Abstract
In the era of cloud computing, protecting message content is insufficient if the metadata—who is communicating with whom—remains exposed to traffic-analysis attacks. This paper introduces LacminCC, a lightweight anonymous communication model designed to provide sender, receiver, and relationship anonymity. By combining a novel anonymous Identity-Based Encryption (IBE) scheme with a strategic grouping and time-slicing architecture, LacminCC overcomes the high overhead and scalability issues common in traditional systems like Tor or DC-nets. Simulation results indicate that LacminCC maintains low latency and constant communication costs, making it a robust solution for secure communication within cloud infrastructures.
Introduction
As cloud computing becomes the backbone of modern data exchange, the threat of traffic analysis has grown. Even when message contents are encrypted, observers can deduce sensitive information by analyzing communication patterns. Existing Anonymous Communication (AC) systems often face a "trilemma": Onion Routing (like Tor) is vulnerable to exit-node attacks; DC-nets are computationally heavy and limit throughput; and flooding algorithms generate excessive network traffic.
LacminCC addresses these gaps by providing a secure, stable, and lightweight model specifically tailored for small-to-medium groups, such as corporate bidders or whistleblowers, operating within cloud environments.
Technical Foundations: Anonymous IBE
The core of LacminCC is a custom Anonymous Identity-Based Encryption (IBE) scheme. Unlike traditional encryption that requires complex digital certificates, IBE allows a user's identity (e.g., an email address) to serve as their public key.
The researchers mathematically proved the security of this scheme under the Bilinear Diffie–Hellman (BDH) Assumption. A critical feature of this IBE variant is its "anonymity": an observer cannot determine which public key was used to generate a specific ciphertext, preventing the leakage of recipient identities during the transmission phase.
The LacminCC Architecture
The model operates through a structured four-phase process ($T_1$ to $T_4$) involving encryption, uploading, downloading, and decryption. Key architectural features include:
- Grouping and Filtering: Users are divided into $M$ groups of $N$ members. To optimize resources, the system uses a "Group Filter" ($C_1$). Instead of downloading every message on the cloud bulletin board, users only download messages tagged with their specific group ID, drastically reducing memory and bandwidth consumption.
- Traffic Masking: To defeat traffic analysis, every user must upload a ciphertext during the upload phase. If a user has no real message to send, they upload a "pseudo-ciphertext" (dummy traffic), ensuring that an observer cannot distinguish between active and idle users.
- Dynamic Security: To prevent long-term cryptographic attacks, the Private Key Generator (PKG) automatically regenerates user keys every 1000 rounds and reshuffles group memberships every 100 rounds.
Performance and Security Analysis
LacminCC was evaluated against Onion Routing, DCARPS, and Anonymous Path Routing. It was the only model to satisfy all three primary anonymity criteria: sender, receiver, and communication relationship anonymity.
From an efficiency standpoint, LacminCC offers significant advantages:
- Storage: It requires only one hashing, one encryption, and one decryption operation per round, making it leaner than multi-layered Onion Routing.
- Communication Cost: The cost remains constant relative to the group size ($N$), avoiding the exponential overhead spikes seen in competing models as message volume increases.
- Throughput: Unlike many DC-net implementations that allow only one sender per round, LacminCC supports multiple concurrent senders without protocol breakdown.
Conclusion
LacminCC provides a scalable and efficient framework for anonymous communication in the cloud. While the current model relies on a trusted Private Key Generator (PKG) and is optimized for small-to-medium groups, its ability to maintain a flat latency profile under increasing load marks a significant step forward. Future work will focus on decentralizing the trust model and further optimizing the protocol for massive-scale public networks.
Summary updated