Implementation and Evaluation of Adaptive Network Steganography for Secure IoT Communications: A Smart Electricity Metering Case Study

The Internet of Things (IoT) has enabled smart electricity metering as part of smart city infrastructure, enabling automated consumption reporting, accurate billing, real-time outage visibility and demand response, and remote operational control. However, this advancement creates an additional cyber–physical attack surface beyond the existing one, including attacks targeting smart meters, gateways, brokers, edge nodes, and utility head-end systems (HESs). The recently reported literature continued to reflect growing concerns over IoT cybersecurity, smart meter threats, edge computing, and covert channel behavior in IoT environments. This article introduces Adaptive Steganography Trust Signaling (ASTS), a defensive network steganography framework that embeds very small, authenticated trust signals within normal smart meter traffic to enable tamper evidence and edge-level integrity verification. The purpose of malicious covert channels is to hide unauthorized communications; however, ASTS is both policy-bound and rate-limited for auditing purposes and to serve as an additional layer of support for conventional cryptographic authentication methods. The flow of smart electricity meter data (and other utility services) is supported by ASTS through bounded timing and ordering, as specified by policy and protocol definitions within that framework. A risk-based decision model embedded in the framework identifies the appropriate method for integrating ASTS into networks based on factors such as anomaly detection, network conditions, device trust levels, and the importance of the services utilized. This paper provides algorithms for adaptive embedding, verification, and policy-driven adaptation. A synthetic evaluation model that simulates periodic meter readings, replay attempts, tampering events, intermittent network degradation, and data manipulation events is used to assess performance before field deployments. The contribution of this work is a complete framework for adaptive defensive network steganography in smart electricity metering. Based on the experiments conducted, the proposed ASTS model demonstrated broader operational awareness and adaptive capabilities than the baseline variants, while closing gaps identified in several related works.