Kai%20Virtanen,%20Raimo%20P.%20H

1
Team Optimal Signaling Strategies in Air Combat

• Kai Virtanen, Raimo P. Hämäläinen and Ville
  Mattila
• Systems Analysis Laboratory
• Helsinki University of Technology

2
Outline

• Signaling game
• The signaling system
• Signaling in the team
• The message prioritization problem
• Importance index
• Age of information
• Own plane information
• Uncertainty in signaling
• Simulation environment
• Conclusions

3
Signaling game

• M-on-N air combat
• Pilots decision making based on information
  received from
• visual detection, radio communication, radar
  measurements
• a signaling system
• Data Link Electronic counter-countermeasure
  Computer (DLEC)
• developed by Finnish Air Force and Finnish
  industry team
• for transmitting messages under jamming
• for encoding and decoding signals

4
Signaling in the team
- Opponents state - Commands
- Opponents state - Own state
J A M M I N G
- Own state
- Battle manager - Overall situation assessment
- Ground measurements
5
Message prioritization problem

• The signaling system aims at helping in the
  situation assessment
• Fixed transmission capacity
• gt The most important information must be
  transmitted gt A decision problem
• Messages ranked based on the positions,
  velocities and accelerations of planes

Select the most important message with respect to
the overall goals of the team! Alternative
messages - States of the opponents observed by
radar - Own state
6
Prioritization environment
Decision at tDt
Decision at t
Alternative messages - Opponent 1 - Opponent 2 -
Opponent 3 - Own state Information - Radar
observations at tDt - Transmitted messages at
t, t-Dt, ...
Alternative messages - Opponent 1 - Opponent 2 -
Own state Information - Radar observations at
t - Transmitted messages at t-Dt, t-2Dt, ...
Opponent 1
Opponent 2
Opponent 3
Decision at t2Dt
Information - Transmitted messages at tDt,
t, ...
Alternatives - Own state data
7
Terminology

• Importance Importance of an individual opponent
  with respect to an individual member of own team
• Importance index Importance of the individual
  opponent with respect to own team
• Selection of the message based on the importance
  index

Own team
Opponents
Importance indices
Importances
Opponents
Own team




8
Index of importance

• A decision analytic value function model
• Attributes based on information available in the
  signaling system

Importance
Distance x1 Angle x2 Velocity xn
Importance of distance v1(x1) Importance of
angle v2(x2) Importance of velocity
vn(xn)

• vk and wk depend on the preferences and goals of
  a pilot
• Index
• Weighted sum of importances
• Maximum importance

9
Age of information

• Quality of relayed information decreases in time
• No confirmation mechanism
• Important messages must be retransmitted
  frequently
• Quality of information ensured
• Success of transmissions ensured
• Time elapsed since the previous transmission
  related to a given plane (age of previous
  message)
• Elapsed time used as a multiplicative attribute
• Importance

Þ
10
Own plane information
A assumes that B is an opponent gt a message is
needed
Team members must be aware of each others
constantly
Þ
B
A
A should send its own state information since B
must know who A is

• Importance index
• Moving average of the importance indices of
  previous transmissions
• Time factor as before

11
Uncertainty in prioritization

• Uncertainty in the attributes
• Inaccurate state variables due to radar errors
• Age of information in the signaling system
• Uncertainty about the weights
• Incomplete preference statements
• Different preferences within a group
• Interval analysis a new method to incorporate
  uncertainty

12
Interval importance index

• Signals error indices are intervals gt Interval
  attributes
• Group or incomplete preference assessment gt
  Interval weights
• The value function is minimized and maximized
  s.t. the attribute and weight intervals gt
  Importance intervals
• Interval importance index based on aggregation
  criteria

Importance intervals
Importance intervals
Own team
Opponents

Importance index


The final selection based on interval importance
indices
13
Information Prioritization in Air Combat (IPAC)

• Matlab-based simulation tool
• M-on-N air combat simulation
• Planes fly given trajectories or follow a
  feedback guidance law
• Radar field assumed to be a cone
• The signaling system model transmits messages
• Prioritization methods
• Importance index, interval importance index,
  range rate/range,
• Construction of the value function
• Visualization of signaling and trajectories

14
Case 4 vs. 4 combat
Settings - number of aircraft - simulation
time - transmission time - radars scanning
angle and range - etc.
15
Initial combat state
team members
opponents
16
Simulated solution
signaling team member
opponent whose info is relayed
17
Results
18
Conclusions

• The decision analytical prioritization model
• The value function captures preferences of pilots
• Uncertainties treated by interval techniques
• Practical use
• True implementation of the signaling system
• Decision making module of air combat simulators
• Evaluation by pilots
• gt This new model improves our situation
  assessment capability
• Future research
• Signaling model incorporated into an air combat
  game
• gt Impact of information structure on the outcome
  of the game