Skip to content
All case studies

Securing a cardiac wearable

My Computer Engineering thesis: can a small ECG wearable encrypt its data without slowing down? Published at SBSEG 2022.

Role
Researcher
Timeline
Thesis · 2022
Stack
C++ · ESP32 · MQTT · cryptography

01 · The problem

Health data on an open network

Medical IoT devices are spreading fast. A cardiac wearable that sends ECG readings over wireless and remote networks is an easy target for man-in-the-middle and eavesdropping attacks, yet it has very little computing power for encryption.

02 · Constraints

Tiny device, real-time data

  • Low-power hardware with limited CPU and memory
  • ECG signals have to arrive on time, so latency matters
  • The security layer must not cut throughput

03 · Options

Three algorithms

  • Chosen

    AES-256 CBC

    The established standard, and the best latency and throughput in our tests.

  • SPECK

    A lightweight block cipher designed for constrained devices.

  • CLEFIA

    A lightweight block cipher standardised for embedded use.

04 · Decision

Measure, don’t guess

“We implemented all three on the wearable and ran load tests, measuring latency and throughput while ECG signals were being sent.”
B. Manoel et al., SBSEG 2022, pp. 97–110

05 · Outcome

What we found

  • 3algorithms implemented and load-tested
  • AES-256best results for latency and throughput
  • 1peer-reviewed paper, SBSEG 2022

All three performed well enough, which shows a security layer is feasible on an IoMT device. AES-256 CBC came out on top, making it the best fit for the cardiac wearable.

06 · What I learned

Trade-offs, measured

This project taught me to back decisions with data. I still use that habit in frontend work: when two options look equal, I measure.

Keep reading