Securing Edge & Physical AI: Post-Quantum IoT Blockchain for Mission-Critical Cyber-Physical Agents

Kürşat Çakal; Suat Özdemir

cs.CROct 24, 2026

Securing Edge & Physical AI: Post-Quantum IoT Blockchain for Mission-Critical Cyber-Physical Agents

Authors: Kürşat Çakal, Suat Özdemir

Organizations: Department of Computer Engineering, Hacettepe University, Ankara, 06800 Turkey

Abstract

Distributed Cyber-Physical Systems (CPS) and Edge AI networks increasingly depend on public-key authentication and key-establishment mechanisms that are vulnerable to cryptographically relevant Quantum Computers (QC). Shor's algorithm threatens widely deployed asymmetric primitives, while Grover's algorithm reduces the effective security margins of symmetric and hash-based mechanisms. Although Post-Quantum Cryptography (PQC) provides a migration path, its larger public keys, ciphertexts, and signatures introduce additional computation, memory, communication, storage, and packet-fragmentation costs that can constrain dynamic multi-agent systems. In this article, we introduce rotational secrecy, an architectural objective that limits temporal key exposure through authenticated transitions between cryptographic epochs. We present Private Authentication for Rotational Secrecy (PARS), a reference mechanism that coordinates policy-triggered key establishment, rotation, and reconciliation among authorized Physical AI agents. For deployments that need distributed agreement on security state, we further develop the PARS-Enabled Edge Security Ledger (PARS-ESL), an optional permissioned architecture maintained by authorized validators or capable edge gateways. PARS-ESL does not store high-volume operational transactions. Instead, it records authenticated state for key epochs, group membership, revocation, algorithm-suite policies, challenge validation, and recovery checkpoints. As a result, resource-constrained agents can take part as lightweight clients, and low-latency operational traffic stays outside the ledger consensus path, protected by the currently active symmetric epoch keys. Our results show that ledger-assisted post-quantum key-epoch management is feasible for selected Edge AI and CPS security-control workloads under the stated trust, compromise, connectivity, and resource assumptions.

Figures & tables

Fig. 1 : Evolution of Physical AI Systems [ 1 , 2 , 3 ] .

Fig. 2 : North Atlantic Treaty Organization Command Structure

Fig. 3 : Quantum Information Science Overview

Fig. 4 : Security Levels of Generic Algorithms for Blockchain

Fig. 5 : Conceptual Overview of Military Edge AI

Fig. 6 : Cyberspace Attributes for Mission-Critical Edge AI Avionics

Fig. 7 : Radio Frequency SIGINT in Space and Defense Environments

Fig. 8 : Blockchain Pipeline Basic Overview

Fig. 9 : Blockchain Types [ 59 ]

