Virtual Sensor Hub

This is a multi-threaded C program that simulates sensors, produces time-stamped samples, computes moving averages and emits alerts. It is designed as a software-only demo of embedded/IoT-oriented C development. It is a compact, testable C project that demonstrates systems-oriented skills relevant to embedded/IoT software such as concurrency, deterministic sampling, simple protocol framing, log-based verification and a small test harness. No hardware required.

Key files

• src/main.c - program entry, duration handling, shutdown logic
• src/sensor.c, sensor.h - deterministic sensor threads
• src/hub.c, hub.h - logging, in-memory queue, processor (moving average + alerts)
• Makefile - one-command build (make)
• data/hub.log - runtime outputs
• tools/check_log.py - Python validator for data/hub.log
• tests/run_tests.sh - orchestrated test harness (runs app + validator)
• tools/parse_logs.py - generates charts and a CSV summarizing the output

Design

• Sensor threads (sensor.c)
  Each sensor (TEMP, HUM, PRESS) runs in its own thread and produces deterministic sample sequences at a configured interval. Determinism enables reproducible runs and stable automated checks.

• Submission & queue (hub.c)
  hub_submit_sample() enqueues incoming samples into a fixed-size circular queue and immediately logs a SAMPLE|... line to data/hub.log. Mutex + condition variable coordinate producer/consumer access.

• Processor thread (hub.c)
  A separate processor consumes queued samples, maintains a sliding moving-average window per sensor (configurable window size) and writes ALERT|...|THRESHOLD_EXCEEDED lines when a windowed average crosses a threshold.

• Logging & verification
  All samples and alerts are appended to data/hub.log (human-readable framed lines). A Python validator (tools/check_log.py) inspects the log to verify expected sample counts and alerts for automated testing.

• Analysis & Visualizations
  The logs are analysed by tools/parse_logs.py and visualizations (histogram + timseries) are generated for all three sensors, along with a timeline of alerts and a CSV summarizing the sensor readings.

Prerequisites

Run in WSL2 (Ubuntu) or any Linux environment with:

sudo apt update
sudo apt install -y build-essential make gdb python3

Execution

To build, run from project root:

make

To run indefinitely (Ctrl+C to stop):

./sensorhub

To run for a fixed duration:

./sensorhub --test-duration 8   # run 8 seconds then exit

The program writes trace lines to data/hub.log (SAMPLE and ALERT framed records).

Testing & Analysis

Run this automated test after building:

./tests/run_tests.sh          # default duration 6s
# or
./tests/run_tests.sh 8        # specify duration in seconds

Run this to perform the analysis after the log file has been generated:

python3 tools/parse_logs.py data/hub.log --outdir outputs --window 5

Visualizations

The following are some sample outputs obtained after executing the program for 35 seconds.

Histograms	Timeseries

Architecture Diagram