Multi-Transmitter Coded Caching with Secure Delivery over Linear Networks - Extended Version

Mohammad Javad Sojdeh, Mehdi Letafati, Seyed Pooya Shariatpanahi, Babak Hossein KhalajarXiv 2022 · DOI · arXiv

Abstract

This paper investigates the challenge of providing information-theoretic security in multi-transmitter coded caching systems. By integrating one-time pad encryption with zero-forcing techniques over linear networks, the authors propose schemes for both centralized and decentralized caching scenarios. The analysis demonstrates that the overhead for securing data against passive eavesdroppers becomes asymptotically negligible as the network size and file library scale. Numerical simulations verify that the proposed methods effectively leverage spatial multiplexing gains to offset security costs, ensuring zero information leakage without sacrificing long-term network efficiency.

Introduction

Coded caching has emerged as a powerful technique to alleviate network congestion by pre-storing popular content in end-user devices during off-peak hours. During the delivery phase, transmitters send "coded" messages that simultaneously satisfy multiple users' requests, achieving a "global caching gain." In multi-transmitter environments, these systems also benefit from "spatial multiplexing gains," which further accelerate delivery by allowing multiple independent data streams to be sent simultaneously.

However, most existing research assumes a perfectly secure delivery channel. In practice, wireless and linear networks are susceptible to passive eavesdroppers who can intercept transmitted packets. This paper addresses a critical security gap: How can we ensure that an eavesdropper gains zero information about the original files while maintaining the efficiency of multi-transmitter coded caching?

System Model and Methodology

The authors model a system where multiple transmitters serve a set of users over a linear network, as illustrated in Figure 1. The model consists of $L$ transmitters, $K$ users with local caches of size $M$, and a library of $N$ files. A transfer matrix $H$ represents the linear network, and a passive eavesdropper is assumed to observe the delivery phase.

System model for secure multi-transmitter coded caching over a linear network featuring $L$ transmitters, $K$ users, and a passive eavesdropper.

The proposed framework relies on two primary pillars to achieve information-theoretic security:

  1. Secret Shared Keys (One-Time Pad): During the placement phase—when the eavesdropper is assumed to be absent—transmitters distribute random secret keys to users. In the subsequent delivery phase, these keys are used to encrypt multi-cast messages, ensuring that only authorized users can decode the content.
  2. Zero-Forcing (ZF) over Linear Networks: The system utilizes multiple transmitters to perform zero-forcing. This technique ensures that a coded message intended for a specific subset of users is "canceled out" at other users who do not require it. Crucially, ZF is also used to manage the signal observed by the eavesdropper, ensuring that the intercepted signals do not reveal the underlying file content.

The paper adapts three state-of-the-art multi-transmitter schemes to incorporate these security measures: a standard secure model, a scheme with reduced subpacketization to address practical file-splitting constraints, and a reduced feedback scheme to minimize the required channel state information.

Theoretical Analysis and Key Insights

A central finding of this work is that the "security cost"—the additional memory for keys and the increased coding delay—becomes asymptotically negligible. As the number of files ($N$) and users ($K$) in the library increases, the performance of the secure scheme converges toward that of the insecure baseline. This suggests that large-scale networks can be secured with minimal impact on throughput.

Furthermore, the use of a "linear network" transfer matrix $H$ provides significant generality. This mathematical abstraction allows the results to be applied to a wide variety of physical layers, including wired networks using network coding and wireless Multiple-Input Single-Output (MISO) broadcast channels. The authors also demonstrate decentralized robustness; even if users cache content independently (decentralized placement), the system remains secure and eventually performs nearly as well as the centralized version in large-scale scenarios.

Results and Discussion

The authors provide exact closed-form expressions for the secure coding delay ($T$) and the fraction of memory dedicated to keys ($M_K$). For instance, Theorem 1 in the paper defines the achievable delay for a centralized system with $L$ transmitters.

Information-theoretic security is rigorously proven using the Mutual Information metric. The authors demonstrate that the eavesdropper's observation ($y_e$) reveals exactly zero bits of information about the library ($W$), satisfying the condition $I(y_e; W_1, ..., W_N) = 0$. Numerical simulations confirm these theoretical derivations, showing that increasing the number of transmitters ($L$) significantly reduces secure coding delay by leveraging multiplexing gains to offset security overhead.

Conclusion and Future Work

This research provides a robust framework for securing multi-transmitter coded caching systems. While the current model assumes the eavesdropper is absent during the placement phase and possesses only a single antenna, it lays the groundwork for more complex scenarios. Future research directions include generalizing the model to account for multi-antenna or randomly located eavesdroppers, as well as deriving converse bounds to prove the optimality of the proposed achievable rates.

Summary updated

Abstract

In this paper, we consider multiple cache-enabled end-users connected to multiple transmitters through a linear network. We also prevent a totally passive eavesdropper, who sniffs the packets in the delivery phase, from obtaining any information about the original files in cache-aided networks. Three different secure centralized multi-transmitter coded caching scenarios namely, secure multi-transmitter coded caching, secure multi-transmitter coded caching with reduced subpacketization, and secure multi-transmitter coded caching with reduced feedback, are considered and closed-form coding delay and secret shared key storage expressions are provided. As our security guarantee, we show that the delivery phase does not reveal any information to the eavesdropper using the mutual information metric. Moreover, we investigate the secure decentralized multi-transmitter coded caching scenario, in which there is no cooperation between the clients and transmitters during the cache content placement phase and study its performance compared to the centralized scheme. We analyze the system's performance in terms of Coding Delay and guarantee the security of our presented schemes using the Mutual Information metric. Numerical evaluations verify that security incurs a negligible cost in terms of memory usage when the number of files and users are scaled up, in both centralized and decentralized scenarios. Also, we numerically show that by increasing the number of files and users, the secure coding delay of centralized and decentralized schemes became asymptotically equal.
BibTeX
@misc{sojdeh2022multi,
  author = {Mohammad Javad Sojdeh and Mehdi Letafati and Seyed Pooya Shariatpanahi and Babak Hossein Khalaj},
  title = {{Multi-Transmitter Coded Caching with Secure Delivery over Linear Networks - Extended Version}},
  year = {2022},
  eprint = {2211.14672},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2211.14672},
}
Figure