Home / Current Issue / Paper 1714174
Communication-Based Theft Detection Models for Nigeria’s Prepaid Metering Infrastructure
Subject area: Science,Engineering and Technology · Area of research: Information and Communication Engineering
DOI: https://doi.org/10.64388/IREV9I8-1714174
Abstract
Electricity theft remains a major contributor to non-technical losses in Nigeria’s power distribution sector despite the widespread deployment of prepaid metering systems. Existing theft-detection approaches are predominantly consumption-based and increasingly ineffective against sophisticated fraud techniques such as communication jamming, token manipulation, and partial meter bypassing. This paper proposes and evaluates a communication-based, multi-model framework for electricity theft detection tailored to Nigeria’s prepaid Advanced Metering Infrastructure (AMI). Four analytical models are developed: the Consumption Pattern Analysis Model (CPAM), the Communication Integrity Model (CIM), the Vending and Token Behaviour Model (VTBM), and the Peer Comparison and Cluster Deviance Model (PCDM). These models are integrated using a risk-based decision framework to improve detection robustness. The proposed approach is evaluated using a real prepaid metering dataset comprising 100 customer records from a Nigerian distribution network. Performance is assessed using accuracy, precision, recall, and F1-score. Results show that communication-based indicators significantly outperform consumption-only analysis, with the integrated framework achieving 90% accuracy and an F1-score of 93%. The findings demonstrate that incorporating communication integrity and vending behaviour analytics into prepaid metering systems offers a practical, scalable solution for detecting electricity theft in developing-country AMI environments.
Keywords
Electricity Theft Detection; Advanced Metering Infrastructure; Communication Integrity; Prepaid Metering; Smart Grid Analytics
How to cite this paper
@article{1714174,
author = {Owoeye Akinyemi Sunday, Eseosa Omorogiuwa},
title = {Communication-Based Theft Detection Models for Nigeria’s Prepaid Metering Infrastructure},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {9},
number = {8},
pages = {501-504},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1714174.pdf},
abstract = {Electricity theft remains a major contributor to non-technical losses in Nigeria’s power distribution sector despite the widespread deployment of prepaid metering systems. Existing theft-detection approaches are predominantly consumption-based and increasingly ineffective against sophisticated fraud techniques such as communication jamming, token manipulation, and partial meter bypassing. This paper proposes and evaluates a communication-based, multi-model framework for electricity theft detection tailored to Nigeria’s prepaid Advanced Metering Infrastructure (AMI). Four analytical models are developed: the Consumption Pattern Analysis Model (CPAM), the Communication Integrity Model (CIM), the Vending and Token Behaviour Model (VTBM), and the Peer Comparison and Cluster Deviance Model (PCDM). These models are integrated using a risk-based decision framework to improve detection robustness. The proposed approach is evaluated using a real prepaid metering dataset comprising 100 customer records from a Nigerian distribution network. Performance is assessed using accuracy, precision, recall, and F1-score. Results show that communication-based indicators significantly outperform consumption-only analysis, with the integrated framework achieving 90% accuracy and an F1-score of 93%. The findings demonstrate that incorporating communication integrity and vending behaviour analytics into prepaid metering systems offers a practical, scalable solution for detecting electricity theft in developing-country AMI environments.},
keywords = {Electricity Theft Detection; Advanced Metering Infrastructure; Communication Integrity; Prepaid Metering; Smart Grid Analytics},
month = {February},
doi = {https://doi.org/10.64388/IREV9I8-1714174}
}