Chapter 20: Online Coded Caching

Learning Objectives

• Distinguish offline (static) vs online (dynamic) coded caching
• Define adversarial and stochastic models for demand
• State regret bounds for online coded caching: $O(\sqrt{T \log K})$
• Understand Follow-the-Perturbed-Leader (FTPL) style algorithms for caching
• Compare LRU/LFU with coded online caching on dynamic workloads
• Analyze learning under non-stationary demand (concept drift)