sustainability factor [\u003cspan additionalcitationids=\"CR9 CR10 CR11 CR12 CR13 CR14\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. PoS improves throughput and lowers latency relative to PoW, but it still faces many challenges like potential centralization due to wealth concentration and latency issues that may not meet the demands of instantaneous financial transactions. PoW, or PoS networks, can still have finality delays that are less than ideal for high-frequency trading or point-of-sale payments.\u003c/p\u003e\u003cp\u003eIn the permissioned blockchain systems, Practical Byzantine Fault Tolerance (PBFT) and its variants are popular for their high throughput and low latency [\u003cspan additionalcitationids=\"CR10 CR11 CR12 CR13 CR14 CR15 CR16 CR17 CR18 CR19\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. The PBFT mechanism reaches consensus through multiple message exchanges among a fixed set of known validators, offering strong consistency and flexibility against Byzantine faults. It has some limitations; like its performance deteriorates quickly as the number of participants grows, limiting its scalability for large, public, and permissionless financial networks [\u003cspan additionalcitationids=\"CR51 CR52 CR53 CR54 CR55 CR56 CR57 CR58 CR59\" citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. Moreover, its reliance on a pre-selected validator set reduces decentralization, making it less suitable for open financial ecosystems.\u003c/p\u003e\u003cp\u003eThere are many other consensus models include, Delegated Proof-of-Stake (DPoS), which speeds up validation by entrusting a small group of elected delegates but further concentrates control and Directed Acyclic Graphs (DAGs). The DPoS boosts scalability and throughput through parallel transaction processing but introduces complexities in maintaining a global order security [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Some authors have presented innovative ideas like Proof of Team Sprint (PoTS) have also been proposed to enhance energy efficiency by enabling collaborative block generation [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eprovides a comparative overview of these prominent consensus mechanisms across key performance and operational metrics[\u003cspan additionalcitationids=\"CR21 CR22 CR23 CR24 CR25 CR26 CR27 CR28 CR29 CR30 CR31 CR32 CR33 CR34 CR35 CR36 CR37 CR38 CR39 CR40 CR41 CR42 CR43 CR44 CR45 CR46 CR47 CR48 CR49\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eConsensus Mechanism\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEnergy Efficiency\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eLatency (ms)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eThroughput (Tps)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eDecentralization\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eScalability\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eFault Tolerance\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eProof-of-Work (PoW)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eVery Low\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eHigh (1000s)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLow (7\u0026ndash;15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eHigh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eLow\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eHigh (51% attack)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eProof-of-Stake (PoS)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHigh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eModerate (100s)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eModerate (20\u0026ndash;100)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eModerate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eModerate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eModerate\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePBFT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHigh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eVery Low (1\u0026ndash;10)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eHigh (1000s)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eLow\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eLow\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eHigh (f\u0026thinsp;\u0026lt;\u0026thinsp;N/3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDPoS\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHigh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eLow (10s)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eHigh (1000s)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eModerate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eModerate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eModerate\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDAG-based\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHigh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eLow (10s)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVery High (1000s+)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eModerate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eHigh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eModerate\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eVoteChain\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eVery High\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003eVery Low (1.2)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003eHigh (65)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003eHigh\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003eHigh\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003eVery High (99.5%)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe some of basic key factors of baseline consensus mechanisms are:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003ePoW: Requires solving cryptographic puzzles, leading to high energy waste.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003ePoS: Vulnerable to \"nothing-at-stake\" attacks.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003ePBFT: Suffers from O(n\u0026sup2;) communication complexity.