Abstract
Recent localized conflicts have shown that autonomously coordinated drone swarms will be a primary approach for unmanned operations in denied environments. To improve secure communications for drone swarms operating under denial conditions, this article first divides mission phases for swarms in such environments. Based on the network topologies present at each phase, it summarizes the communication security issues facing swarm networks and proposes corresponding protection techniques and strategies as a reference for securing drone swarm communications in denied environments.
Introduction
In the 2022 Russia–Ukraine conflict, Ukraine used unmanned combat assets, including a formation of nine aerial drones and seven unmanned surface vehicles, to strike the Russian naval base at Sevastopol, which had area-denial defenses. The operation demonstrated that modern warfare is accelerating toward informationization, intelligence, and autonomous unmanned systems. Autonomous, coordinated unmanned swarms are likely to provide significant operational capability on future battlefields.
Advances in defensive systems and concealment/jamming techniques have reduced battlefield transparency. Gray and black-zone environments are increasingly common. This article reviews the main communication challenges for drone swarms in denied environments, divides mission phases based on typical mission scenarios, provides representative network topologies, analyzes communication security issues at each phase, and proposes mitigation strategies to help secure swarm communications in denial scenarios.
1. Background and Problem Analysis
The concept of an anti-access/area-denial environment (A2/AD) was formally defined by the U.S. Department of Defense in a policy report as an environment that hinders or denies force projection across multiple domains and creates strong localized opposition to external intervention. In such environments, high-intensity electromagnetic jamming, firepower suppression, and cyberattacks can cause node failures, degraded communications, or complete communication loss for drone swarms, creating significant security risks.
Common attacks on drone swarm communication networks include link-level eavesdropping, jamming, message tampering, man-in-the-middle attacks, replay attacks, and Sybil attacks. Node-targeted attacks include backdoors, denial-of-service, and flooding. Network-level attacks include distributed denial-of-service (DDoS), flooding, Sybil attacks, replay attacks, exhaustion attacks, and routing loops. To mitigate physical capture and tampering, researchers have proposed lightweight key authentication and key agreement schemes based on physically unclonable functions (PUFs) to establish session keys among drones and command centers with reduced computational overhead. Other work has proposed pairing-free group key distribution, AEAD/EC/Hash-based authentication protocols for the Internet of Drones, and ECDH/ECDSA-based key authentication protocols combined with HMAC for dual authentication. While these schemes improve general swarm network security, they may not be directly suitable for the constraints and dynamics of denied environments.
In denied environments, swarm data transmission relies on the swarm communication network. The next section divides typical mission phases for swarms in such environments and presents the network topologies corresponding to each phase.
2. Mission Phases and Network Topologies in Denied Environments
Compared with conventional mission environments, denied environments have these characteristics: (1) high-intensity electromagnetic interference and electronic attack, reducing communication reliability; (2) dual denial across geographic and physical domains, making long-distance transport prominent; (3) joint operations across sea, land, and air, increasing the breadth and depth of multi-service collaboration; (4) greater risk to manned forces. Based on these characteristics, mission tasks in denied environments can be divided into three phases: multi-platform transport and delivery, self-organizing flight approach, and autonomous coordinated mission execution. The swarm network adopts hybrid and distributed topologies across these phases.
2.1 Multi-Platform Transport and Delivery Phase
Future combat trends indicate that autonomous, coordinated swarm operations will be a key mode of contested-environment operations. Multiple drone combat groups are transported and deployed from air, sea, or land platforms to form a swarm capable of completing coordinated tasks. Transport platforms include large motherships, strategic transport aircraft, or land-based launch vehicles.
Conventional swarm communication networks typically include satellites, ground control centers, and multiple drone combat units in a command-and-control network. To maximize operational effectiveness in denied environments, reduce swarm energy consumption, and avoid latency introduced by multi-tiered command, the command center can be placed on the transport/delivery platform to plan and direct swarm actions. Before deployment, communication parameters should be preloaded into mission drones to ensure secure information exchange in denied environments. The network topology for this phase is a hybrid topology, where the transport platform outside the denied region serves as a central node controlling flight and communications. The central node and drone nodes form a centralized structure, while central nodes interconnect in a distributed structure. Drones delivered by the same platform form a distributed subgroup topology for intra-group information exchange.
2.2 Self-Organizing Flight Approach Phase
In denied environments, different platforms may deploy heterogeneous drones with varying capabilities. These drones are dynamically grouped according to mission needs to form functionally capable heterogeneous swarms.
Building on the hybrid topology from phase one, heterogeneous drones from different delivery platforms reconstruct into new, functionally viable heterogeneous swarms based on mission requirements and on-board capabilities. Network topology among drone nodes becomes distributed during this phase.
