Fundamental Limits of Coded Caching in Request-Robust D2D Communication Networks.

Area of Science:

• Wireless Communication
• Information Theory
• Network Coding

Background:

• Device-to-Device (D2D) coded caching enhances communication efficiency.
• Prior D2D caching models assume all users request content simultaneously, which is unrealistic.
• Real-world scenarios involve a subset of users requesting content, necessitating robust caching strategies.

Purpose of the Study:

• To address the limitations of existing D2D coded caching by introducing the "request-robust D2D coded caching" problem.
• To minimize average and worst-case delivery rates in D2D networks with partial user requests.
• To develop and analyze caching schemes that are resilient to varying user demands.

Main Methods:

• Formulation of the request-robust D2D coded caching problem with K users and N files, where only r users are requesters.
• Proposal of a novel scheme based on uncoded cache placement, exploiting common demands and one-shot delivery.
• Derivation of information-theoretic converse bounds under uncoded cache placement.
• Adaptation and performance analysis of existing schemes (Yapar et al.) for the request-robust scenario.

Main Results:

• A new caching scheme is proposed for request-robust D2D coded caching.
• The adapted scheme by Yapar et al. is shown to be order optimal within a factor of two (uncoded placement) and four (general case).
• Numerical evaluations demonstrate the proposed scheme's superiority over existing D2D caching methods in request-robust settings for most cache sizes.

Conclusions:

• The proposed scheme effectively addresses the request-robust D2D coded caching problem.
• The developed scheme offers significant performance improvements in terms of delivery rates compared to existing methods.
• This work advances D2D caching by providing a practical solution for scenarios with non-simultaneous user demands.