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eVoting-Based: Tendermint and Algorand improve scalability but lack integrated fraud detection.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003e2.2. Deep Learning in Blockchain\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eCNN-based models detect anomalies in Ethereum transactions (F1-score: 92%) but add latency.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eFederated learning reduces data centralization but increases energy use.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e"},{"header":"3. Methodology and design","content":"\u003cp\u003e\u003cstrong\u003e3. Proposed Consensus Algorithm\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.1. Architectural Design\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eVoteChain decouples transaction validation from consensus voting (Fig. 1). Nodes are dynamically selected based on stake and reputation scores.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.2. Process Workflow\u003c/strong\u003e\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eTransaction Hashing: SHA-3 for immutability.\u003c/li\u003e\n \u003cli\u003eAnomaly Detection: CNN-LSTM model flags suspicious transactions (precision: 99.1%).\u003c/li\u003e\n \u003cli\u003eVoting Phase: Nodes vote in parallel; malicious voters are penalized.\u003c/li\u003e\n \u003cli\u003eFinality: Threshold BLS signatures reduce communication rounds.\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u003cstrong\u003e3.3. Mathematical Model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea. Energy Consumption\u003c/strong\u003e:\u003c/p\u003e\n\u003cp\u003eE\u003csub\u003etx\u003c/sub\u003e= \u0026alpha;\u0026sdot;∣V∣+\u0026beta;\u0026sdot;T\u003csub\u003eproc\u003c/sub\u003e+ \u0026gamma;\u0026sdot;L\u003csub\u003ecomm\u003c/sub\u003e\u003cem\u003e\u0026nbsp; \u0026nbsp;\u003c/em\u003e\u0026nbsp; \u0026nbsp;( 1 )\u003c/p\u003e\n\u003cp\u003eCoefficients (\u0026alpha;=0.15\u003cem\u003e\u0026alpha;\u003c/em\u003e=0.15,\u0026nbsp;\u0026beta;=0.02\u003cem\u003e\u0026beta;\u003c/em\u003e=0.02,\u0026nbsp;\u0026gamma;=0.01\u003cem\u003e\u0026gamma;\u003c/em\u003e=0.01) derived via Ethereum regression (R\u0026sup2;=0.93).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb. Consensus Latency\u003c/strong\u003e:\u003c/p\u003e\n\u003cp\u003eD\u003csub\u003econsensus\u003c/sub\u003e = \u0026nbsp;\u0026delta;log (∣V∣) + ϵ\u0026sdot;D\u003csub\u003evote\u003c/sub\u003e\u003cem\u003e\u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e ( 2 )\u003c/p\u003e\n\u003cp\u003e\u0026delta;=0.8\u003cem\u003e\u0026delta;\u003c/em\u003e=0.8,\u0026nbsp;ϵ=0.05\u003cem\u003eϵ\u003c/em\u003e=0.05\u0026nbsp;optimized via gradient descent.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec. Security Analysis\u003c/strong\u003e:\u003c/p\u003e\n\u003cul type=\"disc\"\u003e\n \u003cli\u003eByzantine Resistance: Tolerates up to 33% malicious nodes.\u003c/li\u003e\n \u003cli\u003eSybil Attack Prevention: Reputation-based node selection.\u003c/li\u003e\n\u003c/ul\u003e"},{"header":"4. Experimental Setup \u0026 Dataset Integration","content":"\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e4.1. Open-Source Financial Transaction Datasets\u003c/h2\u003e\u003cp\u003eWe utilize Ethereum ETL and Google-sourced financial datasets shown in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e for real-world transaction validation.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eOpensource Dataset descriptions\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDataset\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSource\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eDescription\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEthereum ETL\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGitHub\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eHistorical financial transactions\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGoogle Financial X\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOpen Data\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eVerified financial transactions\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003e4.2. Software \u0026amp; Hardware Configuration\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eProgramming: Python, TensorFlow, Solidity (for smart contracts).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eComputing: Intel Xeon 64-core processor, 128GB RAM.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"5. Training, Testing \u0026 Evaluation","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e5.1. Training Deep Lightweight Module\u003c/h2\u003e\u003cp\u003eA CNN-based anomaly detection model is trained using blockchain transactions, identifying fraudulent voting attempts.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e5.2. Evaluation Metrics\u003c/h2\u003e\u003cp\u003eWe have considered the following metrics to evaluate the algorithms:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eEnergy per transaction (mWh/tx)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eThroughput (transactions per second)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eConsensus latency (ms)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eFault tolerance (%)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eAccuracy of anomaly detection (F1-score)\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"6. Experimental Setup and Dataset Integration","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e6.1. Open-Source Financial Transaction Datasets\u003c/h2\u003e\u003cp\u003eTo ensure statistical validity and real-world generalizability, two large-scale open datasets\u0026mdash;Ethereum ETL and Google Financial X\u0026mdash;were integrated (see Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Ethereum ETL provides over 100GB of blockchain-based transaction records, while Google Financial X contributes verified financial transaction logs with a rare fraud incidence (0.2%). The preprocessing pipeline involved cleansing via Pandas and Apache Spark, followed by tokenization and temporal encoding. SMOTE was applied to balance the 99.8:0.2 class skew, crucial for fraud detection efficacy. This integration facilitated robust training for deep models without overfitting to rare fraud cases[\u003cspan additionalcitationids=\"CR56 CR57 CR58 CR59\" citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]\u003c/p\u003e\u003c/div\u003e"},{"header":"7. Results","content":"\u003cp\u003e\u003cstrong\u003e7. 