Table 1 Summary of key findings and techniques of studies on serverless security threat detection.

¹¹

BC, AI, LSTM, Smart Contracts

Monitoring Accuracy, Response Speed, Data Integrity, Anomaly Detection Rate

The system detects GNSS signal anomalies and improves monitoring accuracy and response time.

¹²

BC, Smart Contracts, DL-based IDS, Proof of Voting Consensus, IPS/IDS Modules

Accuracy, Precision, Recall, Sensitivity, Specificity, FPR, F1-Score, MCC

The framework mitigates detection/validation time and boosts attack prevention, making it ideal for real-time use.

¹³

PBAC, BC, Smart Contracts, IPFS

Access Control Efficiency, EHR Management Scalability, Smart Contract Execution Time

The proposed technique enhances EHR management efficiency and scalability.

¹⁴

SMOTE, ENN, RF, XGB, BA, Random and Bayesian Search

Accuracy, Error Rate, Computational Costs, Parameter Search Time

The BA enhanced performance, attaining higher accuracy and a lower error rate in EDoS attack detection.

¹⁵

AI, HE, SIMD, DNN, FedAvg Algorithm, Privacy-preserving Distributed Learning

Detection Accuracy, Training Time, Real-world Adaptability

The framework maintains accuracy, mitigates training time, and adapts to real-world systems.

¹⁶

Ensemble Learning, WCA, DO, Threat Detection

Accuracy, Resilience Against Threats, Real-time Detection

Improved threat detection and system adaptability utilizing ensemble learning and WCA.

¹⁷

EfficientNet, Inception-ResNet-v2, Ensemble DL, BC, Continuous network monitoring, Anomaly detection algorithms

Classification Accuracy, AUC, Training time, Balanced Accuracy Recall Weighted Score (BPRWS) Metric

EffiIncepNet achieved up to 98% accuracy with enhanced scalability and security.

¹⁸

CNN, AdaHessian Optimization, Model Pruning, Post-training Quantization

Accuracy, Recall, Precision, F1-Score

Attained the highest output with low resource usage and high cryptojacking detection.

¹⁹

FODWNN-DoWAD, PBSO, DWNN, HLCCO

Accuracy, Recall, Precision, F1-Score, MCC

Highlighted more significant results in detecting DoW attacks utilizing optimized neural networks.

²⁰

Cryptocurrency market data and Event study analysis, Attack type evaluation

Market Returns, Cryptocurrency Price Stability

Diverse attacks significantly affect cryptocurrency returns and stability.

²¹

FaaSMT, Parallel Processing, MTA, HO

Attack Detection, Function Monitoring, Performance Overhead

FaaSMT effectively detects attacks while mitigating performance overhead.

²²

Dataset Creation, DoW Attack Detection, Containerized Applications, Threat Understanding

Attack Detection, Dataset Utilization, Security Enhancement

The dataset assists in developing stronger models for detecting DoW attacks in serverless environments.

²³

AI, BC, and Compliance Measures

Data Protection, Threat Detection, Security Integrity

Encryption, AI, and BC improve cloud security and ensure compliance.

²⁴

Synthetic Data Generation, ML

Dataset Generation, Anomaly Detection

Created a synthetic dataset for detecting DoW attacks in serverless environments using ML.

²⁵

Bi-3DQRNN, SASOS, EAGTO

Accuracy

The model attained higher accuracy than existing fraud detection techniques in mobile transactions.

²⁶

Warmonger Attack, Egress IP Analysis

Egress IP Usage Patterns, Number of Egress IPs per Serverless Service Providers (SSP)

Small egress IP sets in serverless platforms allow malevolent users to trigger IP blocking, causing DoS.

²⁷

BC, NFTs, Threshold Cryptography, FHE, IPFS, Smart Contracts, Composable NFTs

Privacy Preservation, Data Traceability, Cost Evaluation, Feasibility and User-friendliness

The model ensures efficient, private, traceable genomic data management and monetization.

²⁸

Native Monitoring Tools

Attack Detection Rate, False Alarm Rate

The methodology detects overall simulated attacks with minimal false alarms.