Probabilistic Model Based Caching Strategy for Device-to-Device Communications

Abstract

Many existing literatures of device-to-device (D2D) communication aided caching technologies have focused on the probabilistic caching mechanism, while neglecting some practical attributes of the cellular system. Aiming at overcoming such a deficiency of existing schemes, we propose a sequential unconstrained minimization (SUM) based probabilistic cache (SPC) scheme, which takes into account some characteristics of the cellular system, such as the channel quality requirement, the proportion of caching user equipment (UE), UEs' distribution and cache space constraints. Then, we formulate a probabilistic cache problem with a goal to maximize the successful offloading probability (SOP). To solve it, we derive the closed-form expression of the approximated solution and apply the SPC algorithm to obtain the optimal solution. Finally, we analyze the performance of the SOP by simulations, which show that the proposed SPC scheme can achieve notable benefits in comparison to existing methods.