1 Benchmarking\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea. Preprocessing Pipeline\u003c/strong\u003e:\u003c/p\u003e\n\u003cul type=\"disc\"\u003e\n \u003cli\u003eData cleansing with Pandas and Spark\u003c/li\u003e\n \u003cli\u003eTransactional tokenization\u003c/li\u003e\n \u003cli\u003eFeature standardization and time-window encoding\u003c/li\u003e\n \u003cli\u003eSMOTE (Synthetic Minority Oversampling Technique) applied to address class imbalance in fraudulent vs. legitimate transactions, shown in Figure 3-4\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003e7.2. Software and Hardware Configuration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe experiment utilized a heterogeneous, hybrid setup to validate system-wide scalability, incorporating both local and cloud infrastructures. All models and simulations were executed using Python 3.10, TensorFlow 2.8, and integrated blockchain stacks like Truffle and Ganache. \u0026nbsp;Visual analytics tools such as SHAP and Grad-CAM ensured explainability. Such a versatile configuration is aligned with best practices for hybrid blockchain-AI systems[40-50].\u003c/p\u003e\n\u003col style=\"list-style-type: upper-roman;\"\u003e\n \u003cli\u003eProgramming Tools: Python 3.10, TensorFlow 2.8, Keras, Scikit-learn, XGBoost, SHAP\u003c/li\u003e\n \u003cli\u003eHardware Infrastructure:\u003c/li\u003e\n\u003c/ol\u003e\n\u003cul\u003e\n \u003cli\u003eLocal: Intel Xeon Gold 64-core, 128 GB RAM\u003c/li\u003e\n \u003cli\u003eCloud: AWS EC2 c5.18xlarge (72 vCPUs, 144GB RAM)\u003c/li\u003e\n\u003c/ul\u003e"},{"header":"8. Training, Testing \u0026 Evaluation","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e8.1. Lightweight CNN-LSTM Model Training\u003c/h2\u003e\u003cp\u003eThe hybrid CNN-LSTM model exploits spatial and temporal transaction dependencies.\u003c/p\u003e\u003cp\u003eIts architecture (see Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003e) integrates convolutional layers for spatial encoding and LSTM for sequential feature tracking. The inclusion of attention further refines fraud detection precision. Over 50 epochs, training stabilized at 97.8 percent accuracy. This model is optimized for low-latency, energy-efficient blockchain participation .\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe proposed VoteChain custom build architecture integrates a CNN-LSTM-based anomaly detection engine that leverages temporal patterns in transaction flows. This hybrid deep learning model is optimized for fraud detection and efficient blockchain participation:\u003c/p\u003e\u003cp\u003eDetails of VoteChain Model layered Architecture (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e6\u003c/span\u003e):\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eInput Layer: Time-series vectors of normalized transaction features (size: 32x32)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eConv Layer 1: 32 filters, 3x3 kernel, ReLU\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eConv Layer 2: 64 filters, 3x3, BatchNorm\u0026thinsp;+\u0026thinsp;MaxPooling\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eConv Layer 3: 128 filters, Dropout (0.3)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eLSTM Layer 1: 128 units\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eLSTM Layer 2: 64 units, Attention\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eDense Output: Sigmoid for binary classification (fraud vs. normal)\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThe model training conducted over 10 epochs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e7\u003c/span\u003e with early stopping (patience\u0026thinsp;=\u0026thinsp;7) and Adam optimizer (LR\u0026thinsp;=\u0026thinsp;1e-4). Accuracy converged at ~\u0026thinsp;97.8% with validation loss at ~\u0026thinsp;0.05.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e8.2. Evaluation Metrics\u003c/h2\u003e\u003cp\u003eEvaluation encompasses energy efficiency, detection quality, and consensus robustness metrics.\u003c/p\u003e\u003cp\u003eIncorporating metrics like energy/transaction and consensus latency enables holistic system profiling. Accuracy-related scores (Precision, Recall, F1) ensure detection validity. Fault tolerance was tested under network splits and Byzantine node injections[\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. These metrics form a reproducible benchmarking suite for future DeFi-integrated AI systems.\u003c/p\u003e\u003cp\u003eA multifaceted metric framework ensures the credibility of both blockchain consensus and AI-based anomaly detection:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eEnergy per transaction (mWh/tx)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eThroughput (Transactions/sec)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eConsensus Latency (milliseconds)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eFault Tolerance (% uptime under adversarial/split conditions)\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eDetection Accuracy: F1-Score, Precision, Recall\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eVoteChain model - evaluation metrics summery\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMetric\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eClass 0 9Legit)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eClass 1 (Fraud)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMacro Avg.\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eWt. Avg.