A Moving Target Defense Game with Random Searching Time
to play as the blue player.
Click the "Start" button to start the game.
Click the radio button in front of a state to stay in the same state or move to a new state.
You are playing as the blue player. The board will appear grey until you make the first move.
Rules of the Game
Basics
You play as the blue player (the defender) and the computer plays as the red player (attacker).
There is a single resource that is either controled by the blue player or the red player.
The resource is initially controlled by the blue player.
The blue player can move the resource to a new state at any time to take back the control (with a certain cost defined below).
The red player starts searching the true state once the game starts and whenever the blue player moves to a new state, and immediately gains the control of the resource once it pins down the true state.
Each searching takes a random amount of time and a cost that depends on an intensity level chosen by the red player (defined below).
If the blue player moves before the red player's searching completes, the red player is forced to start a new searching.
The game ends in 60 seconds.
States, gains, and costs
The game has four states (see the figure on the left).
Each player gains 100 points per second when the resource is under its control.
The moving costs of the blue player are given in the table on the right. Staying in the same state costs zero point. The state migration is instaneous.
The red player's searching time is an i.i.d. random variable sampled from an exponential distribution with a parameter λ, which is chosen by the red player and is unknown to the blue player.
The red player incurs a searching cost propotional to λ in each round (independent of the current state and the searching time).
Action spaces
The blue player determins when to move again and the next state of the resource in each round.
The red player picks the parameter λ in each round. In the current game, λ is chosen in the beginning of the game and is fixed afterwards.
How to win
The objective of each player is to gain as many points as possible.
To win the game, the blue player wants to be in control for as long as possible while moving only when necessary.
Hint to the blue player: try to learn the red player's searching time distribution and decide when and where to move accordingly.
The Board
The board displays the information that are currently known to the blue player during the game.
The blue (resp. red) rectangles represent the periods of time when the blue (resp. red) player is in control.
The latest moving time and attacking time are revealed to the blue player only when it moves to a new state. That is, the attacker remains stealthy until the defender makes a move.
About the Game
Our game can be viewed as an adaption and extension of the following two games:
The FlipIt game invented by Marten van Dijk, Ari Juels, Alina Oprea, and Ronald L. Rivest in their paper FLIPIT: The Game of “Stealthy Takeover” for modelling Advanced Persistent Threats (APTs).