Comparing Communication Technologies for Electricity Smart Meters

Smart Meter Communication: Reproducible Results Reproduces Tables I–VIII and Figure 1 from the smart-meter comparison paper. Numerical results use the paper's scores and weights; descriptive tables use transcribed text. The PDF is not required, and its SHA-256 fingerprint is recorded in inputs.json. Quick start Requires Python 3.11+. The included run used Python 3.12.14. After extracting the ZIP: cd smart_meter_reproduction python -m pip install -r requirements.txt python generate_results.py --check-paper Use python3 if needed. Results are saved in outputs/; generated outputs are also included. LaTeX is not required to run the script. --check-paper checks calculated results against paper_expected.json and raises an error on a mismatch. Expected values are used only for verification. Main files File Purpose generate_results.py Calculations, sensitivity tests, tables, and chart inputs.json Scores, weights, assumptions, technology order, and seeds paper_expected.json Published results for verification requirements.txt Pinned dependency versions outputs/ CSV, Markdown, and LaTeX tables; PNG/SVG chart; detailed results, summary, and run metadata Expected results W varies weights; W+S varies weights and scores. Each test uses 10,000 runs per scenario. Scenario First choice Weighted score TOPSIS W wins W+S wins Urban Wi-SUN FAN 4.29 0.663 99.98% 57.47% Rural LoRaWAN 4.38 0.761 100.00% 69.92% Remote LoRaWAN 4.61 0.971 100.00% 83.16% Change assumptions Copy and edit inputs.json, then run: python generate_results.py --inputs my_inputs.json --out my_results Omit --check-paper when changing inputs. Use --runs 50000 to change the simulation size; percentages may differ from the paper. Reproducibility and scope The simulations use NumPy PCG64 with seeds 20260922 (W) and 20260923 (W+S). Preserve input order, random-draw order, and the generator's continuity across scenarios. Exact ties follow the fixed technology order. Higher scores always mean better suitability, including lower CapEx and OpEx. Do not invert cost scores again in TOPSIS. Scores and weights are illustrative judgments. Reproducing the calculations does not validate field performance, and simulation percentages are not confidence levels. All five technologies are assumed feasible. Check local coverage, permissions, power lines, and backhaul before applying the rankings. Table I retains the source PDF's reference numbers. Check bibliography accuracy and update these numbers if references are reordered.