\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrecision\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e98.94%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e94.69%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e96.82%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e96.84%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRecall\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e94.60%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e98.96%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e96.78%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e96.75%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eF1-Score\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e96.72%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e96.78%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e96.75%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e96.75%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSupport\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2964\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2883\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026ndash;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5847\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAccuracy\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026ndash;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026ndash;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026ndash;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e96.75%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003e8.3. Benchmarking Results\u003c/h2\u003e\u003cp\u003eVoteChain outperforms traditional models on energy, latency, and fault tolerance.\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e4\u003c/span\u003e illustrates comparative performance across multiple consensus protocols. VoteChain achieves 250x lower energy use than PoS and near-instantaneous latency (1.2ms). Such performance validates the architectural fusion of anomaly detection with voting-based blockchain consensus mechanisms.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003esummery of evaluation metrics [\u003cspan additionalcitationids=\"CR51 CR52 CR53 CR54 CR55 CR56 CR57 CR58 CR59\" citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAlgorithm\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEnergy (mWh/tx)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eThroughput (Tps)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLatency (ms)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eFault Tolerance (%)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVoteChain (Ours)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e20.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e99.5\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eProof-of-Work (PoW)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5000\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e59,500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e99.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eProof-of-Stake (PoS)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e300\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e10,670\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e98.5\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePBFT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e150\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1000\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1200\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e99.9\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThis graph illustrates a significant energy-efficiency leap in VoteChain, consuming 250x less energy than PoS and 99.6% less than PoW.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003e8.4. Component-Wise Performance Analysis\u003c/h2\u003e\u003cp\u003eThe ablation studies quantify the role of each VoteChain component,r emoving anomaly detection led to 12.2% fraud reintroduction, confirming its necessity. Eliminating dynamic node selection decreased fault tolerance, underlining its operational synergy with detection modules. These tests validate each architectural layer's contribution to overall system integrity[\u003cspan additionalcitationids=\"CR56 CR57 CR58 CR59\" citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eTwo critical ablation variants were tested:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eWithout anomaly detection: Fraud rate increased from 0.19\u0026ndash;12.2%\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eWithout dynamic node selection: Fault tolerance reduced from 99.5\u0026ndash;95.3%\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThis validates the synergistic role of deep learning and node adaptability in the VoteChain framework.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003e9.5. Explainability via SHAP\u003c/h2\u003e\u003cp\u003eExplainable AI enhances model transparency for high-stakes financial systems.\u003c/p\u003e\u003cp\u003eSHAP scores identify dominant features (e.g., time-gap, gas-fee anomalies) shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e13\u003c/span\u003e. These interpretable outputs are vital for regulatory compliance and post-hoc forensic analysis.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eInterpretability in real-world deployment\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe proposed model deployment testing has been done through following steps:\u003c/p\u003e\u003cp\u003ea. SHAP (DeepExplainer) was applied to CNN-LSTM outputs:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eKey transactional features like time gaps, gas fees, and sender-receiver history were top fraud predictors.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eLocal explanations provided per transaction enable forensic traceability.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eIt has been shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e13\u003c/span\u003e SHAP Summary Plot, it shows feature importance for fraud detection across samples.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003e10.6. Statistical Validation\u003c/h2\u003e\u003cp\u003eStatistical validation underpins model reliability across varying data partitions.\u003c/p\u003e\u003cp\u003eWelch\u0026rsquo;s t-test confirmed performance gaps were significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), establishing VoteChain\u0026rsquo;s consistent edge. Additionally, 10-fold cross-validation yielded accuracy fluctuations within \u0026plusmn;\u0026thinsp;1.8%, confirming strong generalization potential across unseen blockchain transaction windows.