PaperlayerView exact revision on arXiv

Source-linked AI summary

On the Security of the Automatic Dependent Surveillance-Broadcast Protocol

Martin Strohmeier, Vincent Lenders, Ivan Martinovic

arXiv:1307.3664v2cs.CRcs.NI

TL;DR

ADS-B lacks security measures despite its expanding role in air traffic control, creating a need to understand its reported attacks and vulnerabilities. This survey synthesizes ADS-B security research, assesses countermeasures from related wireless networks, and concludes that existing approaches provide only partial improvements while comprehensive security would require protocol changes.

  • Problem

    ADS-B is an increasingly deployed air-traffic-control protocol whose inherent lack of security enables vulnerabilities including eavesdropping and jamming.

  • Method

    The survey catalogs ADS-B attacks and vulnerabilities and evaluates theoretical and practical countermeasures from ADS-B and related wireless-network research.

  • Results

    Existing solutions under consideration or used in practice can provide only a fill-in security improvement, while full cryptographic PKI is the only approach identified as guaranteeing received-data integrity.

  • Takeaways & Limitations

    Comprehensive long-term security and possibly privacy would require new message types or completely new protocols rather than relying solely on current approaches.

Abstract

from arXiv · show

Automatic dependent surveillance-broadcast (ADS-B) is the communications protocol currently being rolled out as part of next generation air transportation systems. As the heart of modern air traffic control, it will play an essential role in the protection of two billion passengers per year, besides being crucial to many other interest groups in aviation. The inherent lack of security measures in the ADS-B protocol has long been a topic in both the aviation circles and in the academic community. Due to recently published proof-of-concept attacks, the topic is becoming ever more pressing, especially with the deadline for mandatory implementation in most airspaces fast approaching. This survey first summarizes the attacks and problems that have been reported in relation to ADS-B security. Thereafter, it surveys both the theoretical and practical efforts which have been previously conducted concerning these issues, including possible countermeasures. In addition, the survey seeks to go beyond the current state of the art and gives a detailed assessment of security measures which have been developed more generally for related wireless networks such as sensor networks and vehicular ad hoc networks, including a taxonomy of all considered approaches.

I. INTRODUCTION

ADS-B is replacing independent radar with cooperative, dependent surveillance, promising lower deployment costs and greater accuracy while introducing substantial security concerns. This survey catalogs ADS-B vulnerabilities, evaluates countermeasures from ADS-B and related wireless-network research, and examines the constraints of the 1090ES protocol.

  • I. INTRODUCTION: ADS-B supports the transition from primary to secondary surveillance by using aircraft-derived positional data broadcast to ground stations and nearby aircraft.ADS-B Out transmits the data, while ADS-B In receives and processes it.
  • I. INTRODUCTION: The survey catalogs reported ADS-B vulnerabilities, analyzes a threat catalogue and Mode S vulnerability categories, and evaluates countermeasures from related wireless-network fields.It also asks why traditional cryptography cannot simply be transferred to ADS-B.
  • I. INTRODUCTION: Commercial aviation primarily uses 1090ES, which integrates ADS-B with traditional Mode S, while UAT requires new hardware and is currently used for general aviation.The survey focuses on the commercially used 1090ES data link.
  • I. INTRODUCTION: 1090ES broadcasts 112-bit ADS-B messages using pulse-position modulation, with a 24-bit CRC supporting correction of up to five bit errors.The format includes aircraft addressing and a 56-bit ME field for arbitrary data.
  • I. INTRODUCTION: Only the 56-bit ME field can carry arbitrary security data, but it is small and typically occupied by positional and other information.This makes the currently deployed format highly limiting for many security solutions.

Relation to Legacy Systems

ADS-B shifts surveillance from independent radar sensing to aircraft-dependent broadcasts, reducing deployment costs but exposing ATC to attacks inherent in unsecured wireless communication.

