Cluster Forming of Multiagent Systems in Rolling Horizon Games with Non-uniform Horizons

TL;DR

This paper investigates how multiagent systems form clusters under jamming attacks and reactive defenses in repeated rolling horizon games with non-uniform horizons, highlighting the impact of strategic timing and computational abilities.

Contribution

It introduces a novel analysis of cluster formation in multiagent systems considering non-uniform horizons and strategic interactions in rolling horizon games.

Findings

Horizon lengths significantly influence cluster formation.

Different computational abilities affect strategic outcomes.

Simulation results confirm theoretical predictions.

Abstract

Consensus and cluster forming of multiagent systems in the face of jamming attacks along with reactive recovery actions by a defender are discussed. The attacker is capable to disable some of the edges of the network with the objective to divide the agents into a smaller size of clusters while, in response, the defender recovers some of the edges by increasing the transmission power. We consider repeated games where the resulting optimal strategies for the two players are derived in a rolling horizon fashion. The attacker and the defender possess different computational abilities to calculate their strategies. This aspect is represented by the non-uniform values of the horizon lengths and the game periods. Theoretical and simulation based results demonstrate the effects of the horizon lengths and the game periods on the agents' states.