\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eWelch\u0026rsquo;s t-test (two-tailed, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01) confirms the performance differences (F1-score, latency) between VoteChain and traditional consensus models are statistically significant.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e10-fold cross-validation shows\u0026thinsp;\u0026plusmn;\u0026thinsp;1.8% deviation in accuracy, indicating generalizability.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe VoteChain framework introduces a synergistic fusion of deep learning-based fraud detection and energy-efficient voting-based consensus for blockchain financial transactions. Despite its strong empirical performance, certain limitations were identified. The system's scalability currently caps at 500 active nodes. Moreover, its latency sensitivity under artificial network delays (\u0026gt;\u0026thinsp;600ms) causes a marginal performance drop (\u0026sim;3%) in CNN-LSTM precision, indicating the potential value of incorporating GRU-based or attention-driven temporal models for future iterations.\u003c/p\u003e\u003cp\u003eOn the positive spectrum, the broader implications of VoteChain are both impactful and forward-looking. Environmentally, it achieves a 98% CO₂ reduction compared to traditional Proof-of-Work (PoW) systems, making it highly viable for carbon-sensitive financial markets. With its lightning-fast architecture delivering latency under 2 milliseconds, VoteChain easily handles microtransactions\u0026mdash;perfect for real-time demands in DeFi, IoT, and digital payments. Beyond finance, its flexible modular design allows it to adapt to other critical areas. By applying the right input transformations, VoteChain can be extended to secure healthcare records or manage blockchain-based supply chain audits. These results highlight VoteChain\u0026rsquo;s promise not just as a scalable financial solution, but as a versatile, energy-efficient platform ready to power AI-driven blockchain applications across a wide range of industries.\u003c/p\u003e\u003cp\u003e\u003cb\u003e11.1. Limitations\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eScalability Bound: The system was validated up to 500 active nodes.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eLatency Sensitivity: Under artificial delay injections (\u0026gt;\u0026thinsp;600ms), the CNN-LSTM precision dropped by 3%. In future variants may require GRU-LSTM hybridization or temporal attention.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\n\u003ch3\u003e11.2. Broader Impact\u003c/h3\u003e\n\u003cp\u003eThe proposed model VoteChain\u0026rsquo;s novel integration of AI-driven fraud detection with voting-based consensus demonstrates observed as:\u003c/p\u003e\u003cp\u003ea. Environmental Gains -The proposed model has gained 98% CO₂ emission reduction compared to PoW. The proposed model is an ideal strategy for carbon-regulated financial markets.\u003c/p\u003e\u003cp\u003eb. Microtransaction Support -The transactional latency is less than 2ms, it allows sub-second confirmations ideal for retail payments, IoT finance, and DeFi services.\u003c/p\u003e\u003cp\u003ec. Cross-domain Utility -The proposed model architecture is modular and can be deployed in healthcare or supply chain blockchain systems after fine-tuning input features.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study introduced VoteChain, a hybrid framework that combines CNN-LSTM-based anomaly detection with a voting-based consensus mechanism, tailored for energy-efficient, real-time financial transaction validation. By integrating deep learning techniques with blockchain consensus logic, the framework achieved high fraud detection accuracy (~\u0026thinsp;97.8%), ultra-low latency (\u0026lt;\u0026thinsp;2ms), and substantial reductions in energy consumption (250x lower than PoS and 99.6% lower than PoW). The experimental results, supported by rigorous statistical validation, demonstrated VoteChain\u0026rsquo;s superiority across critical performance metrics such as throughput, energy per transaction, and fault tolerance. Furthermore, explainability techniques (SHAP and Grad-CAM) provided transparency, reinforcing trust in fraud detection decisions.\u003c/p\u003e"},{"header":"Future Work","content":"\u003cp\u003eSeveral avenues are proposed to enhance and extend the VoteChain framework. Future work includes model optimization through the integration of GRU-LSTM hybrids or transformer-based architectures to improve temporal sensitivity, particularly under high-latency conditions. Scalability can be addressed by incorporating lightweight containerization solutions for efficient deployment in distributed environments. The framework also shows strong potential for cross-industry applications, including fraud detection in healthcare billing, supply chain financing, and insurance claims. To ensure long-term security, quantum-resilient cryptographic layers will be explored to safeguard against post-quantum threats. Furthermore, regulatory integration will be prioritized by aligning VoteChain with international compliance standards such as GDPR and PSD2, facilitating its adoption in institutional and financial sectors.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eCompeting Interests / Conflict of Interest:\u003c/strong\u003e The authors declare that there are no conflicts of interest regarding the publication of this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical Approval and Consent to Participate:\u003c/strong\u003e Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical Trial Registration:\u003c/strong\u003e Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to Publish Declaration:\u003c/strong\u003e Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to Participate Declaration:\u003c/strong\u003e Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Declaration:\u003c/strong\u003e Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of Data and Materials:\u003c/strong\u003e The datasets used in this study are publicly available open-source datasets. No proprietary or restricted datasets were used. The processed data and analysis results presented in this study are available upon reasonable request from the corresponding author. Raw datasets were accessed from publicly available blockchain benchmarking and consensus evaluation sources.