  • Relation to Legacy Systems: Legacy PSR independently detects reflected signals, whereas SSR depends on aircraft transponder replies; ADS-B extends this cooperative, aircraft-derived model.ADS-B aircraft collect position and velocity data and periodically broadcast them to ground stations and nearby aircraft.
  • Relation to Legacy Systems: ADS-B’s broadcast medium enables passive eavesdropping, which is difficult to prevent without encryption and practically impossible to detect.The same vulnerability also creates a widely recognized privacy concern.
  • Relation to Legacy Systems: Jamming can disable ADS-B transmitters or receivers, including through high-power transmission on Mode S’s 1090MHz frequency.Targeted attacks can create major denial-of-service problems at airports, although distributed receivers make complete area blackout more difficult.
  • Relation to Legacy Systems: Message injection is feasible because ADS-B lacks data-link authentication, allowing attackers to transmit correctly modulated and formatted non-legitimate messages.Reported effects include ground-station or aircraft ghost injection and flooding.
  • Relation to Legacy Systems: Attackers can delete messages through interference or modify them through overshadowing and bit-flipping, potentially producing corrupted or arbitrary data.Constructive interference can exceed Mode S extended squitter’s five-bit CRC correction limit, while modification attacks alter transmitted bits or replace message portions.

C. Identification of System Requirements

ADS-B operates as a highly mobile, long-range, unidirectional broadcast network with substantial packet loss and few device resource constraints, defining the setting that security solutions must accommodate.

  • C. Identification of System Requirements: The real-world network model uses single-hop, unidirectional broadcasts in which aircraft periodically beacon position, velocity, and direction in plaintext.The paper focuses on this deployed model rather than emerging multi-hop Aeronautical Ad hoc Networks.
  • C. Identification of System Requirements: ADS-B devices are modeled without energy or significant computational constraints at aircraft or ground stations.These assumptions distinguish ADS-B from many constrained wireless sensor networks.
  • C. Identification of System Requirements: The protocol provides no collision prevention, retransmission, or delivery guarantees, and the packet error rate averages about 33%.Packet loss is handled at higher layers, so candidate security schemes must operate despite unreliable physical-layer delivery.
  • C. Identification of System Requirements: The network is highly mobile and ad hoc, with aircraft reaching 1,000 km/h or more and communications between two nodes potentially lasting only seconds.Aircraft trajectories are not generally physically restricted, although common routes and restricted airspaces exist.
  • C. Identification of System Requirements: ADS-B is a long-range system, typically considered feasible at distances of 100 NM and more, while legitimate-node capture is less concerning than legal access to a node.The paper notes that legally accessing a legitimate ADS-B node is not considered very difficult, particularly in general aviation.

Security Attributes:

The survey frames ADS-B security around authenticating senders and locations while preserving operational compatibility, rapid detection, scalability, and resilience to packet loss and denial of service.

  • Security Attributes:: Secure broadcast must establish that information comes from the appropriate sender while allowing senders to limit broadcast recipients when needed.Confidentiality is identified as a separate concern that was not considered in ADS-B’s original development.
  • Security Attributes:: The paper identifies data integrity, source integrity, and data-origin authentication as core security properties for ADS-B.These respectively concern unmodified data, the claimed participant, and the location claimed in the message.
  • Security Attributes:: Candidate schemes should remain compatible with current installations, detect incidents quickly and correctly, and resist denial-of-service attacks against computing resources.These requirements constrain both deployment impact and defensive processing costs.
  • Security Attributes:: Security approaches must scale with rising aircraft density and global traffic without measurably increasing strain on the heavily used 1030 MHz channel.The requirement covers both local density growth and globally increasing aircraft traffic.
  • Security Attributes:: Security and reliability should not deteriorate when wireless channels are jammed, while non-repudiation is treated as desirable but not an immediate priority.The survey also organizes proposed protections into secure broadcast authentication and secure location verification.
  • Security Attributes:: The survey broadens prior ADS-B work by cataloguing threats, assessing countermeasures from related wireless fields, and weighing their advantages and disadvantages.Its taxonomy distinguishes secure broadcast authentication from secure location verification.

B. ADS-C

ADS-C is a connection-oriented alternative for suspicious ADS-B participants, but its infrastructure, performance, cost, visibility, and communication model limit its usefulness for NextGen ATC.

