The proliferation of low-cost, capable small unmanned aerial vehicles (UAVs) has introduced a disruptive and complex threat to modern air defense paradigms. The concept of drone swarms, or clusters, involves the coordinated use of numerous, often disposable, UAVs to overwhelm traditional defense systems through sheer numbers, low radar cross-sections, and distributed, intelligent behaviors. This presents a formidable challenge, particularly for the protection of high-value assets and critical infrastructure. In this context, integrated missile and gun air defense systems, often serving as the last line of defense in a layered protective shield, are thrust into the forefront of anti-drone warfare. This study delves into the operational capabilities of such combined arms systems against UAV swarm attacks, employing queuing theory to model the engagement dynamics, conducting simulation-based analysis, and proposing pathways for enhancing anti-drone effectiveness from both a technical and tactical perspective.
The fundamental challenge in anti-drone swarm defense lies in the mismatch between the high cost and limited magazine depth of defensive systems and the low cost and high quantity of attacking drones. A swarm may consist of dozens or even hundreds of UAVs, arriving in a coordinated yet temporally random fashion. Traditional air defense metrics focused on engaging single high-value targets are insufficient. The evaluation must shift towards assessing the system’s capacity to manage a sustained, high-rate engagement sequence and its probability of preventing swarm penetration over a defined period and defensive sector. The integrated system, typically combining a rapid-fire cannon (e.g., 30-40mm) with a battery of short to medium-range missiles, leverages the strengths of both: the gun provides a high-rate, cost-effective solution for close-in threats, while missiles extend the engagement envelope and handle higher-priority or more maneuvering targets at longer ranges. The synergy is key for an effective anti-drone strategy.
Methodological Foundation: A Queuing Theory Model for Swarm Engagements
To analytically capture the stochastic nature of a swarm attack and the serial service process of a defense system, queuing theory provides a robust mathematical framework. We model the airspace sector defended by an integrated system as a service facility. The invading drone swarm constitutes the arrival stream of “customers,” while the firing units (missile launchers and gun turrets) are the “servers.” Crucially, this is a queue with finite patience: a drone that spends too much time waiting in the engagement zone (or being engaged) without being destroyed is considered to have penetrated the defense.
We make the following foundational assumptions for the anti-drone engagement model:
- The drone swarm attacks over a mission time $T_H$. The inter-arrival times of individual drones follow an exponential distribution, resulting in a Poisson arrival process with mean rate $\lambda$ (drones per unit time).
- The service time for the defense system, defined as the total time required to complete target acquisition, tracking, firing, and assessing engagement of one drone, is also a random variable. It follows an exponential distribution with a mean service rate $\mu = 1 / \bar{T}_{sj}$, where $\bar{T}_{sj}$ is the average service (or shot) cycle time.
- A drone will only remain in the weapon system’s lethal engagement zone for a limited, random time $T_{zl}$, following an exponential distribution with parameter $\nu = 1 / \bar{T}_{zl}$, where $\bar{T}_{zl}$ is the average dwell time in the kill zone. If not destroyed within this time, it exits the zone and penetrates.
Let the integrated defense system have $s$ parallel firing channels (e.g., $s$ missile launchers, with the gun considered as a separate but overlapping service channel). The state of the system, $n$, represents the number of drones currently being engaged or waiting to be engaged within the zone. The system transitions between states based on arrivals and service completions. This can be described as a birth-death process. For states where all servers are busy ($n \ge s$), the effective “service” rate for drones in the queue is influenced by their limited dwell time $\nu$. The steady-state balance equations, derived from the Markov chain, are as follows:
For $0 \le n < s$:
$$ \lambda P_n = (n+1)\mu P_{n+1} $$
For $n \ge s$:
$$ \lambda P_n = [s\mu + (n+1-s)\nu] P_{n+1} $$
With the normalization condition:
$$ \sum_{n=0}^{\infty} P_n = 1 $$
where $P_n$ is the steady-state probability of having $n$ drones in the system (in service and in queue).
Solving this system yields the state probabilities. Let $\rho = \lambda / \mu$ and $\sigma = \nu / \mu$. The probability that the system is idle is:
$$ P_0 = \left[ \sum_{n=0}^{s-1} \frac{\rho^n}{n!} + \frac{\rho^s}{(s-\rho)(s-1)!} \right]^{-1} \quad \text{(for $\rho < s$)} $$
The probabilities for other states are:
$$ P_n = P_0 \frac{\rho^n}{n!}, \quad \text{for } n = 1, 2, …, s $$
$$ P_{s+j} = P_0 \frac{\rho^s}{s!} \left(\frac{\rho}{s}\right)^j, \quad \text{for } j = 1, 2, … $$
Modeling the Integrated Missile-Gun System for Anti-Drone Operations
The integrated system features two distinct but coordinated weapon systems: missiles and a gun. They share common surveillance and tracking sensors but have independent firing channels and lethal zones. The gun’s zone is nested inside the missile’s longer-range zone. A drone must survive both layers to penetrate. Therefore, we model two sequential, overlapping queuing systems. The engagement process is illustrated conceptually below, where a drone first enters the missile zone and, if not destroyed, proceeds into the gun zone.