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding Declaration:\u003c/strong\u003e This research received no external funding. However, one of the co-authors is affiliated with the Australian National University (ANU), which is a member of Springer Nature\u0026rsquo;s institutional open-access program. Therefore, the article processing charges (APC) are expected to be fully covered (100%) under ANU\u0026rsquo;s institutional agreement.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eNakamoto, S.: Bitcoin: A Peer-to-Peer Electronic Cash System, 2008. [Online]. Available: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://bitcoin.org/bitcoin.pdf\u003c/span\u003e\u003cspan address=\"https://bitcoin.org/bitcoin.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eButerin, V.: A Next-Generation Smart Contract and Decentralized Application Platform. Ethereum White Paper, (2014)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDwork, C., Naor, M.: Pricing via Processing or Combatting Junk Mail. in In: Advances in Cryptology \u0026mdash; CRYPTO\u0026rsquo; 92, vol. 740, pp. 139\u0026ndash;147. Springer (1993)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCastro, M., Liskov, B.: Practical Byzantine Fault Tolerance, in \u003cem\u003eOSDI\u003c/em\u003e, pp. 173\u0026ndash;186. (1999)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBenet, J., Addressed, I.P.F.S.C., Versioned, P.P., File System: [Online]. Available: https://ipfs.tech (2014)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBano, S., Sonnino, A., Al-Bassam, M., Azouvi, S., McCorry, P., Meiklejohn, S., Danezis, G.: SoK: Consensus in the Age of Blockchains, in \u003cem\u003eProceedings of the 1st ACM Conference on Advances in Financial Technologies\u003c/em\u003e, (2019)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZheng, Z., Xie, S., Dai, H., Chen, X., Wang, H.: An Overview of Blockchain Technology: Architecture, Consensus, and Future Trends, in \u003cem\u003eProceedings of IEEE BigData Congress\u003c/em\u003e, (2017)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKiayias, A., Russell, A., David, B., Oliynykov, R.: Ouroboros: A Provably Secure Proof-of-Stake Blockchain Protocol, in \u003cem\u003eAnnual International Cryptology Conference\u003c/em\u003e, Springer, (2017)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMettler, M.: Blockchain Technology in Healthcare: The Revolution Starts Here, in \u003cem\u003eIEEE 18th International Conference on e-Health Networking\u003c/em\u003e, pp. 1\u0026ndash;3. (2016)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChaum, D., Mail, U.E., Addresses, R., Pseudonyms, D.: Commun. ACM. \u003cb\u003e24\u003c/b\u003e(2), 84\u0026ndash;88 (1981)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcCorry, K., Shahandashti, S.F., Hao, F.: A Smart Contract for Boardroom Voting with Maximum Voter Privacy. in In: Financial Cryptography and Data Security. Springer (2017)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eH\u0026ouml;lbl, M., Welzer, T., Kos, A., Kamišalić, M.: A Systematic Review of the Use of Blockchain in Healthcare. Symmetry, \u003cb\u003e10\u003c/b\u003e, 10, (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNeudecker, T., Hartenstein, H.: Network Layer Aspects of Permissionless Blockchains. IEEE Commun. Surv. Tutorials. \u003cb\u003e21\u003c/b\u003e(1), 838\u0026ndash;857 (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcCorry, P., Shahandashti, S.F., Hao, F.: A Smart Contract for Boardroom Voting with Maximum Voter Privacy, in \u003cem\u003eFC 2017\u003c/em\u003e, Springer\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKalodner, H., Goldfeder, S., Chator, A., M\u0026ouml;ser, M., Narayanan, A.: BlockSci: Design and applications of a blockchain analysis platform, in \u003cem\u003eUSENIX Security Symposium\u003c/em\u003e, (2017)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRouhani, S., Deters, R.: Security, Performance, and Applications of Smart Contracts: A Systematic Survey, in \u003cem\u003eIEEE Access\u003c/em\u003e, vol. 7, pp. 50759\u0026ndash;50779, (2019)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGervais, A., et al.: On the Security and Performance of Proof of Work Blockchains, in \u003cem\u003eACM CCS\u003c/em\u003e, (2016)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eConti, M., Lal, C., Ruj, S.: A Survey on Security and Privacy Issues of Bitcoin. IEEE Commun. Surv. Tutorials, \u003cb\u003e20\u003c/b\u003e, 4, (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen, J., Del Molino, A., Hicks, B., Perrig, A.: Securing the Internet of Things with Decentralized Trust. in IEEE Secur. Priv. Workshops, (2015)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePopov, S.: The Tangle. IOTA Whitepaper, (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKumar, D.D., Chandini, D.V., Reddy, D.: Secure Electronic Voting System Using Blockchain Technology. Int. J. Smart Home. \u003cb\u003e14\u003c/b\u003e(2), 31\u0026ndash;38 (2020)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGurubasavanna, M.G., Shariff, S.U., Mamatha, R., Sathisha, N.: Multimode Authentication Based Electronic Voting Kiosk Using Raspberry Pi, in \u003cem\u003eProc. 2nd Int. Conf. I-SMAC (IoT in Social, Mobile, Analytics and Cloud)\u003c/em\u003e, Palladam, India, IEEE, pp. 1\u0026ndash;8. (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGiraldo, F.D., Barbosa, M.C., Carlos, Y., Gamboa, E.: Electronic Voting Using Blockchain and Smart Contracts: Proof of Concept. Latin Am. Transact. \u003cb\u003e18\u003c/b\u003e(10), 1\u0026ndash;9 (2020)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePawlak, M., Maranda, A.P., Kryvinska, N.: Towards the Intelligent Agents for Blockchain e-Voting System, in \u003cem\u003eProc. 