Fig. 10 : Blockchain Use Cases for Edge AI

Fig. 11 : Quantum vs Classical Computing Base

Fig. 12 : Fundamental Concepts for Quantum Computers

Fig. 13 : Cryptography Systems for Blockchain Architecture

Fig. 14 : Lattice Based Cryptography Overview

Fig. 15 : Code Based Cryptography Overview

Fig. 16 : Supersingular Isogeny Based Cryptography Overview

Fig. 17 : Multivariate and Hash Based Cryptography Overview

Type	Methodology	Pros & Cons	Target	Contribution	Status
S-QM	• Intertwined Architecture for IoT, BC and QCMPs • Identity Management Mechanism • Data Integrity Before Storage • Heterogeneity Gateway Support Among Edges • Data Transfer Over Quantum Distributed Ledger	Pros: • Ensuring Data Integrity Before Storage • Facilitates Decentralization in IoT • Focuses IoT Resource Limitation Cons: • Depends on Quantum Computers	IoT [ 70 ]	• Propose Multilevel Distributed-Ledger • Integration Support for Heterogeneous Edges • Building Secure IoT Applications with BC and QCMPs	Empirical
	• Quantum Communication Infrastructure • Information Processing with Blockchain • Quantum-Reflection Key Distribution Protocol	Pros: • No Need for Qubits • Traditional Device Compliance • Resist Various Attacks • High Computational Efficiency • High Level of Practicability Cons: • Lack of Native Quantum Resistance • Implementation Difficulty • Complex Infrastructure Cost	IoV&IoT [ 71 ]	• Two Novel Lightweight Quantum-Reflection Protocols • Reduced Communication Rounds • Overcome Rare Q-Bits • Offering Compatibility for Traditional Device	Theoretic Notional
	• Quantum Key Distribution • Secure Authentication Algorithm	Pros: • Anonymity of Network Peers • Supports Large Scale Blockchains Cons: • Lack of Attack Mitigation Analysis • Lack of Technical Details • Lack of Functional Supremacy	Generic [ 72 ]	• Efficient Blockchain for Quantum Attacks • Experimental Realization	Notional Empirical
	• Usage of QDS and QM • Consensus Algorithm • Entanglement Operations	Pros: • Weighted Graph States • Undeniability of Nodes • Resistant to Hash Rate Attacks • Solves Heterogeneity Problem • Resource Friendly • Lower Computation Volume • Wide-Spectrum of Applicability Cons: • Technically Challenging • Complex Architecture	Generic [ 73 ]	• Present Higher Efficiency • Novel Consensus Mechanism	Theoretic Empirical
	• Quantum-native chain via temporal GHZ states • Entanglement-in-time block links	Pros: • Past blocks physically tamper-proof • Subcomponents experimentally realizable Cons: • No implementation needs quantum network • No key rotation	Generic [ 69 ]	• Pioneering quantum-native blockchain concept	Conceptual
	• Quantum-blockchain framework for EMRs • Quantum-authenticated transactions	Pros: • Privacy-preserving EMR sharing • Quantum-secured integrity Cons: • Assumes quantum channels • No key rotation	IoMT [ 83 ]	• Quantum-BC framework for private EMRs in IoMT	Theoretical

TABLE I : Quantum Security Architectures - Quantum Mechanics (S-QM) Schemes

Type	Methodology	Pros & Cons	Target	Contribution	Status
S-PQCL	• Lattice Based Delegation Algorithm • Double Signature Design • Preimage Sampling Algorithm	Pros: • Reduced Correlation Between Message and Signature • Shorter Signature and Key Size • Decreased Computational Complexity • Higher Security and Efficiency • Resist Various Attacks Cons: • Implementation Complexity and Time • Lack of Security Analysis • Lack of Quantum Effects on Financial Backbone	Finance [ 80 ]	• Novel Signature Schema • Quantum Secure Cryptocurrency Architecture	Theoretic
	• Lattice Based Encryption Architecture	Pros: • Reduced Communication Cost • Ensure Forward Security • Resist Private Key Attacks • Higher Efficiency • Lower Execution Time Cons: • Reduced Trapdoor • Complex Notation Flow for Security Analysis • Possible Backdoors	Health [ 82 ]	• BC Based Searchable Encryption Schema for QAs • Solving Key Leakage Problem • Robust Ontological Backbone for Healthcare	Theoretic Empirical
	• Lattice RingCT (Module-SIS/LWE) • Efficient PQ private payments	Pros: • Compact proofs vs prior lattice RingCT • Confidential transactions Cons: • Heavier than classical (ECC) RingCT • No key rotation	Finance [ 81 ]	• Efficient post-quantum private blockchain payments (IEEE S&P)	Empirical

TABLE II : Quantum Security Architectures - Lattice-Based Post-Quantum (S-PQCL) Schemes

