Analysis of Secret Key Agreement Protocol for Massive MIMO Systems

Abstract

In this paper, we propose and analyze a secret key agreement (SKA) protocol designed for multiple-input multiple-output (MIMO) scenarios, in which a legitimate transmitter estimates the amount of information that can be leaked to potentially hazardous legitimate receivers in the network, generates a secret key, and transmits the generated key to a specific legitimate receiver. For the latter key transmission step, a new beamforming method is proposed that regards other legitimate receivers in the network as untrustworthy nodes and show the attack algorithm they can perform. The proposed protocol requires only one-way transmission and no mutual interaction to share the secret key. Through numerical simulations, we evaluate two scenarios for malicious users, where a part of them are other legitimate receivers. Namely, we evaluate the information leakage risk of our model under two conditions: a disadvantage for them, where the malicious user is unaware of the existence of beamforming, and an advantage, where the user can jam directly to the beamformer. Finally, to demonstrate the practicality of the protocol, we demonstrate that the beamforming advantage dramatically reduces the computational complexity requirement at the legitimate receiver.