  • ADS-C requires additional avionics systems such as FANS 1/A or ATN for data communications.
  • Its performance may be limited by the communications medium, and each transmission can incur a data-link service-provider cost.
  • Unlike ADS-B IN, ADS-C messages are not directly available to other aircraft.
  • ADS-C over ACARS considerably limits usefulness because it sends on demand rather than periodically broadcasting, sacrificing ADS-B’s cost, scalability, and ease-of-use advantages.
  • Military Mode 5 adopts broadcast capability and cryptographic protections, but incomplete performance information and likely disadvantages in cost, scalability, and ease of use limit direct comparison.

IV. SECURE BROADCAST AUTHENTICATION

The survey examines broadcast-authentication approaches for ADS-B, emphasizing node authentication while assessing cryptographic and physical-layer alternatives against the protocol’s unidirectional, dynamic environment.

  • Broadcast authentication requires an asymmetric mechanism so receivers can verify messages without generating authentic messages themselves.
  • Reactive authentication could preserve ADS-B’s open nature by adding security only when suspicious behavior is detected, reducing overhead during ordinary operation.
  • The survey focuses mainly on node-based schemes that authenticate hardware rather than user-based schemes that authenticate humans.
  • Non-Cryptographic Schemes: Fingerprinting can identify device classes or individuals through software, hardware, or channel characteristics, but it does not provide surefire security.
  • Channel/Location-Based Fingerprinting: Physical-layer schemes requiring bidirectional communication are incompatible with current ADS-B, whose 1090ES channel leaves only about 0.6188 ms of coherence time at 800 km/h.
  • Randomized/Uncoordinated Frequency Hopping / Spreading: Frequency-hopping schemes are unsuitable for worldwide civil ATC because secret spreading codes or hopping patterns would likely not remain secret.

B. Public Key Cryptography

Public-key cryptography is considered for authenticating ADS-B broadcasts, but certificate management, compatibility, message-size constraints, and preserving open broadcast create substantial deployment barriers.

  • Proposed ADS-B PKI designs include ECDSA signatures, X.509 certificates, certificate distribution through aviation assets, and retroactive signature transmission.
  • Broadcast authentication favors public-key cryptography because receivers must verify messages without sharing a key that would let them generate valid messages.
  • Centralized certificate authorities face breach risks and cross-airspace interoperability challenges when different aviation authorities mandate separate systems.
  • Encryption would leave management and control frames unprotected, break compatibility with the installed base, and make ad hoc key exchange costly and difficult.
  • A cryptographic design comparable to ADS-S would not preserve public broadcast communication, conflicting with ADS-B’s open nature.
  • One-time signatures are infeasible on standardized 1090ES because signing 60 bits requires at least 80 bytes of overhead; traditional cryptography remains difficult overall.

C. Retroactive Key Publication

Retroactive key publication enables broadcast authentication by sending MACs first and disclosing keys later; TESLA-family methods reduce overhead but introduce timing and operational trade-offs.

  • Retroactive authentication sends an encrypted MAC with each broadcast message and publishes its decryption key after a set time or number of messages.
  • TESLA provides scalable broadcast authentication for lossy, real-time networks, while µTESLA adapts it for wireless sensor networks.
  • TESLA and µTESLA use one-way key chains, with secret interval keys disclosed after a delay so receivers can authenticate earlier messages.
  • µTESLA tolerates packet loss on the jam-prone 1090 MHz channel and requires less communication overhead and protocol modification than traditional asymmetric cryptography.
  • µTESLA may require reinitialization and can be susceptible to memory-based denial-of-service attacks, motivating a forward-hashing variant that avoids reinitialization.
  • In VANET simulations, TESLA with keys attached to subsequent broadcasts performed better on channel congestion and MAC-layer delay than ECDSA and separately published keys.

V. SECURE LOCATION VERIFICATION

Secure location verification checks whether aircraft location claims are authentic, primarily using multilateration from time-difference measurements. The approach is deployed on the ground but faces accuracy, coverage, cost, and communication-security challenges.