Type	Methodology	Pros & Cons	Target	Contribution	Status
S-PQC	• PARS Protocol (Rotational Secrecy) • PoE Consensus Mechanism • Layer-1 Private Quantum Blockchain • Light Node Executive Architecture	Pros: • Algorithm-agnostic (modular structure) • High energy efficiency • Low storage footprint • Robust quantum attack mitigation • Optimized for edge hardware • Strong empirical baseline Cons: • Lacks holistic empirical • Performance metrics • (deferred to this study) • Coordination overhead for pattern consensus	IoMT& Military& Generic [ 35 ]	• Proposes Layer-1 Blockchain for military systems • Dynamic key reconciliation for tactical swarms • Introduces novel PoE consensus • Comprehensive benchmark of NIST PQC candidates	Empirical
	• Lattice Based DSA Blockchain • IPFS	Pros: • Lower Transaction Size • Increased Performance • Optimized Block Size in Traditional Architectures Cons: • Limitations for Parallel Transactions • Lack of Performance and Scalability Analysis	Generic [ 74 ]	• Propose Quantum-Resistant DSA Based Blockchain • Plan to Resist Quantum Attacks by qTesla • Ensure Higher Number of Transactions	Empirical
	• Novel Quantum PoW Consensus Mechanism	Pros: • Solve Random Quadratic Equations • Enhanced Security • Higher Transaction Amount Cons: • Holds Possible Back Doors of Rainbow	Generic [ 39 ]	• Memory Mining Capabilities • Reduce Validity Concerns on Transactions • Efficient Block Expansion	Theoretic Empirical
	• Conditional Privacy-Preserving Authentication • Lattice Based Architecture • Authentication Algorithm	Pros: • Energy and Storage Efficiency • Lower Computation Cost • Rich Set Amount Attacks Mitigation Capability • Rich Set Amount Attacks Mitigation Capability • Sophisticated Security Functionalities • Strong Analytical Analysis Cons: • Loosely Correlated Cost Results • Loosely Correlated Proofs • Complex Model for Implementation Phase	IoV& IoT [ 75 ]	• Propose Quantum-Resistant Authentication Schema • Reliable Consortium Blockchain for Data Management • Enhance Security Functionalities of Literature	Notional Theoretic Empirical
	• Quantum-Safe XMSS • Novel Mining Algorithm • Key Migration from Old to XMSS Based Keys • W-OTS+ Schema for Avoiding One-Time • Non-Interactive ZKP for Privacy	Pros: • Lack of Transaction Malleability • Suitable for Lightning Network • Protected Network Efficiency • Burn Operation for Non-Migrated Coins Cons: • Extended XMSS Signature Migration is Required • Mix of Classical Approaches like Equihash, SHA3 • Higher Witness Scale Factor	Finance [ 77 ]	• Quantum-Safe Cryptocurrency Schema • Serves for Avoiding Possible Defects of Bitcoin for QAs	Theoretic Notional
	• Multi-sensor input fusion (RF/IR/radar/RGB) • Post-Quantum Cryptography (Kyber/Dilithium2) • YOLOv8, Swin-T, and Quantum Variational Classifier	Pros: • On-device threat detection and active response • Low average inference latency (52.9 ms) • High accuracy (F1-score of 90.0) Cons: • High local computational/thermal demand • Heavy hardware requirements for multimodal execution	Edge AI Surveillance [ 92 ]	• Proposes first fully integrated, PQC-secure Edge AI threat mitigation platform (Karana) • Hardware-validated on NVIDIA Jetson architectures	Empirical

TABLE III : Quantum Security Architectures - Post-Quantum Cryptography (S-PQC) Schemes

Fig. 18 : Modular Overview of Proposed Study

Fig. 19 : Defense Roadmap Security Paradigm Shift

Fig. 20 : Quantum Attacks - Denial of Engagement 1

Fig. 21 : Quantum Attacks - Denial of Engagement 2

Fig. 22 : Quantum Attacks - Denial of Command Control

Fig. 23 : Quantum Attacks - Denial of System

Fig. 24 : Quantum Attacks - Denial of Line-of-Sight

Fig. 25 : Quantum Attacks - Denial of Cooperation

Fig. 26 : Comparison of Key Exchange Methods

Fig. 27 : PARS Protocol Operational Flow.

