We study a$K$-user coded-caching broadcast problem in a joint source-channel coding framework. The transmitter observes a database of files that are being generated at a certain rate per channel use, and each user has a cache, which can store a fixed fraction of the generated symbols. In the delivery phase, the transmitter broadcasts a message so that the users can decode their desired files using the received signal and their cache content. The communication between the transmitter and the receivers happens over a (deterministic) time-varying erasure broadcast channel, and the channel state information is only available to the users. We characterize the maximum achievable source rate for the 2-user and the degraded$K$-user problems. We provide an upper bound for any caching strategy’s achievable source rates. Finally, we present a linear programming formulation to show that the upper bound is not a sharp characterization. Closing the gap between the achievable rate and the optimum rate remains open.

A two-user coded caching problem is studied in a joint source-channel coding framework. A source generates symbols at a certain rate for each file in the database, and a fixed fraction of the symbols are cached at each user. The delivery phase of coded caching takes place over a time-varying erasure broadcast channel, where the channel state information is only available at the receivers. The maximum source rate to keep up with the ergodic rate of both users is characterized.

Coded caching is a new approach to decrease the communication load during the peak hours of the network. It provides a significant gain, that is maximized in the centralized setting, where the server controls the placement. In many situations, each user fills its cache without any information about the placement of other users. We show that subspace precoding for placement improves the delivery load of a decentralized caching system compared to uncoded placement. Surprisingly, the proposed scheme achieves the delivery load of the centralized placement for K = 3 users for the entire range of cache size.

Coded caching can significantly decrease the communication load in peak hours of the network. The gain of caching is maximized in a centralized setting, where the cache content of users are opportunistically designed. In the absence of a centralized placement, users' caches are filled with randomly selected packets of the files. This yields to a loss in the caching gain, especially for small cache size. A novel placement scheme is introduced in this work which is based on (within file) precoding of the files at the server, followed by random cache placement. It is shown that the proposed technique improves the caching gain compared to the uncoded placement. Surprisingly, the performance of the proposed decentralized placement matches with that of the centralized placement for small cache size.