  • A. Multilateration: Multilateration estimates an aircraft’s 3D position from time differences of the same signal received at four or more antennas.The measurements define hyperboloids, whose intersection identifies the aircraft position.
  • A. Multilateration: TDOA multilateration is the preferred ground-based location-verification solution and can reuse existing aircraft communications.It requires new ground receiver and central processing stations rather than aircraft infrastructure changes.
  • A. Multilateration: At 90 NM, multilateration provides roughly 30 m accuracy compared with 20 m for ADS-B, with practical accuracy deteriorating over longer distances.WAMLAT is relatively easy and cost-effective to deploy on the ground, but coverage over vast open spaces can be infeasible.
  • A. Multilateration: Ground-based multilateration has difficulty estimating aircraft altitude and is affected by multipath, missed detections, and the need for links between receivers and processing stations.Angle-of-arrival measurements have been proposed to improve unsatisfactory height estimates.
  • A. Multilateration: Attacking multilateration requires modified TCAS receivers, precise relative timing, and a track-determination algorithm, imposing cost and engineering effort.Unsecured communication among localization participants nevertheless leaves open the possibility of Sybil attacks.

B. Distance Bounding

Distance bounding uses challenge-response timing to establish that a prover is within a specified distance. Although related schemes can combine timing, plausibility, and signal-strength checks, ADS-B’s long distances, high velocities, and broadcast-only design limit direct applicability.

  • B. Distance Bounding: Distance bounding has a prover respond to a verifier’s challenge so the verifier can establish an upper bound on physical distance.A man-in-the-middle can add processing delay and appear farther away, but cannot reduce the measured distance.
  • B. Distance Bounding: Distance-bounding protocols remain exposed to relay attacks including distance fraud, mafia fraud, terrorist fraud, and distance hijacking.These attacks are documented in the broader distance-bounding literature.
  • B. Distance Bounding: A VANET-derived scheme combines traditional distance bounding with transmission-range and speed-based plausibility checks, followed by location verification.Trusted entities can also collaborate to estimate a prover’s location through trilateration.
  • B. Distance Bounding: Under idealized assumptions, secure multilateration combining distance bounding and RSS differences can detect false location claims and mitigate distance-enlargement and collusion attacks.The result demonstrates a theoretical combination of physical-layer techniques, while practical ATC challenges remain.
  • B. Distance Bounding: Distance bounding is unsuitable for the long distances and high velocities of air traffic control and requires a prover response absent from ADS-B.As an on-demand feature, it could still provide legitimacy information where primary radar is unavailable or phased out.

C. Kalman Filtering and Intent Verification

Kalman filtering recursively predicts and updates aircraft states to smooth noisy measurements, bridge missing data, and support plausibility and intent checks. Its usefulness is constrained by modeling assumptions, computational cost, limited history, and susceptibility to gradual trajectory manipulation.

  • C. Kalman Filtering and Intent Verification: Kalman filtering recursively predicts state variables and uncertainties, then updates estimates using weighted measurements.The procedure can operate in real time while retaining only the previous state.
  • C. Kalman Filtering and Intent Verification: Kalman filtering assumes a linear observed system with normally distributed variables and errors, although adaptations for nonlinear systems exist.Aircraft-to-aircraft positional-claim filtering is more difficult than several other ADS-B-related filtering uses.
  • C. Kalman Filtering and Intent Verification: In ADS-B systems, Kalman filters smooth noisy signals, coast over missing data, and verify reported state vectors and trajectory changes.They are used in multilateration and ground systems for plausibility checks.
  • C. Kalman Filtering and Intent Verification: Intent verification compares aircraft motion with plausible intent models across horizontal, vertical, and velocity dimensions.The approach also evaluates geometric conformance against specified horizontal and vertical limits.
  • C. Kalman Filtering and Intent Verification: A frog boiling attack jams the correct signal and gradually modifies transmitted positions until the filter accepts them as valid trajectory changes.Filtering still detects obviously bogus manoeuvres, but its trust-establishment threshold remains open research.

D. Group Verification

Group verification lets trusted aircraft use multilateration to check location claims from group members and nearby outsiders. It can increase attack difficulty, but requires additional communication, membership authentication, and resilience to jamming.

