---

VitalCache: Multi-Vital Sign Cache Analyzer

Bridging Computer Architecture and Biomedical Systems through real-time embedded cache simulation.

Overview

VitalCache is an embedded system project that integrates real-time biomedical sensing with cache memory architecture concepts.
The system collects live physiological data — heart rate (PPG), temperature, and SpO₂ — and simulates how this data is stored and retrieved using different cache mapping techniques.
This project demonstrates how cache performance impacts real-time healthcare systems, making it both a learning tool and a research prototype.

Team Members

PRAGALYA M , YOUVASHREE K , YOUVASHREE K

Objectives

Simulate cache memory interaction with biomedical data

Compare:

• Direct Mapping
• Fully Associative Mapping
• Set-Associative Mapping

Analyze cache performance metrics (hit, miss, conflict)
Demonstrate real-time embedded system integration

Key Features

• Real-time sensor data acquisition (PPG, Temp, SpO₂)
• Cache simulation with 3 mapping techniques
• FIFO-based replacement policy
• Live performance metrics (Hit / Miss / Conflict)
• Health alerts:
  • Fever detection
  • Abnormal heart rate
  • Low SpO₂

Hardware Used

• ESP32 Microcontroller
• PPG Sensor (Analog Pulse Sensor)
• MAX30102 / MAX30105 (SpO₂ + HR)
• LM35 Temperature Sensor

Pin Configuration

Component	Pin
PPG Sensor	GPIO 34
LM35 Temp	GPIO 35
MAX30102 SDA	GPIO 21
MAX30102 SCL	GPIO 22

(Based on system design from project PPT)

Concepts Implemented

• Cache Memory Fundamentals
• Temporal & Spatial Locality
• Cache Lines / Blocks
• Tag, Index, Offset

Mapping Techniques

• Direct Mapping
• Fully Associative Mapping
• 2-Way Set Associative Mapping

Metrics

• Cache Hit
• Cache Miss
• Conflict Miss

System Workflow

Sensors collect real-time data
Data stored in main memory blocks
User requests block via serial input
Cache mapping algorithm processes request

Output:

• Hit / Miss / Conflict
• Sensor values
• Execution result
• Health alerts triggered if abnormal

How to Run

1. Setup

Install Arduino IDE
Install required libraries:

• MAX30105
• spo2_algorithm

2. Upload Code

Select board: ESP32 Dev Module
Upload the provided .ino file

3. Open Serial Monitor

Baud rate: 115200

Input Format

Enter in Serial Monitor:

<mappingType> <blockIndex>

Example:

• 0 5 → Direct Mapping
• 1 3 → Fully Associative
• 2 7 → Set Associative

Sample Output

Fetch: Mapping 1, Block 4
Decode: 36.5°C, 78 bpm, 98%
Execute: Sum = 212
Hit:1 Miss:0 Conflict:0

Results & Insights

From experimental observations:

Mapping Type	Hit Ratio
Direct Mapping	~0.32
Fully Associative	~0.36
Set-Associative	~0.36

Associative and Set-Associative mapping perform better due to reduced conflicts.

Health Monitoring Alerts

• Temperature > 38°C → Fever
• HR < 60 or > 100 → Abnormal
• SpO₂ < 95% → Low Oxygen

Applications

• Smart healthcare monitoring systems
• Embedded system education
• Cache architecture visualization tool
• IoT-based health analytics

Innovation / Contribution

Combines hardware-level biomedical sensing with cache simulation
Provides real-time evaluation of memory architectures
Acts as a practical learning bridge between:

• Computer Architecture
• Embedded Systems
• Biomedical Engineering

References

Key research areas:

• Cache-assisted health monitoring
• IoT healthcare systems
• Memory hierarchy optimization

Future Work

• Add cloud integration (IoT dashboard)
• Implement LRU replacement policy
• Improve signal processing accuracy
• Add mobile/web visualization

License

MIT

Acknowledgment

Developed as part of:
24AIM204 – Foundations of Computer Architecture

---