Pure threshold strategies for a two-node tandem network under partial information

TL;DR

This paper investigates pure threshold strategies in a two-node tandem network where customers make joining decisions based on partial information, specifically when they are informed of the total number of users, and proves the existence of equilibrium policies.

Contribution

It demonstrates the existence of pure threshold equilibrium policies under partial information in a tandem queue network, extending understanding of strategic customer behavior.

Findings

Pure threshold equilibrium policies exist under partial information.

Customers informed of total network occupancy follow threshold-based joining strategies.

The analysis applies to scenarios where customers lack full system state knowledge.

Abstract

In a two node tandem network, customers decide to join or balk by maximizing a given profit function whose costs are proportional to the sojourn time they spend at each queue. Assuming that their choices are taken without knowing the complete state of the system, we show that a pure threshold equilibrium policy exists. In particular we analyze the case when the partial information consists in informing the arrival customers of the total number of users in the network.