Socially Motivated Energy-Efficient Wireless Caching Strategy With Saddle Point Approximation

Wireless caching has attracted substantial attention due to the potential for reducing service delays. However, it will waste energy if the cached content item is not requested by users. In order to improve the energy efficiency of wireless caching, we present a socially motivated wireless caching strategy from the cross-domain perspective in this paper. To this end, we first study the evolutionary requesting model based on users’ preference and their social ties. Then we derive the approximated expression for the delay-outage probability with a given power allocation strategy over Rayleigh block fading channels based on the saddle point technology. Based on the derived expression, a two-layer binary search method is presented to determine a suitable caching volume with a given outage probability and a power constraint. After that, we formulate a cross-layer optimization problem to jointly allocate the storage and transmission resources. By solving this optimization problem, we can determine the caching strategy to minimize the power consumption while promising the hit ratio performance. In addition, we extend the energy-efficient caching system over additive white Gaussian noise channels. Finally, numerical results are presented to demonstrate the potential of the presented caching strategy.