On the Effect of Correlation in IRS-Aided SWIPT Networks

In this work, we focus on the effect of correlation between the elements of an intelligent reflecting surface, in the context of simultaneous wireless information and power transfer (SWIFT). The two extreme cases are considered, that is, when the elements are uncorrelated (independent) and when the elements are billy correlated. For each case, we study the performance of employing random phase shifts as well as with no phase shifts. Closed-form analytical expression are derived for the outage probability and the average harvested energy. When the elements are uncorrelated, random phase shifts provide the same performance as having no phase shifts. However, we show that the performance of correlated elements attains significant gains when there are no phase shifts and exceeds the performance of the uncorrelated case.