Abstract:Integrating coded caching (CC) into multiple-input multiple-output (MIMO) communications may significantly enhance the achievable degrees of freedom (DoF) of the wireless networks. In this paper, we consider a cache-aided MIMO configuration with a CC gain $t$, where a server with $L$ Tx antennas communicates with $K$ users, each with $G$ Rx antennas. In the proposed content-aware MIMO strategy, we carefully adjust the number of users $\Omega$ and the number of parallel streams decoded by each user $\beta$ served in each transmission to maximize the DoF. As a result, we achieve a DoF of ${\max_{\beta, \Omega }}{\Omega \beta}$, where ${\beta \le \mathrm{min}\big(G,\frac{L \binom{\Omega-1}{t}}{1 + (\Omega - t-1)\binom{\Omega-1}{t}}\big)}$. To prove the achievability of the proposed DoF bound, we provide a novel transmission strategy based on the simultaneous unicasting of multiple data streams. In this strategy, the missing data packets are scheduled such that the number of parallel streams per transmission is maximized while the decodability of all useful terms by each target user is guaranteed. Numerical simulations validate the findings, confirming the enhanced DoF and improved performance of the proposed design.
Abstract:A practical and scalable multicast beamformer design in multi-input multi-output~(MIMO) coded caching~(CC) systems is introduced in this paper. The proposed approach allows multicast transmission to multiple groups with partially overlapping user sets using receiver dimensions to distinguish between different group-specific streams. Additionally, it provides flexibility in accommodating various parameter configurations of the MIMO-CC setup and overcomes practical limitations, such as the requirement to use successive interference cancellation~(SIC) at the receiver, while achieving the same degrees-of-freedom~(DoF). To evaluate the proposed scheme, we define the symmetric rate as the sum rate of the partially overlapping streams received per user, comprising a linear multistream multicast transmission vector and the linear minimum mean square error~(LMMSE) receiver. The resulting non-convex symmetric rate maximization problem is solved using alternative optimization and successive convex approximation~(SCA). Moreover, a fast iterative Lagrangian-based algorithm is developed, significantly reducing the computational overhead compared to previous designs. The effectiveness of our proposed method is demonstrated by extensive simulations.
Abstract:A location-aware coded caching scheme is introduced for applications with location-dependent data requests. An example of such an application is a wireless immersive experience, where users are immersed in a three-dimensional virtual world and their viewpoint varies as they move within the application area. As the wireless connectivity condition of the users also varies with their location due to small- and large-scale fading, a non-uniform memory allocation process is used to avoid excessive delivery time in the bottleneck areas. Then, a well-defined location-aware placement and delivery array (LAPDA) is used for data delivery to utilize unicast transmission with a fast converging, iterative linear beamforming process. An underlying weighted max-min transmit precoder design enables the proposed scheme to serve users in poor connectivity areas with smaller amounts of data while simultaneously delivering larger amounts to other users. Unlike previous studies in the literature, our new scheme is not constrained by the number of users or network parameters (users' cache capacity, number of antennas at the transmitter, etc.) and is suitable for large networks due to its linear transceiver structure. Despite non-uniform cache placement, the proposed scheme achieves a coded caching gain that is additive to the multiplexing gain and outperforms conventional symmetric CC schemes with only a moderate degree of freedom (DoF) loss.
Abstract:Efficient usage of in-device storage and computation capabilities are key solutions to support data-intensive applications such as immersive digital experience. This paper proposes a location-dependent multi-antenna coded caching -based content delivery scheme tailored specifically for wireless immersive viewing applications. First, a memory assignment phase is performed where the content relevant to the identified wireless bottleneck areas are incentivized. As a result, unequal fractions of location-dependent multimedia content are cached at each user. Then, a novel packet generation process is carried out given asymmetric cache placement. During the subsequent delivery phase, the number of packets transmitted to each user is the same, while the sizes of the packets are proportional to the corresponding location-dependent cache ratios. Finally, each user is served with location-specific content using joint multicast beamforming and multi-rate modulation scheme that simultaneously benefits from global caching and spatial multiplexing gains. Numerical experiments and mathematical analysis demonstrate significant performance gains compared to the state-of-the-art.