2.3 Autonomous Coordinated Mission Phase
Swarm datalinks are typically divided into payload communications and non-payload communications. Both link types rely on wireless communications and thus require stable and secure channels. In denied environments, intense and complex electronic attacks can degrade or sever links between command centers and drones. As a result, swarms must autonomously coordinate to complete assigned missions based on payloads and mission parameters.
The network topology in this phase continues as a distributed structure. Reconstructed swarms entering denied areas will, due to strong electromagnetic interference, lose communication with transport-platform-based command centers and must use distributed topologies for intra-swarm information exchange and autonomous coordination to complete missions.
3. Communication Security Issues for Drone Swarms in Denied Environments
As autonomy increases, multiple drone groups form mission swarms to conduct complex tasks in denied environments through coordinated search, target tracking, and strike missions. However, wireless swarm networks inherently use broadcast/multicast communication, making them vulnerable to attacks that can limit operational effectiveness. The following analyzes key threats faced by swarm communication networks in denied environments.
3.1 Link Security
Attacks on swarm communication links include maritime command-and-control links, aerial command-and-control links, and intra-swarm links. Passive attacks such as eavesdropping can capture sensitive mission and targeting data, enabling traffic analysis to locate command centers for precision strikes. Active attacks include radio-frequency jamming and path manipulation, which can tamper with control commands or data to deceive drones into accepting false data or losing data, causing incorrect threat assessments, wrong commands or flight paths, and mission failure. Common techniques include eavesdropping, message tampering, man-in-the-middle attacks, wormhole attacks, and satellite navigation signal spoofing.
3.2 Node Security
The most prominent denial factor is strong electromagnetic attack, with electromagnetic pulses being a primary means. Strong pulses can cover drone onboard operating bands, increase internal noise, or introduce interference, degrading or disabling equipment and causing single or multiple node failures. For example, pulses may irreversibly damage onboard data transceivers, removing communication capability and causing drones to descend or crash, impacting the integrity of the swarm communication network.
3.3 Network-Level Security
Network-level attacks in denied environments resemble traditional network attacks, including DDoS, flooding, Sybil attacks, replay attacks, exhaustion attacks, and routing loop attacks. Attackers can exploit constrained resources in denied environments by fabricating or controlling multiple malicious nodes to flood the swarm network with redundant data, causing communication congestion and exhausting network or compute resources. This prevents autonomous coordination and can neutralize the swarm's combat capability.
4. Countermeasures for Secure Swarm Communication in Denied Environments
Because swarm communications rely primarily on wireless links, their openness introduces many security challenges. The following countermeasures address link security, node security, and network information security for swarms operating in denied environments.
4.1 Encrypted Transmission to Secure Links
Depending on mission needs, swarm links can be separated into command-and-control links and data links. Command links handle control orders and emergency communications, while data links carry routine sensing data. Channel separation can reduce interference risks. Encrypting both data and command links ensures that any network information traversing a drone node is decrypted and re-encrypted by nodes equipped with cryptographic modules (software- or hardware-based). Link encryption protects against tampering and eavesdropping.
4.2 Distributed Node Authentication and Autonomous Coordination
Distributed swarms operate autonomously without fixed central nodes to resist single-point failures. During the transport phase, the command center on the transport platform should provision mission drones with secure initial parameters (for example, public/private key pairs, PUF-based secrets) via secure channels. Identity authentication methods such as one-time passwords or digital signatures should ensure each network node is legitimate, improving node-level security.
4.3 Holistic Key Management to Strengthen System Defenses
Design an integrated key and security parameter management scheme that reflects differing data exchange requirements. Allocate public/private keys and pre-provisioned authentication parameters appropriately, and use encryption at the data-link layer to protect information in transit and at rest. At the node level, identity-based authentication (for example, ID-based credentials) should ensure only legitimate drones participate in swarm exchanges to prevent spoofed-node attacks. Optimize key management to support key updates when nodes join or leave, or when subgroups merge or split, ensuring forward and backward secrecy. Select lightweight but robust key-management solutions to enhance swarm survivability and defensive capability.
Conclusion
Future battlefields will be highly contested, complex, dynamic, and characterized by strong denial measures. This article divided drone swarm missions in denied environments into distinct phases, analyzed the network topologies at each phase, identified communication security issues, and proposed strategies for securing links, nodes, and the overall network. These measures aim to enable safer communications for drone swarms in denied scenarios and to preserve operational effectiveness.
Reference: Feng Z., Yuan L., Liu J. Analysis of Secure Communication Issues for Drone Swarms in Denied Environments. Information Security and Communication Confidentiality, 2023(4):66-72.