Let the subscript $D$ denote parameters for the missile system and $G$ for the gun system. The probability that a drone is intercepted by missiles is not simply the kill probability of one missile. It depends on the system’s ability to service the drone before it leaves the missile zone. We define an anti-drone interception probability for the missile layer, $P_{D,X}$, which accounts for queueing delays and the single-shot kill probability $P_{D,HJ}$:
$$ P_{D,X} = 1 – \exp\left( -\nu_D \cdot \bar{T}_{D,eff} \right) $$
where $\nu_D$ is the exit rate from the missile zone, and $\bar{T}_{D,eff}$ is the effective mean time available for missile engagement, which is influenced by the queue state and the probability of kill per engagement attempt. A detailed derivation leads to a formula involving the sum of probabilities for states where drones are waiting:
$$ P_{D,X} \approx 1 – \frac{\nu_D}{\lambda \cdot (-\ln(1-P_{D,HJ}))} \sum_{j=1}^{\infty} j \cdot P_{D, s+j} $$
A similar equation holds for the gun layer interception probability $P_{G,X}$, with its parameters $\nu_G$, $P_{G,HJ}$, and state probabilities $P_{G, s+j}$.
The overall probability that a drone from the swarm successfully penetrates the entire integrated anti-drone defense is:
$$ P_{AN} = (1 – P_{D,X}) \cdot (1 – P_{G,X}) $$
The operational capacity of the integrated system, a key metric for anti-drone swarm defense, is the expected number of drones killed during the swarm attack period $T_H$. If $N_a = \lambda T_H$ is the expected total number of attacking drones, the system’s kill capacity $N_{ca}$ is:
$$ N_{ca} = N_a \cdot (1 – P_{AN}) = \lambda T_H (P_{D,X} + P_{G,X} – P_{D,X}P_{G,X}) $$
Simulation Analysis and Anti-Drone Capability Insights
We conducted extensive simulation studies using the queuing model to evaluate the sensitivity of the integrated system’s anti-drone performance to key technical parameters. The baseline scenario assumes a swarm of 100-200 small UAVs attacking over 30 minutes ($T_H=1800$ s). The defense system has $s=6$ missile firing channels. The gun is treated as a separate server with a very high service rate due to its rapid fire capability.
1. Impact of Missile and Gun Engagement Range
The primary effect of increasing missile range is to enlarge the defended area, thereby increasing the average dwell time $\bar{T}_{zl}$ (decreasing $\nu$) for drones in the missile zone. This gives the system more time to engage each target. We simulated with missile ranges of 5km, 10km, and 15km, with corresponding adjustments to dwell time.
| Simulation Parameter | Missile Range = 5 km | Missile Range = 10 km | Missile Range = 15 km |
|---|---|---|---|
| Missile Interception Prob. $P_{D,X}$ | 62.5% | 62.9% | 64.4% |
| Gun Interception Prob. $P_{G,X}$ | 97.3% | 97.3% | 97.3% |
| Swarm Penetration Prob. $P_{AN}$ | 1.02% | 1.01% | 0.97% |
| System Kill Capacity $N_{ca}$ (out of 100) | 98.98 | 98.99 | 99.03 |
The results show diminishing returns for increased missile range in this anti-drone context. The gun layer is overwhelmingly responsible for the final interception, thanks to its high service rate. This suggests that for cost-effective anti-drone swarm defense, optimizing the gun component is paramount. A missile range of 10km offers a balanced solution, providing a meaningful engagement layer without the excessive cost associated with longer-range missiles better suited for larger aircraft.
2. Impact of Gun Kill Probability (PK)
The single-shot kill probability of the gun, $P_{G,HJ}$, is a critical factor. It depends on ammunition type (e.g., high-explosive vs. programmable airburst), fire control accuracy, and target vulnerability. We analyzed the effect of varying $P_{G,HJ}$ against a larger swarm of 200 drones.
| Simulation Parameter | $P_{G,HJ}$ = 40% | $P_{G,HJ}$ = 60% | $P_{G,HJ}$ = 80% |
|---|---|---|---|
| Missile Interception Prob. $P_{D,X}$ | 59.7% | 59.7% | 59.7% |
| Gun Interception Prob. $P_{G,X}$ | 73.3% | 95.2% | 99.7% |
| Swarm Penetration Prob. $P_{AN}$ | 10.78% | 1.94% | 0.12% |
| System Kill Capacity $N_{ca}$ (out of 200) | 178.4 | 196.1 | 199.8 |
The improvement is dramatic. Increasing the gun’s lethality from 40% to 60% reduces swarm penetration by nearly 80%. This strongly advocates for investment in advanced ammunition (like time-fuzed or electronically fused airburst rounds) and high-precision tracking systems specifically designed for small, agile anti-drone targets. This is arguably the most direct path to enhancing anti-drone swarm capability.
