Figure 1.
Figure 2.
Figure 3.
Datasets for Evaluating Privacy-Preserving Data Aggregation Techniques in WBANs_ All Datasets were down sampled to 1Hz temporal resolution and normalized across channels prior to integration into the Ns-3 Traffic Generation model
| Dataset Name | Source | Data Contents | Size | Use Case | Accessibility |
|---|---|---|---|---|---|
| MIMIC-III | Beth Israel Deaconess MC | Patient vitals, lab tests, survival data, etc. | ~40,000 patients | Health monitoring in intensive care | Free (requires approval) |
| PhysioNet Challenge Datasets | PhysioNet | Various physiological signals | Varies | Testing algorithms on real signals | Free |
| Arrhythmia Dataset | UCI Machine Learning Repository | Attributes with heart arrhythmia features | 452 instances | Cardiac condition analysis | Free |
| Heart Disease Dataset | UCI Machine Learning Repository | Clinical heart disease records | 303 instances | Disease prediction and analysis | Free |
| Healthcare Activity Monitoring | Kaggle | Activity data, heart rate, sleep patterns | ~10,000 instances | Activity-based health tracking | Free |
| Simulated WBAN Data | Custom Simulation | Simulated vital signs and patient data | Configurable | Customizable scenarios for WBANs | Custom generated |
Datasets Usage Overview
| Dataset | Extracted Features | Sampling Rate Used | Payload Size in NS-3 |
|---|---|---|---|
| MIMIC-III | HR, BP, SpO2, respiration signals | 1–10 Hz | 32–64 bytes |
| PhysioNet ECG | ECG waveform samples | 125–250 Hz (down sampled to 50 Hz) | 64 bytes |
| UCI Arrhythmia | Attribute vectors | 1 Hz | 32 bytes |
| UCI Heart Disease | Feature vectors | 1 Hz | 32–40 bytes |
| Kaggle Activity | HR, activity, sleep indicators | 0.5–1 Hz | 40–64 bytes |
Indicative Network-level performance characteristics of Data Aggregation Techniques in WBANs as reported in existing literature
| Technique | Packet Delivery Ratio (%) | Error Rate (%) | Network Latency | System Efficiency |
|---|---|---|---|---|
| Edge Computing-Based Aggregation | 98.5±0.6 | 0.5±0.12 | 85±5.1 | 4.0 |
| Quantum Key Distribution (QKD) | 96.0±1.2 | 0.2±0.08 | 130±8.2 | 3.0 |
| Lightweight Homomorphic Encryption | 94.5±1.5 | 0.8±0.15 | 150±9.7 | 3.0 |
| AI-Enhanced Anonymization | 97.0±0.9 | 0.6±0.11 | 95±6.3 | 4.0 |
| Differential Privacy Techniques | 95.0±1.1 | 0.7±0.13 | 105±7.1 | 4.0 |
| Adaptive Streaming Filters | 99.0±0.8 | 0.4±0.09 | 80±4.3 | 5.0 |
Comparative Analysis of Recent Techniques in Privacy-Preserving Data Aggregation for WBANs
| References | Key Algorithm or Methodology | Dataset Used | Research Gap Addressed | Result Outcomes |
|---|---|---|---|---|
| Reddy & Ram (2020) [11] | DAP-DS (Data Aggregation and Precedence by Delay Sensitivity) | Simulated WBAN topology using NS-3 | Absence of energy-efficient, delay-aware data aggregation in WBANs | More energy efficiency, less latency, and a higher packet delivery ratio. |
| Zhang & Dong (2023) [12] | Lightweight and secure aggregation scheme for anonymous multireceivers using ECC and pseudo-identities | Simulated data, no real-world dataset | WBANs lack anonymous, secure multi-receiver aggregation | Decreased transmission and communication costs and guaranteed data integrity and anonymity. |
| Almalki & Soufiene (2021)[13] | EPPDA (Efficient Privacy-Preserving Data Aggregation) using authentication tokens and ECC | Custom simulation (IoT-based healthcare data) | In IoT healthcare, privacy-preserving, verified data aggregation is required. | Verified by simulations, high security, low overhead, and effective aggregation. |
| Soufiene et al. (2020) [16] | PEERP (Priority-based Energy-Efficient Routing Protocol) | Simulated IoT-based healthcare scenario | Insufficient energy-efficient and QoS aware routing in healthcare IoT. | Reliable data transfer, minimal energy use, and high packet delivery. |
| Salehi Shahraki et al. (2023) [2] | Key management and access control protocols | Laboratory-based WBAN setups | Inefficiency in existing key management systems | Streamlined key management, improving trust and security. |
| Herbst et al. (2024) [3] | Token-based authentication and threat mitigation | Simulated medical data scenarios | Lack of device-specific security measures | Increased security through device authentication. |
| Pawar and Kalbande (2023) [4] | ECEBA protocol for optimizing QoS | Test bed with real sensor data | Poor quality of service in data transmission | Enhanced quality and efficiency of data transmission. |