1:	function initParams( $Node_{i}$ , $Count_{\text{Node}}$ )
2:	$PK_{i},M_{\text{id}},P_{\mathrm{dp}}\leftarrow 0$
3:	for each $Node_{i}\in\text{Military Space}$
4:	$PK_{i}\leftarrow\text{ProtectorKeysForNewSession}$
5:	$M_{\text{id}_{i}}\leftarrow\text{GenerateMachineID}$
6:	$P_{\mathrm{dp}}\leftarrow\text{GeneratePatternDesignParams}$
7:	return $PK_{i}$ , $M_{\text{id}}$ , $P_{\mathrm{dp}}$

Algorithm 1 PARS Protocol (Alignment)

Fig. 28 : ZKP Hardness for Challange Response Mechanism

Fig. 29 : PARS Protocol Generation and Protection Metrics

CNRME Encapsulation

Algorithm 3 PARS Protocol (CNMRE - Traditional) - Deterministic Post-Quantum Obscurity Encapsulation

CNRME Key Derivation

Algorithm 4 PARS Protocol (CNMRE - PQC) - Deterministic Post-Quantum Obscurity Encapsulation

Fig. 31 : Edge AI Nodes Quantum Security Flow

Fig. 32 : Sequence Diagram Proposed Security Stack

Fig. 33 : Private Authentication for Rotational Secrecy

Fig. 34 : PQC Algorithm Runtime Analysis

Algorithm	Security Level	Key Pair	Enc e /Sign f	Dec e /Verify f	Scheme Type	Attack/Mitigation
BIKE e,d	256 – 128 (bits)	0.001458s	0.000099s	0.009767s	Code Based	Structural Attacks
Key Recovery Attacks
Squaring Attacks
Power Analysis Attack
Side Channel Attacks
Reaction Attacks

TABLE IV : Quantum Age Cryptosystem Analysis for Architecture

Fig. 35 : Average Consensus Time as a Function of Swarms

Fig. 36 : Size Metric Footprints of NIST Schemes for Edge AI

Fig. 37 : Accumulated Memory Footprints of NIST Schemes

Fig. 38 : Denial of Consensus Rounds

Fig. 39 : ESL Failure Results

Platform	Compute core(s)	Accel. path	Lattice (KEM/DSA)	Hash (SLH-DSA)
Jetson Orin	Cortex-A78AE + Ampere GPU	NEON / GPU	$\approx$ 2.5–4 $\times$ (CPU); $\leq$ 1 $\times$ batched	$\approx$ 1.5–3 $\times$ (crypto ext.)
Jetson AGX Orin Ind.	Hardened Cortex-A78AE + GPU	NEON / GPU (ECC)	$\approx$ 2.8–4.5 $\times$ (CPU); $\leq$ 1.1 $\times$ batched	$\approx$ 1.6–3.3 $\times$ (crypto ext.)
Jetson Nano	Cortex-A57 + Maxwell GPU	NEON (no crypto ext.)	$\approx$ 12–18 $\times$	$\approx$ 25–40 $\times$
i.MX RT1064 (MCU)	single Cortex-M7 @600 MHz	assembly only	$\approx$ 1–2 ms/op (ML-KEM-768); 5–15 ms ML-DSA-65 sign [ 121 ]	impractical for fast re-keying
TI Sitara AM69x	8 $\times$ Cortex-A72 @2 GHz + R5F + HSM	NEON / SW parallel + HSM	$\approx$ 3.5–5.5 $\times$ (projected)	$\approx$ 2–3.5 $\times$ (projected)
PolarFire SoC (RISC-V)	quad RISC-V U54 (in-order, 64-bit)	none	$\approx$ 20–45 $\times$ (projected)	several seconds (projected)

TABLE V : Architecture-Based Projection of PQC Core Cost Across Edge AI Platforms

Platform	ML-KEM	ML-DSA	SLH-DSA	FN-DSA / HQC	Primary support path
Intel i5-12450H (baseline)	SW	SW	SW	draft (liboqs)	OpenSSL 3.5 / liboqs (AVX2)
Jetson Orin	GPU, SW	GPU, SW	SW	draft (liboqs)	cuPQC + OpenSSL 3.5 (NEON/Ampere)
Jetson AGX Orin Ind.	GPU, SW	GPU, SW	SW	draft (liboqs)	cuPQC + OpenSSL 3.5 (rugged ECC)
Jetson Nano	GPU, SW	GPU, SW	SW	draft (liboqs)	cuPQC + OpenSSL 3.5 (NEON)
i.MX RT1064 (MCU)	asm, IP	asm, IP	verify (IP)	draft	pqm4 + PQMicroLib ( $<$ 5 KB)
TI Sitara AM69x	SW, HSM	SW, HSM	SW, HSM	draft (liboqs)	OpenSSL 3.5 + HSM PQC extensions