3. Impact of System Reaction and Service Time
The average service cycle time $\bar{T}_{sj}$ encapsulates the total system reaction time: detection-to-track latency, fire control solution computation, slew-and-point time, and the time of flight/sustained fire needed for a kill. Reducing this time increases the service rate $\mu$, allowing the system to handle a higher arrival rate $\lambda$. We first analyze shortening the gun’s cycle time.
| Simulation Parameter | Gun Cycle $\bar{T}_{G,sj}$ = 10s | Gun Cycle $\bar{T}_{G,sj}$ = 8s | Gun Cycle $\bar{T}_{G,sj}$ = 6s |
|---|---|---|---|
| Missile Interception Prob. $P_{D,X}$ | 59.7% | 59.7% | 59.7% |
| Gun Interception Prob. $P_{G,X}$ | 73.3% | 78.5% | 94.2% |
| Swarm Penetration Prob. $P_{AN}$ | 10.78% | 8.65% | 2.34% |
| System Kill Capacity $N_{ca}$ (out of 200) | 178.4 | 182.7 | 195.3 |
Similarly, we analyze the effect of improving the missile system’s reload or engagement cycle time, keeping the gun cycle at 8s.
| Simulation Parameter | Missile Cycle $\bar{T}_{D,sj}$ = 60s | Missile Cycle $\bar{T}_{D,sj}$ = 45s | Missile Cycle $\bar{T}_{D,sj}$ = 30s |
|---|---|---|---|
| Missile Interception Prob. $P_{D,X}$ | 59.7% | 61.0% | 67.6% |
| Gun Interception Prob. $P_{G,X}$ | 78.5% | 78.5% | 78.5% |
| Swarm Penetration Prob. $P_{AN}$ | 8.65% | 8.37% | 6.94% |
| System Kill Capacity $N_{ca}$ (out of 200) | 182.7 | 183.3 | 186.1 |
The analysis confirms that reducing service time, especially for the gun, significantly boosts anti-drone performance. Improvements come from faster sensors, more agile turret drives, automated threat assessment, and, for missiles, vertical launch systems or rapid reload mechanisms. However, the missile cycle time improvement shows less relative impact, reinforcing the notion that the gun is the workhorse in dense anti-drone swarm scenarios.
Synthesis and Recommendations for Enhanced Anti-Drone Swarm Defense
Based on the queuing theory modeling and simulation analysis, several key conclusions and recommendations emerge for optimizing integrated missile-gun systems for the anti-drone swarm mission:
- Prioritize the Gun System: The simulation results consistently show that the rapid-fire gun is the most critical component for defeating dense UAV swarms. Its high service rate ($\mu_G$) makes it the primary layer for attrition. Therefore, resource allocation should focus on maximizing gun performance.
- Optimize for Lethality and Speed: The two most effective technical improvements are increasing the gun’s single-shot kill probability $P_{G,HJ}$ (via advanced ammunition) and decreasing its total engagement cycle time $\bar{T}_{G,sj}$ (via automation and agility). The mathematical model quantifies the disproportionate payoff of these enhancements on the overall anti-drone swarm kill capacity $N_{ca}$.
- Right-Size the Missile Component: Missiles are essential for extending the defended footprint and engaging higher-priority or longer-range elements of a mixed threat. However, against pure low-cost swarms, extreme range is not cost-effective. A focus on improving missile system readiness (reducing $\bar{T}_{D,sj}$) and ensuring seamless integration with the gun’s fire control for efficient target handover is more valuable than simply extending maximum range.
- Tactical Deployment Implications: The model supports tactical concepts such as “gun-centric” engagement policies for confirmed drone swarms. It also highlights the importance of sensor performance against small, low-flying targets; effective detection and tracking at maximum range is a prerequisite that underpins all the queuing dynamics modeled. Networked sensors providing early warning can effectively increase the dwell time parameter $\bar{T}_{zl}$, improving interception probabilities for both layers.
In conclusion, the threat of drone swarms necessitates a reevaluation of terminal air defense design and tactics. By applying stochastic models like queuing theory, the complex interaction between swarm offense and integrated defense can be analyzed systematically. This analysis demonstrates that a well-optimized, gun-heavy integrated system represents a robust and cost-effective solution for the pressing anti-drone challenge. The insights gained provide a quantitative foundation for setting technical performance requirements during system design and for developing effective tactical doctrines for counter-swarm operations in modern conflicts.