9th Int. Conf. Emerging Ubiquitous Systems and Pervasive Networks (EUSPN)\u003c/em\u003e, Leuven, Belgium, Elsevier, pp. 239\u0026ndash;246. (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYi, H.: Securing e-Voting Based on Blockchain in P2P Network. EURASIP J. Wirel. Commun. Netw. vol. no. \u003cb\u003e137\u003c/b\u003e, 1\u0026ndash;9 (2019)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRomero, M.T., Orden, D., Maestre, I.M., Sanchez, J.J., Guzman, J.M.G.: Distributed Remote E-Voting System Based on Shamir\u0026rsquo;s Secret Sharing Scheme. Electronics. no. \u003cb\u003e10\u003c/b\u003e(3075), 1\u0026ndash;19 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTas, R., Tanriover, O.O.: A Systematic Review of Challenges and Opportunities of Blockchain for E-Voting. Symmetry. no. \u003cb\u003e12\u003c/b\u003e(1328), 1\u0026ndash;24 (2020)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLarriba, A.M., Cuco, A.C.I., Sempere, J.M., Lopez, D.: Distributed Trust, a Blockchain Election Scheme. Informatica. no. \u003cb\u003e32\u003c/b\u003e(2), 1\u0026ndash;35 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, C., Xiao, J., Dai, X., Jin, H.: AMVchain: Authority Management Mechanism on Blockchain-Based Voting Systems. Peer-to-Peer Netw. Appl. no. \u003cb\u003e14\u003c/b\u003e, 2801\u0026ndash;2812 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRehman, E., Khan, M.A., Soomro, T.R., Taleb, N., Afifi, M.A., Ghazal, T.M.: Using Blockchain to Ensure Trust Between Donor Agencies and NGOs in Under-Developed Countries. Computers. no. \u003cb\u003e10\u003c/b\u003e(98), 1\u0026ndash;20 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGonzalez, C.D., Mena, D.F., Munoz, A.M., Rojas, O., Gomez, G.S.: Electronic Voting System Using an Enterprise Blockchain. Appl. Sci. no. \u003cb\u003e12\u003c/b\u003e(531), 1\u0026ndash;14 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLykidis, I., Drosatos, G., Rantos, K.: The Use of Blockchain Technology in e-Government Services. Computers. no. \u003cb\u003e10\u003c/b\u003e(168), 1\u0026ndash;17 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAlshehri, A., et al.: Privacy-Preserving E-Voting System Supporting Score Voting Using Blockchain. Appl. Sci. no. \u003cb\u003e13\u003c/b\u003e(1096), 1\u0026ndash;18 (2023)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRamya, P., Jashwanth, T., Sathvik, D.V.: A Hybrid Proof of Stake Trust Block Chain Model in Pervasive Social Networking for E-Voting System. Int. J. Intell. Eng. Syst. no. \u003cb\u003e15\u003c/b\u003e(4), 1\u0026ndash;10 (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eUmar, B.U., Olaniyi, O.M., Olajide, D.O., Dogo, E.M.: Paillier Cryptosystem Based ChainNode for Secure Electronic Voting. Front. Blockchain. no. \u003cb\u003e5\u003c/b\u003e, 1\u0026ndash;11 (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVemula, S., Kovvur, R.M.R., Marneni, D.: Secure E-Voting System Implementation Using CryptDB. SN Comput. Sci. no. \u003cb\u003e2\u003c/b\u003e(3), 1\u0026ndash;6 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePathak, A., et al.: Design and Implementation of a Secure and Robust Voting System Based on Blockchain. Int. J. Adv. Res. Ideas Innov. Technol. no. \u003cb\u003e4\u003c/b\u003e(5), 1\u0026ndash;7 (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCao, H.J., et al.: A New Proxy Electronic Voting Scheme Achieved by Six-Particle Entangled States. Int. J. Theor. Phys. no. \u003cb\u003e57\u003c/b\u003e, 674\u0026ndash;681 (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGupta, S., Gupta, K.K., Shukla, P.K.: BSeVS\u0026mdash;A Smart e-Voting System Based on Blockchain Technology, in \u003cem\u003eProc. 14th ICCCNT Conf.\u003c/em\u003e, Delhi, India, IEEE, pp. 1\u0026ndash;7. (2023)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGupta, S., Gupta, K.K., Shukla, P.K.: BVKCS\u0026mdash;A Blockchain Based e-Voting System With Key Covering Scheme, in \u003cem\u003eProc. ICTBIG Conf.\u003c/em\u003e, Indore, India, IEEE, pp. 1\u0026ndash;7. (2023)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYajam, H., et al.: Improvement on Bitcoin\u0026rsquo;s Verifiable Public Randomness With Semi Trusted Delegates, in \u003cem\u003eProc. 9th IST Symp.\u003c/em\u003e, Tehran, Iran, IEEE, pp. 1\u0026ndash;7. (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFu, J., et al.: A Study on the Optimization of Blockchain Hashing Algorithm Based on PRCA. Secur. Commun. Netw., vol. pp. 1\u0026ndash;12. (2020)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, B., et al.: Large-Scale Election Based on Blockchain, in \u003cem\u003eProc. Int. Conf. Identification, Information and Knowledge in the Internet of Things\u003c/em\u003e, vol. 129, pp. 234\u0026ndash;237, (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKhan, K.M., Arshad, J., Khan, M.M.: Secure Digital Voting System Based on Blockchain Technology. Int. J. Electron. Gov. Res. no. \u003cb\u003e14\u003c/b\u003e(1), 1\u0026ndash;10 (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKshetri, N., Vows, J., Blockchain-Enabled, E.-V.: IEEE Softw. no. \u003cb\u003e35\u003c/b\u003e(4), 95\u0026ndash;99 (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFerrer, E.C.: The Blockchain: A New Framework for Robotic Swarm Systems, in \u003cem\u003eProc. Future Technologies Conf. (FTC) 2018\u003c/em\u003e, vol. 881, Cham: Springer, pp. 1\u0026ndash;12. (2019)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSolankar, A., Javheri, S.: Secure E-Voting System Using Visual Cryptography \u0026amp; Block Chain Ledger. Turk. J. Comput. Math. Educ. no. \u003cb\u003e12\u003c/b\u003e(1S), 7\u0026ndash;12 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBao, Z., Wang, B., Shi, W., Privacy-Preserving, A.: Decentralized and Functional Bitcoin E-Voting Protocol, in \u003cem\u003eProc. IEEE SmartWorld\u003c/em\u003e (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e/SCALCOM/UIC/ATC/CBDCom/IOP/SCI\u003c/span\u003e\u003cspan address=\"http:///SCALCOM/UIC/ATC/CBDCom/IOP/SCI\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, Guangzhou, China, pp. 1\u0026ndash;6\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbuidris, Y., et