Optimal Control of Dynamic Defensive Maneuvers

This direction is split into other directions: security-performance trade-off (overhead of dynamic defense). Its last version stays readable and citable here.

Version 1 of 135 papers45 conceptsfrozen 2026-10-04

Traditional cybersecurity relies on a "patch-and-pray" model, where systems remain static until a vulnerability is discovered and fixed. However, static systems give attackers an asymmetric advantage: they have unlimited time to study a fixed target. To counter this, researchers have proposed Moving Target Defense (MTD) and Defensive Maneuvers, which involve constantly changing a system's configuration—such as IP addresses, software stacks, or network topologies—to increase uncertainty for the adversary.

The fundamental challenge of this approach is that "moving" is not free. Every defensive reconfiguration consumes computational resources, introduces latency, and risks disrupting legitimate services. Modern research focuses on the mathematical optimization of this trade-off: finding the exact frequency and type of maneuver that maximizes security while minimizing operational degradation.

The Cost of Movement

In a distributed system, every defensive action competes for the same resources as the system's primary mission. William Clay Moody and Amy Apon demonstrate this using Stochastic Petri Nets, showing a "boundary" where increasing security maneuvers directly cannibalizes computational throughput.

These costs manifest in several ways:

  • Migration Penalties: Switching between different web stacks or database engines requires data migration, which Sailik Sengupta and Subbarao Kambhampati identify as a major deterrent to frequent rotation.
  • Service Disruption: In real-time systems, security tasks must be scheduled carefully to avoid violating timing constraints. Monowar Hasan and Sibin Mohan propose the Contego framework to allow security tasks to run opportunistically, treating security as a tunable performance requirement rather than a fixed overhead.
  • Economic Impact: In critical infrastructure like power grids, changing line reactances to detect stealthy attacks moves the system away from its "optimal power flow," resulting in higher operational costs (Subhash Lakshminarayana and Elena Veronica Belmega).

Security as a Quantifiable Performance Metric

To manage these trade-offs, researchers like Alfred O. Hero and José M. F. Moura suggest treating security as a quantifiable performance metric, similar to latency or bandwidth. By doing so, the "security posture" of a system can be dynamically adjusted based on the current threat level.

For example, Michael Bloem and Tansu Alpcan use $H^\infty$-optimal control theory to automatically adjust firewall rules. Their framework balances the risk of infection against the "cost" of accidentally blocking legitimate traffic. If the threat is low, the system relaxes its filters to maximize performance; if an attack is detected, it tightens them, accepting the temporary performance hit.

Strategic Switching via Game Theory

If a defender moves too frequently, they bankrupt their own resources; if they move too slowly, they become an easy target. Finding the "equilibrium" requires modeling the interaction between defenders and attackers as a game.

Megha Bose and Akshat Kumar use Factored Markov Decision Processes (MDPs) to calculate defensive strategies that explicitly incorporate "switching costs" into the reward function. This prevents the system from making unnecessary movements that don't significantly reduce risk. Similarly, Chenyuan Wu and Dahlia Malkhi developed BFTBrain, which uses reinforcement learning to switch between different consensus protocols in real-time, optimizing for throughput while maintaining resilience against Byzantine faults.

The Threshold of Justification

A central theme across this research is the Threshold of Justification: at what point does the reduction in risk outweigh the overhead of a reconfiguration?

Amin Ghafouri and Yevgeniy Vorobeychik argue that in dynamic environments, security parameters should only be updated when the change in the environment is significant enough to justify the "operational reconfiguration cost." This prevents system instability caused by "chattering"—frequent, low-value changes that consume resources without providing meaningful protection.

By formalizing these costs, the research shifts cybersecurity from a purely defensive discipline to an optimization problem. The goal is no longer just to be "secure," but to be optimally secure—maintaining the highest possible level of protection that the system's mission and resources can sustain.

Papers in this version

History & citationVersion 1 · 2026-10-04

Split into other directions · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_optimal_control_of_dynamic_defensive_man_10b9f116,
  title = {Optimal Control of Dynamic Defensive Maneuvers},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/optimal-control-of-dynamic-defensive-maneuvers/rdr_10b9f1164f3d4bbd845c8036c0af0849},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_optimal_control_of_dynamic_defensive_man_10b9f116_v1,
  title = {Optimal Control of Dynamic Defensive Maneuvers},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/optimal-control-of-dynamic-defensive-maneuvers/rdr_10b9f1164f3d4bbd845c8036c0af0849/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}