Performance Analysis of Smart Home Security Using Dynamic Authentication QoS Network Classification
Implementation Plan:
************************
Step 1: Initially, we constructed a smart-home network with 30 IoT devices, 1 edge gateway, and 1 server, communicating via encrypted MQTT protocols in MATLAB.
Step 2: Then, we collect behavioral fingerprinting data, flow-level encrypted MQTT QoS features, and authentication logs from IoT nodes using a ML Edge-IIoTset dataset
Step 3: Next, we preprocess the collected data by extracting timing, statistical, and behavioral features, perform normalization, and generate dynamic device signatures for continuous authentication and QoS-driven traffic classification.
Step 4: Then, we train LightGBM models for node-level authentication and encrypted traffic classification, and quantize a Phi-3 Mini SLM (INT4) for gateway-level alert reasoning and policy orchestration based on collected data.
Step 5: Next, we optimized models by applying swarm-based DNN hyperparameter tuning methods to balance detection accuracy, latency, and resource usage based on collected data.
Step 6: Then, we implement privacy-preserving federated learning updates to analyze improving detection accuracy
Step 7: Finally, we evaluate and plot performance metrics for the following:
7.1: Number of IoT Devices vs. Authentication Accuracy (%)
7.2: Number of IoT Devices vs. End-to-End Latency (ms)
7.3: Number of IoT Devices vs. Throughput (Kbps)
7.4: Number of IoT Devices vs. CPU Utilization (%)
7.5: Number of IoT Devices vs. Detection Accuracy (%)
Software Requirements:
***************************
1. Development Tool: Matlab-R2023a or above
2. Operating System: Windows-10 (54-bit) or above
Dataset:
*********
Link : https://www.kaggle.com/datasets/sibasispradhan/edge-iiotset-dataset?select=ML-EdgeIIoT-dataset.csv
Note:
*******
1) If the proposed plan does not fully align with your requirements, please provide all necessary details—including steps, parameters, models, and expected outcomes—in advance. Once implementation begins, modifications will not be feasible without prior input. Kindly ensure that any missing configurations or specifications are clearly outlined in the plan before confirming, as post-implementation changes will not be accommodated.
2) If there’s no built-in solution for what the project needs, we can always turn to reference models, customize our own, different math models or write the code ourselves to fulfil the process.
3) If the plan satisfies your requirement, Please confirm with us.
4) Project based on Simulation only, not a real time project.