al.: Secure Large-Scale E-Voting System Based on Blockchain Contract Using a Hybrid Consensus Model Combined With Sharding. ETRI J. no. \u003cb\u003e43\u003c/b\u003e, 1\u0026ndash;14 (2020)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTas, R., Tanriover, O.O.: A Manipulation Prevention Model for Blockchain-Based E-Voting Systems. Secur. Commun. Netw., vol. pp. 1\u0026ndash;16. (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAhn, B.: Implementation and Early Adoption of an Ethereum-Based Electronic Voting System for the Prevention of Fraudulent Voting. Sustainability. no. \u003cb\u003e14\u003c/b\u003e(2917), 1\u0026ndash;16 (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNeziri, V., et al.: Assuring Anonymity and Privacy in Electronic Voting With Distributed Technologies Based on Blockchain. Appl. Sci. no. \u003cb\u003e12\u003c/b\u003e(5477), 1\u0026ndash;12 (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGupta, S., Gupta, K.K., Shukla, P.K.: Blockchain Based Voting System Powered by Post Quantum Cryptography (BBVSP-PQC), in \u003cem\u003eProc. ICPC2T Conf.\u003c/em\u003e, Raipur, India, IEEE, pp. 1\u0026ndash;8. (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChoi, S., Kang, J., Chung, K.S.: Design of Blockchain Based e-Voting System for Vote Requirements, \u003cem\u003eJ. Phys.\u003c/em\u003e, vol. pp. 1\u0026ndash;7, 2021. (1944)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAhubele, B.O., Oghenekaro, L.U.: Secured Electronic Voting System Using RSA Key Encapsulation Mechanism. Eur. J. Electr. Eng. Comput. Sci. no. \u003cb\u003e6\u003c/b\u003e, 1\u0026ndash;7 (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eToma, C., et al.: Secure and Anonymous Voting D-App With IoT Embedded Device Using Blockchain Technology. Electronics. no. \u003cb\u003e11\u003c/b\u003e(1895), 1\u0026ndash;25 (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSrikrishnaswetha, K., Kumar, S., Mahmood, M.R.: A Study on Smart Electronics Voting Machine Using Face Recognition and Aadhar Verification With IoT. Innov. Electron. Commun. Eng. no. \u003cb\u003e65\u003c/b\u003e, 87\u0026ndash;95 (2019)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRathee, G., et al.: On the Design and Implementation of a Blockchain Enabled E-Voting Application Within IoT-Oriented Smart Cities. IEEE Access. no. \u003cb\u003e9\u003c/b\u003e, 1\u0026ndash;12 (2021). Special Section Blockchain Technology\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQu, W., et al.: A Electronic Voting Protocol Based on Blockchain and Homomorphic Signcryption. Concurrency Comput., \u003cb\u003e34\u003c/b\u003e, 3, e5817, (2020)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChaieb, M., et al.: Design and Practical Implementation of Verify-Your-Vote Protocol. Concurrency Comput., \u003cb\u003e33\u003c/b\u003e, 1, e5813, (2020)\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"cluster-computing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Cluster Computing](https://www.springer.com/journal/10586)","snPcode":"10586","submissionUrl":"https://submission.nature.com/new-submission/10586/3","title":"Cluster Computing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Blockchain, Consensus Algorithm, Energy Efficiency, VoteChain, Deep Learning, Fraud Detection, Lightweight, Financial Transactions, Decentralized Finance (DeFi), Latency Optimization, Throughput, Fault Tolerance","lastPublishedDoi":"10.21203/rs.3.rs-7012159/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7012159/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eBlockchain consensus mechanisms like Proof-of-Work (PoW) and Proof-of-Stake (PoS) face critical challenges in energy efficiency and latency for financial transactions. In the study we devises VoteChain, a novel voting-based consensus algorithm enhanced by deep lightweight anomaly detection. We derive a rigorous energy-latency tradeoff model and integrate a Convolutional Neural Network (CNN) for real-time fraud detection. In the study we use opensource datasets and tested that VoteChain reduces energy consumption to 20.0 mWh/tx (98% lower than PoW) while achieving 65 Tps throughput and 1.2 ms latency, outperforming PoW, PoS, and PBFT. Whereas fault tolerance reaches 99.5% and validated via Byzantine attack simulations. In the study, we explored a scalable, eco-friendly solution for decentralized financial transactions.\u003c/p\u003e","manuscriptTitle":"A Deep Approach to an Energy-Efficient Voting-Based Consensus Algorithm for Secure Financial Transactions","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-10 17:02:04","doi":"10.21203/rs.3.rs-7012159/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-13T03:45:20+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-29T06:58:36+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"43327396670781547425358812805856002485","date":"2026-01-28T14:33:20+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-13T12:54:55+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"227011303019289655561162374880514323668","date":"2025-08-11T09:02:56+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-09T06:45:42+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-01T06:41:58+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-01T01:08:42+00:00","index":"","fulltext":""},{"type":"submitted","content":"Cluster Computing","date":"2025-06-30T15:17:32+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"cluster-computing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Cluster Computing](https://www.springer.com/journal/10586)","snPcode":"10586","submissionUrl":"https://submission.nature.com/new-submission/10586/3","title":"Cluster Computing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"06d5dd86-b5c9-4cb8-85c4-e555e0733962","owner":[],"postedDate":"July 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-12T06:41:06+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-10 17:02:04","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7012159","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7012159","identity":"rs-7012159","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}