Skip to Main Content
In wireless networks with energy limited nodes, user cooperation is usually exploited to reduce the network energy consumption. In many practical scenarios, however, nodes' selfishness raises doubts on whether each node will be willing to spend its valuable energy in forwarding packets for other users. To analyze this problem, a non-cooperative game theoretic framework is adopted in our work. Using this framework, the critical role of altruistic nodes in encouraging cooperation is established, both for small and large scale networks. In a small network, where nodes utilize the Decode-Forward scheme to cooperate, we show that a relay node, with appropriate strategy and location, successfully turns the Nash Equilibrium from no- cooperation to full-cooperation. In the large scale network, we show that it is sufficient to have a vanishingly small fraction of the nodes to be altruistic, i.e., relay nodes, in order to ensure full cooperation from all the nodes in the network. This result hinges on using the appropriate forwarding policies by the altruistic nodes, as detailed in the sequel. Our work also establishes the sub-optimality of traditional relaying strategies, which ignore the game-theoretic aspect of the problem. An important aspect of our work is that only reward/punishment policies that can be realized on the physical layer are used, and hence, our results establish the achievability of full cooperation without requiring additional incentive mechanisms at the application layer.