TABLE VI : PQC Support Across the Edge Platforms

Data Flow Class	Operational Transfer Profile	Cryptographic Pattern	PQC Per-Object Overhead & Steady-State Cost
Sensor / Streaming Imagery	Large volume, continuous streaming, soft real-time constraint.	ML-KEM-768 $\rightarrow$ AEAD bulk stream; ML-DSA-65-signed keyframes.	$\approx$ 2.3 KB per session (ek + ct) $+$ 3.3 KB keyframe signature; 16 B tag per packet.
Federated Learning Updates	Periodic batch transfer, MB-scale parameter tensors, delay-tolerant.	ML-DSA-65-signed model tensors over established KEM channel.	3.3 KB update signature ( $\ll$ 1% of an MB-scale tensor); negligible relative overhead.
Telemetry / C4ISR Control	Small payload, high-frequency broadcast, hard real-time constraint.	ML-KEM-512 / ML-DSA-44 $+$ symmetric AEAD stream.	$\approx$ 1.6 KB per session (ek + ct) $+$ 2.4 KB key signature; 16 B tag per packet.
Configuration Database	Infrequent synchronization, KB–MB scale, integrity-critical.	ML-DSA-65/87-signed snapshot manifests and state hashes.	3.3–4.6 KB per snapshot signature; amortized across rare sync events.
Software / Firmware (OTA)	Rare distribution, large image size, very high integrity requirement.	Stateful LMS/XMSS ( $h=10$ ) or ML-DSA-87, offline signed ⋆ .	$\approx$ 1.5–2.5 KB per image (LMS/XMSS) or 4.6 KB (ML-DSA-87); 7.9 KB (SLH-DSA-128s).
Analysis / Mission Logs	Variable batch size, archival storage, provenance-critical.	ML-DSA-87 or SLH-DSA-128s $+$ timestamped state hash ⋆ .	4.6–7.9 KB per record batch signature; long-term tamper-evident proof.

TABLE VII : Cryptographic Patterns and Overhead Across Representative Edge AI Data Classes

Data Flow Class	Primary Target Substrate	PARS-ESL Architectural Role	Key Epoch Management & Rekeying Trigger
Sensor / Streaming Imagery	GPU / AI Engines (Inference) $+$ Spatial PL Fabric (AEAD Bulk)	Negotiates ephemeral session keys via PARS; records state anchors on ESL.	Time-Based (TB) Periodic Security Threshold Control (PSTC) windowed re-keying.
Federated Learning Updates	GPU / AI Engines (Tensors) $+$ Host Processing System (PS)	Provides ledger-anchored model provenance; validates client updates immutably.	Windowed re-keying aligned with federated learning round completion epochs.
Telemetry / C4ISR Control	Deterministic PL Datapaths $+$ Real-Time Cores (e.g., Cortex-R5F)	Enforces low-latency authentication filters with minimal memory footprint.	Action-Based (AB) / Group-Based (GB) threshold triggers upon operational state shift.
Configuration Database	Host Processing System (PS) Secure Enclave / Flash Storage	Maintains versioned, immutable configuration snapshot roots on-chain.	Event-driven reconciliation upon policy update or node group reconfiguration.
Software / Firmware (OTA)	Parallel PL Keccak Engines (Verify) $+$ Hardened RoT Enclave	Writes immutable release image hashes and verification receipts to the ledger.	Offline signed; validated on boot/update prior to key epoch transition.
Analysis / Mission Logs	Host PS Core $+$ High-Throughput PL Cryptographic Hashing	Commits archival root hashes to ESL for post-mission auditing and traceability.	Continuous epoch-stamped logging decoupled from live command channels.

