Table 2 Issues OF SG and its countermeasures.
From: A hybrid AI-Blockchain security framework for smart grids
Attack | Attack description | Countermeasure | Ref. |
|---|---|---|---|
TSA | Because of this attack, the PMU’s timing has been thrown off. | Shift-invariant data transmission and error-correlated average estimate. | |
Spoofing | When challenged by the intended victim, attackers may pretend to be a trustworthy third party. | Access control, anti-virus software, firewall, VPN, IDS, fusion-based defense strategy, game-theoretic analysis, and message encoding. | |
Sniffing | This attack steals transferred data between sender and destination. | Message encryption. | |
MITM | This cyber threat arises because an adversary can impersonate an authorized user and access sensitive data using their credentials. | In SG, lightweight physical unlovable function (PUF) authentication prevents MITM attacks. The smart meter uses PUF authentication to transfer enrolment credentials. PUF scheme shares unique keys as digital footprints. Keys allow nodes to share encrypted data. | |
LAA attack | Load-altering assaults (LAAs) boost load power consumption to overwhelm lines. Historically, LAAs have used two methods: direct hacking and indirect alteration through exploitation. | The DLAA requires a power system assault since two vulnerable loads were chosen to test system assaults. A robust observer mode is then created to monitor load frequency and residual signals. The evaluation was done on three generators and six power system buses to prove detection is achievable. DLAA can be identified by validating an SG model and developing an adaptive fading Kalman filter (AFKF) to forecast the SG’s state. | |
Malicious command injection attack | Controlling energy flow in electrical grids can be done with phase-shifting transformers, also called phase shifters. Phase shifters reduce electricity congestion in transmission lines and regulate based on contractual agreements. | Detecting malicious code in smart meters using a LSTM network-based technique on the CPU or MCU power channel. This technique uses LSTM. A real-case smart meter evaluation found 92% efficiency. A lightweight approach is recommended for detecting stealthy malicious tap-changed commands. | |
Load redistribution attack | According to the author of [34], load redistribution attacks are linked to state estimation-false data injection attacks. These assaults compromise load bus and electrical flow measurements, while total power demand remains intact. Due to this, the assault causes a load redistribution across the network. | A closest neighbor-based load redistribution attack detection approach is offered. This approach maintains promising detection performance as it scales from small to large systems. On an LR assault with unsystematic anomalous load variations, sensitivity analysis and broad testing are done. Additionally, statistical approaches can localize attacks and predict the likelihood of each load being attacked. | |
Brute force | An attacker will probe every part of a system to find holes in its defenses. | User authentication, encryption, and complex passwords are advised. | |
Eavesdropping | Unauthorized people can access transferred data by pretending to be lawful hardware like monitoring sensors. | MILP, access control list, encryption, VPN, and tree model to determine attack profitability. | |
CPU overload | The target central processing unit is swamped with irrelevant data, wasting computing power and electricity. | Anti-virus solutions. | |
TCP-SYN flooding | This attack delays the victim’s equipment’s synchronization, violating timing limitations. | The authentication of messages. | |
Jamming | Attackers can modify a device’s signal frequency or loudness. Other approved devices cannot contact the victim. | Replace the compromised smart meter and change the channels. |
