Abstract
As power grids transition toward decentralized architectures, Transactive Energy Systems (TES) have emerged as a critical mechanism for balancing supply and demand through market-based interactions. However, the reliance on insecure communication links introduces significant privacy and security vulnerabilities. This paper presents a unified cryptographic framework designed to protect individual bidding information and detect malicious data injection attacks. By integrating an enhanced Paillier encryption scheme with a novel "block design" and a stamp-concatenation digital signature mechanism, the proposed system achieves high computational efficiency and robust cyber-resiliency.
Introduction
Transactive Energy Systems (TES) facilitate the optimal allocation of energy resources by engaging suppliers and consumers in market operations. In a typical TES, participants submit bidding curves—representing their willingness to buy or sell energy at specific prices—to a central coordinator. While promising, this approach faces two primary challenges when deployed over public or insecure networks: the leakage of sensitive consumer habits or business secrets (privacy) and the potential for data tampering or reordering by malicious actors (security).
Problem Statement
Existing research in TES often prioritizes coordination mechanisms over cyber-resiliency. Standard encryption methods frequently introduce prohibitive computational overhead, making them unsuitable for real-time market clearing. Furthermore, traditional digital signatures may verify the sender's identity but fail to detect sophisticated temporal attacks, such as message reordering or replay attacks, which can destabilize the grid or lead to financial fraud.
Proposed Methodology
The authors propose a unified framework leveraging the additively homomorphic properties of Paillier cryptography. This allows a coordinator to perform necessary mathematical operations on encrypted data without ever accessing the underlying plaintext.
1. Privacy Preservation via Block Design
To address the computational bottleneck of point-wise encryption, the framework introduces a "block design." Instead of encrypting each point on a bidding curve individually, multiple data points are concatenated into a single large integer. This allows the entire curve to be encrypted and aggregated in a single operation, significantly reducing processing time. A semi-honest Third-Party Aggregator (TP) is utilized to sum these encrypted bids, ensuring the coordinator only receives the aggregate total.
2. Security via Stamp-Concatenation
To counter data injection and reordering attacks, the framework employs an enhanced digital signature scheme. Each message is appended with a unique index "stamp" before being signed. This mechanism ensures that the receiver can verify not only the integrity of the data but also its correct sequence and timeliness, preventing attackers from successfully replaying or swapping valid messages.
Experimental Validation
The framework was tested using a simulated distribution feeder with 1,000 residential air conditioning (AC) units. The simulation incorporated:
- Load Dynamics: A second-order equivalent thermal parameter (ETP) model for AC units.
- Environmental Data: Typical Meteorological Year (TMY2) data for Columbus, Ohio.
- Market Data: Real-world wholesale electricity prices from the PJM market.
Key Results
The simulation results demonstrated that the proposed cryptographic enhancements do not compromise market accuracy; clearing prices remained identical to those calculated using plaintext data.
The "block design" provided a massive efficiency boost. For a bidding curve with 101 sampled points using a 4000-bit key, the processing time per agent was reduced from 128 seconds to 1.4 seconds—a 101-fold improvement. Additionally, the stamp-concatenation mechanism successfully detected four distinct attack types, including data replacement and reordering, which standard signature schemes failed to identify.
Conclusion and Future Work
This research provides a practical path for the secure deployment of TES in modern power systems. While motivated by TES, the proposed framework is broadly applicable to enhancing privacy and security in any general hierarchical control system. Future work will address "Byzantine attacks" where legitimate participants may intentionally submit distorted data. Further investigation into preventing collusion between the coordinator and the third-party aggregator will also be essential for ensuring long-term system integrity.
Summary updated