TABLE VIII : Hardware Substrate Allocation and PARS-ESL Protocol Roles for Edge AI Traffic

Operational Scenario	Target Hardware Platform	Swarm Size	KEM Choice	DSA Choice	ZK Level	PSTC Configuration
Peacetime Base Security / C2	NXP i.MX RT1064 (Cortex-M7 MCU)	$\leq$ 4 nodes	ML-KEM-512	ML-DSA-44	Level 1	$\approx$ 60 min (TB, Low Footprint)
Low-Power Tactical Comms	Microchip PolarFire SoC (RISC-V + PL)	$\leq$ 8 nodes	ML-KEM-512	ML-DSA-44	Level 2	$\approx$ 30 min (TB, Power-Optimized)
Civilian Edge Logistics Swarms	NVIDIA Jetson Nano (Cortex-A57 Base)	$\leq$ 8 nodes	ML-KEM-768	ML-DSA-65	Level 1	$\approx$ 30 min (TB, Software EVP Path)
Active Tactical Mobile Maneuvers	AMD/Xilinx Zynq-7000 (Cortex-A9 + PL)	$\leq$ 12 nodes	ML-KEM-768	ML-DSA-65	Level 2	$\approx$ 15 min (AB, Spatial PL Acceleration)
Tactical Drone Swarm	TI Sitara AM69x (Cortex-A72 + HSM)	$\leq$ 12 nodes	ML-KEM-768	ML-DSA-65	Level 2	$\approx$ 15 min (AB, Real-time Isolated)
Tactical Recon Radar	Altera Agilex 9 Direct RF (Cortex-A53 + PL)	$\leq$ 12 nodes	ML-KEM-768	ML-DSA-65	Level 2	$\approx$ 15 min (AB, Direct-RF Mode)

TABLE IX : Platform-Specific Parameter Extrapolations Based on GCS

Operational Tier	Target Platform	Compute Architecture & Substrate	PQC Support Status & Primary Path	Key Architectural Pros (Advantages)	Key Constraints & Cons (Trade-offs)	Applied Protocol Paradigm	Recommended Algorithm Suite
Gateway / GCS	Intel Core i5-12450H (GCS Baseline)	8C/12T @ 3.20 GHz (64 GB DDR4)	Full PQC (Native) OpenSSL 3.5 / liboqs (AVX2 Vectorized)	High raw throughput, native AVX2 SIMD acceleration, full support for stateful and stateless post-quantum schemes.	High active power draw (45–95 W), bulky form factor, unsuited for SWaP-constrained field or airborne platforms.	Full PQC Node (PARS-ESL Root / Validator)	ML-KEM-1024 / SLH-DSA-128s ZKP Lvl 4, Continuous / Event
Strategic Space & Orbital AI	AMD XQR Versal (Space Adaptive SoC)	Rad-Hard Cortex-A72 + TMR-Hardened PL + AIE	Full PQC (Accelerated) Hardened PL Cores + PS OpenSSL 3.5	MIL-PRF-38535 Class B space-flight qualified, TMR-hardened spatial PL eliminates SEUs, AI Engines isolate orbital vision pipelines.	High unit procurement cost, steep learning curve for AI Engine graph programming and spatial routing toolchains.	Full PQC Node (Mode A, Accelerated)	ML-KEM-1024 / ML-DSA-87 ZKP Lvl 4, PSTC: 5–10 min (GB)
	Microchip PIC64-HPSC (NASA Space MPU)	8 $\times$ SiFive X280 @ 1.0 GHz (512-bit Vector RVV)	Full PQC (Accelerated) 512-bit RVV ISA + Hardened Enclave	512-bit vector registers accelerate polynomial NTT loops without discrete FPGA fabric, Secure Enclave provides a hardware RoT.	Emerging RISC-V space software ecosystem, non-reconfigurable logic (fixed ISA), early-stage RVV auto-vectorizing compilers.	Full PQC Node (Mode A, Vectorized)	ML-KEM-1024 / ML-DSA-87 ZKP Lvl 4, PSTC: 5–10 min (GB)
	Microchip RT PolarFire (Rad-Tolerant RISC-V)	Quad SiFive U54 (64-bit) + Flash-FPGA Fabric	Full PQC in PL / Hybrid Custom PL Core + RISC-V Linux	Non-volatile flash FPGA is inherently immune to configuration memory SEUs (no scrubber needed), lowest static power ( $<$ 3 W).	In-order 64-bit RISC-V application cores exhibit severe software-only execution penalties without custom PL accelerator cores.	Full PQC (PL Mode A) / Classical (PS Mode B)	ML-KEM-512 / ML-DSA-44 ZKP Lvl 2, PSTC: $\approx$ 30 min (TB)
Tactical Defense & Electronic Warfare	Altera Agilex 9 (Direct RF SoC FPGA)	Quad Cortex-A53 (HPS) + Direct RF PL (64 Gsps)	Full PQC (Accelerated) Direct-RF PL + Secure Device Manager	Direct RF sampling (up to 64 Gsps) enables real-time spectral EW integration, Secure Device Manager (SDM) enforces PQC boot.	High static/dynamic power draw (25–50 W), requires advanced conduction cooling enclosures in sealed tactical missile/pod bays.	Full PQC Node (Mode A, Direct RF)	ML-KEM-768 / ML-DSA-65 ZKP Lvl 2, PSTC: $\approx$ 15 min (AB)
	AMD Versal AI Edge Gen 2 (Commercial SoC)	Cortex-A78AE + R52 Cores + Expansive PL + AI Engines	Full PQC (Accelerated) PL IP Cores + PS OpenSSL 3.5	High functional safety (ASIL-D / SIL-3), optimal workload partitioning across dual real-time cores, application complex, and vector engines.	Power budget (10–25 W) exceeds ultra-low SWaP drone batteries, complex AXI interconnect bus arbitration under heavy multimodal load.	Full PQC Node (Mode A, Accelerated)	ML-KEM-1024 / ML-DSA-87 ZKP Lvl 3, PSTC: 5–10 min (GB)

