Conventional power grids are being superseded by smart grids, which have smart meters as one of the key components. Currently, for the smart metering communication, wireless technologies have predominantly replaced the traditional Power Line Communication (PLC). Different vendors manufacture smart meters using different wireless communication technologies. For example, some vendors use WiMAX, others prefer Low-Power Wireless Personal Area Networks (Lo-WPAN) for the Media Access Control (MAC) and physical layer of the smart meter network, also known as Advanced Metering Infrastructure (AMI) network. Different communication techniques are used in various components of an AMI network. Thus, it is essential to create a testbed to evaluate the performance of a new wireless technology or a novel protocol to the network. It is risky to study cyber-security threats in an operational network. Hence, a real-time simulation testbed is considered as a substitute to capture communication among cyber-physical subsystems. To design the communication part of our testbed, we explored a Cellular Internet of Things (CIoT) : Co-operative Ultra NarrowBand (C-UNB) technology for the physical and the MAC layer of the Neighborhood Area Network (NAN) of the AMI. After successful evaluation of its performance in a Simpy python simulator, we integrated a module into Network Simulator-3 (NS-3). As NS-3 provides a platform to incorporate real-time traffic to the AMI network, we can inject traffic from power simulators like Real Time Digital Simulator (RTDS). Our testbed was used to make a comparative study of different wireless technologies such as IEEE 802.11ah, WiMAX, and Long Term Evolution (LTE). For the traffic, we used HTTP and Constrained Application Protocol (CoAP), a widely used protocol in IoT. Additionally, we integrated the NS-3 module of Device Language Message Specification - Companion Specification for Energy Metering (DLMS-COSEM), that follows the IEC 62056 standards for electricity metering data exchange. This module which comprises of application and transport layers works in addition with the physical and MAC layer of the ii C-UNB module. Since wireless communication is prone to eavesdropping and information leakages, it is crucial to conduct security studies on these networks. Hence, we performed some cyber-attacks such as Denial of Service (DoS), Address Resolution Protocol (ARP) spoofing and Man-in-the-Middle (MiTM) attacks in the testbed, to analyze their impact on normal operation of AMI network. Encryption techniques can alleviate the issue of data hijacking, but makes the network traffic invisible, which prevents conventional Intrusion Detection Systems (IDS) from undertaking packet-level inspection. Thus, we developed a Bayesian-based IDS for ARP spoof detection to prevent rogue smart meters from modifying genuine data or injecting false data. The proposed real time simulation testbed is successfully utilized to perform delay and throughput analysis for the existing wireless technologies alongwith the evaluation of the novel features of C-UNB module in NS-3. This module can be used to evaluate a broad range of traffic. Using the testbed we also validated our IDS for ARP spoofing attack. This work can be further utilized by security researchers to study different cyber attacks in the AMI network and propose new attack prevention and detection solution. Moreover, it can also allow wireless communication researchers to improve our C-UNB module for NS-3.
