Abstract
Device-to-Device (D2D) computation offloading, or D2D offloading, exploits spare computing resources of nearby user devices to enhance mobile computing performance. Its success relies on user participation in costly collaborative service provisioning, thus mandating an incentive mechanism that can compensate for these costs. Although incentive mechanism design has been studied extensively in the literature, this paper considers a more challenging yet less investigated problem in which selfish users are also facing interdependent security risks that depend on the collective behavior of all users. To this end, we build a novel mathematical framework by combining the power of game theory and epidemic theory to investigate the interplay between user incentives and interdependent security risks in D2D offloading, thereby enabling the design of security-aware incentive mechanisms. Our analysis discovers an interesting "less is more" phenomenon: although giving users more incentives promotes more participation, it may harm the network operator's utility. This is because too much participation may foster persistent security risks and as a result, the effective participation level does not improve.