TABLE X : Cross-Platform Hardware Co-Design, PQC Support Tier, and PARS-ESL Overview

CPS Security

See all 27 papers

Mar 12, 2026

Quantum-Secure-By-Construction (QSC): A Paradigm Shift For Post-Quantum Agentic Intelligence

As agentic artificial intelligence systems scale across globally distributed and long lived infrastructures, secure and policy compliant communication becomes a fundamental systems challenge. This challenge grows more serious in the quantum era, where the cryptographic assumptions built into today's AI deployments may not remain valid over their operational lifetime. Here, we introduce quantum secure by construction, or QSC, as a design paradigm that treats quantum secure communication as a core architectural property of agentic AI systems rather than an upgrade added later. We realize QSC through a runtime adaptive security model that combines post quantum cryptography, quantum random number generation, and quantum key distribution to secure interactions among autonomous agents operating across heterogeneous cloud, edge, and inter organizational environments. The approach is cryptographically pluggable and guided by policy, allowing the system to adjust its security posture according to infrastructure availability, regulatory constraints, and performance needs. QSC contributes a governance aware orchestration layer that selects and combines link specific cryptographic protections across the full agent lifecycle, including session bootstrap, inter agent coordination, tool invocation, and memory access. Through system level analysis and empirical evaluation, we examine the trade offs between classical and quantum secure mechanisms and show that QSC can reduce the operational complexity and cost of introducing quantum security into deployed agentic AI systems. These results position QSC as a foundational paradigm for post quantum agentic intelligence and establish a principled pathway for designing globally interoperable, resilient, and future ready intelligent systems.