  • D. Group Verification: A trusted group of four or more aircraft can use TDOA or RSS multilateration to verify location claims from members and nearby outsiders.If a forged report is detected, nearby aircraft may need a larger avoidance circle because the reported position is no longer precisely trusted.
  • D. Group Verification: Around 91% of aircraft at a given time could belong to a sufficiently large group of at least four aircraft in the studied US airspace.The finding comes from a study examining substantial traffic-density differences across the United States.
  • D. Group Verification: Group verification requires extra messages and a new protocol because ADS-B communication is purely unidirectional broadcast.L-DACS is mentioned as a possible future air-to-air communication technology, but no medium-term specification was available.
  • D. Group Verification: Securely admitting new members, excluding malicious aircraft, and handling intentional jamming remain central group-verification challenges.These issues arise even if a suitable air-to-air communication protocol is implemented.
  • D. Group Verification: Even without a perfectly secure solution, group verification could raise the difficulty and engineering effort of some airborne attacks by orders of magnitude.The proposed benefit is therefore increased attack resistance rather than complete security.
  • D. Group Verification: Data fusion can combine ADS-B, multilateration, flight-plan, and radar data to improve security and accuracy near airports without amending ADS-B.Its principal downside is the increased cost of additional redundant systems.

F. Traffic Modeling and Plausibility Checks

Traffic modeling and plausibility checks use historical data, statistical position comparisons, and simple behavioral rules to flag unusual ADS-B claims. These indicators can support investigation and reduce attacks, but require careful scrutiny because no individual rule is necessary or sufficient.

  • Traffic Modeling: Historical data and machine learning can build ground-station heat maps that flag aircraft whose claimed locations do not match expected RSS patterns.The models provide hints about non-matching location claims rather than definitive attack determinations.
  • Statistical Verification: Statistical verification compares sequential claimed and estimated positions, treating sufficiently sampled differences as normally distributed random error.The approach is intended for fixed receivers, especially ground stations able to collect enough measurements.
  • Operational Use: Plausibility rules are neither necessary nor sufficient alone, but technically or physically impossible packets may be dropped to reduce system strain and prevent poorly crafted spoofing or denial-of-service attacks.Multiple checks can increase an attacker’s alert risk and attack cost, while suspected attacks still require human or additional technical scrutiny.
  • Plausibility Rules: Simple red flags include impossible locations, impossible velocities, excessive aircraft density, unusual map positions, flight-plan mismatches, and discontinuities in state-vector fields.These checks can be applied across physical, operational, and application-layer information.

VI. SUMMARY

The survey finds that ADS-B security approaches differ in attack coverage, security properties, and deployment feasibility, with no single low-impact solution providing comprehensive protection. It therefore favors incremental, backward-compatible measures for the current system while reserving comprehensive security for future protocols.

  • Security Coverage: Most examined security schemes target message injection or modification, while passive attacks remain difficult to protect against without full cryptography.The survey attributes this focus partly to ADS-B’s desirable open nature and the difficulty of securing eavesdropping.
  • Security Features: Only a full cryptographic public key infrastructure can guarantee received-data integrity; other approaches secure source integrity or independently verify location claims.Spread spectrum and cryptography can also directly protect against flood denial-of-service attacks, whereas other methods rely on higher layers.
  • Feasibility: Wide-area multilateration, Kalman filters, and data fusion are already in use, whereas distance bounding and full public key infrastructure remain technically difficult and costly.The feasibility comparison emphasizes a correlation between implementation difficulty and cost, alongside choices between protocol replacement, modifications, and parallel systems.
  • Future Protocols: Because air-traffic systems take decades to certify and deploy, future protocols should incorporate authentication, suitable cryptography, communication patterns, key management, and lossy-channel constraints from the outset.The survey uses ADS-B as a case study for planning long-term protocol security.
  • Near-Term Measures: For the current system, backward-compatible incremental changes such as multilateration, fingerprinting, and improved data fusion are more practical than a complete overhaul.These measures can improve location verification, intrusion detection, safety, or security, but may sacrifice some original ADS-B advantages.
  • Scope Boundary: The survey primarily addresses attack detection and prevention, leaving attack reaction for future research.Deployment cost and complexity are important but are not the only considerations because insufficient security can be more expensive.
  • Conclusion: The survey concludes that current solutions are temporary improvements, while comprehensive security and possibly privacy require new message types or entirely new protocols.Long-term designs should assess secure broadcast authentication, secure location verification, traffic density, and communication overhead.
Loading 1307.3664v2…