On the Fundamental Limits of Cache-Aided Multiuser Private Information Retrieval

We consider the problem of cache-aided Multiuser Private Information Retrieval (MuPIR) which is an extension of the single-user cache-aided PIR problem to the case of multiple users. In cache-aided MuPIR, each of the
$K_{\mathrm{ u}}$
cache-equipped users wishes to privately retrieve a message out of
$K$
messages from
$N$
databases each having access to the entire message library. Demand privacy requires that any individual database learns nothing about the demands of all users. The users are connected to each database via an error-free shared-link. In this paper, we aim to characterize the optimal trade-off between user cache memory and communication load for such systems. First, we propose a novel approach of
cache-aided interference alignment (CIA)
, for the MuPIR problem with
$K=2$
messages,
$K_{\mathrm{ u}}=2$
users and
$N\ge 2$
databases. The CIA approach is optimal when the cache placement is uncoded. For general cache placement, the CIA approach is optimal when
$N=2$
and 3 verified by the computer-aided converse approach. Second, for the general case, we propose a
product design
(PD) which incorporates the PIR code into the linear caching code. The product design is shown to be order optimal within a multiplicative factor of 8 and is exactly optimal in the high memory regime.