Arit Kumar Bishwas, Mousumi Sen, Albert Nieto-Morales +1

PricewaterhouseCoopers USA, San Francisco, CA · Independent Researcher, USA, Fremont, CA

Multi-Agent System Security

Ask PDF

Jun 5, 2026

Blockchain Infrastructure for Intelligent Cyber--Physical--Social Systems:Post-Quantum Security, Interoperability, and Trustworthy Data Economies in the Era of Embodied AI

The deployment of embodied artificial intelligence via world-model-based robotics presents a transformative opportunity for blockchain infrastructure, establishing urgent demand for trustworthy data provenance, cross-organizational governance, and incentive-compatible sharing across decentralized ecosystems. Simultaneously, quantum computing advances recognized by the 2025 Nobel Prize in Physics and the Turing Award threaten the cryptographic primitives securing these data economies, creating an interdependent imperative: long-lived verification for embodied AI depends on crypto-agile architectures capable of withstanding quantum adversaries. This tutorial examines blockchain as the coordination layer bridging this dual transition, from financial substrate to foundational Cyber-Physical-Social Systems infrastructure that simultaneously secures against quantum cryptanalysis and enables scalable, trustworthy data economies. The session opens with an immersive AWS Braket demonstration engaging participants with superconducting, trapped-ion, and neutral-atom hardware to assess cryptographic threat timelines and witness ECDSA-to-post-quantum signature transitions. Five integrated modules progress from embodied AI and world-model requirements through quantum hardware reality and evidence-based security migration, to scalable cross-shard architectures via BrokerChain protocols, trustworthy data economies implementing Croissant metadata standards and robotic learning provenance, and industry ecosystem integration for multi-modal cloud deployment. By bridging quantum hardware realities with embodied AI data requirements, this tutorial charts blockchain as unified infrastructure for next-generation decentralized intelligent environments, providing open-source frameworks and roadmaps for architecting quantum-resistant, interoperable, and data-trustworthy systems.

Song Guo, Huawei Huang, Dongping Liu +2

Hong Kong University of Science and Technology · Sun Yat-sen University · Amazon Web Services +1

Post-Quantum Cryptography

Ask PDF

Apr 28, 2026

Lightweight Quantum Agent for Edge Systems: Joint PQC and NOMA Resource Allocation

In the context of quantum secure scenarios, existing research on mobile edge devices and intelligent computing and edge (ICE) systems based on the Non-Orthogonal Multiple Access (NOMA) communication model have overlooked the energy consumption overhead of Post-Quantum Cryptography (PQC) modules, and the high complexity of traditional resource allocation algorithms fails to meet the demands of real-time decision-making. To address these challenges, this paper proposes a lightweight agentic AI framework designed for online joint optimization within ICE-enabled mobile devices. The scheme constructs a multi-stage stochastic Mixed Integer Nonlinear Programming (MINLP) model that incorporates static power-consumption constraints for PQC modules. Based on Lyapunov optimization theory, the long-term optimization problem is decoupled, and a linear complexity algorithm is proposed to solve the nonconvex challenges of NOMA power allocation . Simulation results verify that the proposed scheme significantly improves computational throughput while ensuring system queue stability and energy consumption constraints. Compared with traditional Successive Convex Approximation (SCA) algorithms, the complexity is reduced to

\mathcal{O}(N)

, achieving a speedup of approximately 46 times when the number of devices

N=35

, thereby meeting the real-time decision-making requirements in dynamic wireless environments.

Yongtao Yao, Wenjing Xiao, Miaojiang Chen +5

School of Computer, Electronics and Information, Guangxi University, and also with the Guangxi Key Laboratory of Multimedia Communications and Network Technology, Nanning 530004, China · School of Computer Science and Engineering, Central South University, Changsha 410083, China · College of Cyber Security, Jinan University, Guangzhou 510632, China +3

Edge Computing Wireless Communications Wireless Resource Allocation Post-Quantum Cryptography

Ask PDF

Abstract

CPS Security

Explore similar work

Quantum-Secure-By-Construction (QSC): A Paradigm Shift For Post-Quantum Agentic Intelligence

Blockchain Infrastructure for Intelligent Cyber--Physical--Social Systems:Post-Quantum Security, Interoperability, and Trustworthy Data Economies in the Era of Embodied AI

Lightweight Quantum Agent for Edge Systems: Joint PQC and NOMA Resource Allocation