| Nawaz et al. (2024) [17] | Methodology for patient prescreening in WBANs | Hospital trial data | Inefficient early detection mechanisms | Improved early detection leading to better patient outcomes. |
| Mehmood et al. (2023) [18] | QoS-based multi-path routing scheme | Simulated healthcare monitoring | Inadequate network reliability and performance | Increased reliability and enhanced network performance. |
| Memon et al. (2023) [6] | Probabilistic route stability protocol | Real-world WBAN implementations | Frequent route failures impacting data delivery | Stabilized routing, improving data delivery reliability. |
| Masood et al. (2024) [9] | Software-defined network management for energy efficiency | Energy consumption test scenarios | High energy consumption reducing operational longevity | Significantly reduced energy usage in WBANs. |
| Singh et al. (2024) [10] | Stable matching algorithms for revenue maximization in federated networks | Economic model simulations | Inefficient economic handling in network operations | Optimized revenue outcomes enhancing economic efficiency. |
| Verma & Gupta (2023) [19] | Pairing-free authentication and aggregation using bilinear mapping and hash functions | No standard dataset; simulation used | Pairing based techniques with high overhead in intelligent healthcare system. | Reduced processing costs, scalable, safe, and effective data aggregation. |
| Mehmood et, al. (2020) [14] | Trust-based ERCS (Trust-based Energy-Efficient and Reliable Communication | Real WBANtestbedwith8-12sensor nodes | Lack of trust management in energy-efficient routing; vulnerability to | 15¬20%energy savings, >97%PDR with20%maliciousnodes, enhanced security through behavioral trust scoring |
| Khan et al. (2024) [15] | Optimized Dynamic Attribute-Based Searchable Encryption (DABSE) using ECC on BN-256curve | Simulatedmedicaldatabase (10,000encrypted records) | Highcomputational overhead insearchableencryptionlimitingreal-timequeries; need for fine-grained-accesscontrol | 65%faster search (45-60mslatency), O(n log n) complexity,128 bitsecurity with 32 byte keys, maintains data utility postencryption |
Comparative summary of reported performance characteristics of selected data Aggregation Techniques for WBANs on tends and representative values from prior studies
| Technique | Privacy Preservation | Data Integrity | Throughput | Latency (ms) | Energy Consumption | Accuracy of Data Aggregation |
|---|---|---|---|---|---|---|
| Edge Computing Based Aggregation | 4.0 | 4.0 | 340±12.5 | 85±5.1 | 0.38±0.03 | 92.0±1.8 |
| Quantum Key Distribution (QKD) | 5.0 | 5.0 | 280±15.3 | 130±8.2 | 0.60±0.04 | 95.0±1.2 |
| Lightweight Homomorphic Encryption | 5.0 | 4.0 | 250±18.7 | 150±9.7 | 0.65±0.05 | 90.0±2.1 |
| AI-Enhanced Anonymization | 4.0 | 3.0 | 320±14.2 | 95±6.3 | 0.42±0.03 | 88.0±2.5 |
| Differential Privacy Techniques | 5.0 | 4.0 | 300±13.8 | 105±7.1 | 0.45±0.04 | 91.0±1.9 |
| Adaptive Streaming Filters | 3.0 | 3.0 | 350±11.3 | 80±4.3 | 0.33±0.02 | 85.0±2.8 |
Privacy Preserving Scoring Rubric
| Score | Definition & Criteria |
|---|---|
| 1 – Very Low | No encryption; vulnerable to eavesdropping; no anonymity or key management. |
| 2 – Low | Basic symmetric encryption; susceptible to replay or compromise; no resistance to traffic analysis. |
| 3 – Moderate | Lightweight ECC or token-based authentication; moderate resistance to passive attackers. |
| 4 – High | Differential privacy (ε ≥ 1), CP-ABE, or authenticated encryption; resists common WBAN adversaries; low leakage probability. |
| 5 – Very High | Formal guarantees such as Homomorphic Encryption, QKD-based keying, or DP with ε < 1; resistant to strong adversaries; cryptographically provable security. |
Data Integrity Scoring Rubric
| Score | Definition & Criteria |
|---|---|
| 1 | No checksum or integrity check. |
| 2 | Basic checksum; vulnerable to tampering. |
| 3 | MAC or hash-based verification (SHA-1/SHA-256). |
| 4 | ECC/ABE-based authenticated integrity; tamper detection with low false-positive rate. |
| 5 | Fully verifiable aggregation (HE or signature-based); strong resistance to forgery. |
Implementation Parameters for Each Technique
| Technique | Library/Tool | Key Size/Parameter | Notes |
|---|---|---|---|
| QKD | Simulated BB84 protocol (QKDsim v2.1) | 128 bit | Quantum bit error rate = 2% |
| ABE | CP-ABE (Charm-Crypto 0.50) | 256-bit ECC keys | Attribute Count = 5 |
| Homomorphic Enc. | PySEAL 4.1 | 4096 polynomial modulus | CKKS Scheme |
| Differential Privacy | Numpy + Laplace noise | ε = 0.5–1.0 | - |
| Adaptive Streaming Filters | Custom NS-3 Module | - | Sampling interval = 1 s |