1 / 11
CPG No. 187  |  BE Third Year – CSE
SentinelMesh

A Secure & Tamper-Resistant Communication Framework for UAV Swarms with Blockchain-Based Logging

TIET Patiala  |  Dept. of CSE  |  March 2026

Mentor: Dr. Garima Singh, Assistant Professor

UAV Swarm ECC+ASCON ML + Trust Blockchain Log Dashboard
Team Members
AS

Arjun Singh

102317110

PP

Priyanshu Patel

102303595

AV

Arpita Verma

102303001

BB

Bhumika Bandyopadhyay

102315167

KG

Krrish Garg

102303698

Project Overview

What is SentinelMesh?

System Architecture

Dashboard & Monitoring Real-time swarm health & trust scores Blockchain Logging Layer Private chain · Async queue · SHA-256 Trust & Anomaly Detection Isolation Forest · Dynamic trust scores Cryptographic Security Layer ECC P-256 key exchange · ASCON-128 UAV Swarm Communication Decentralised mesh · Wireless inter-drone Fig 1. SentinelMesh architecture layers

Open Wireless

UAV comms lack auth & encryption — vulnerable to attacks.

No Unified Solution

Crypto, detection & logging handled in isolation only.

Resource Constrained

Heavy solutions don't fit UAV-class hardware.

No Trust Layer

No mechanism to distinguish legitimate from rogue drones.

No Accountability

Existing systems lack tamper-proof audit trails.

SentinelMesh

Five integrated security layers in one real-time system.

Literature Survey

Related Work & Research Gap

Literature Coverage Map Research Gap No unified real-time system UAV Security [1][2] Vulnerabilities mapped ASCON Crypto [3][8] Lightweight encrypt ECC [4] Small key sizes ML Anomaly [5] Behaviour analysis Blockchain [7] Tamper-proof logs Trust Systems [6] Node isolation Fig 2. Literature coverage map & identified gap

[1][2] UAV Communication Security

Vulnerabilities documented. No integrated real-time protection exists.

[3][8] ASCON Lightweight Crypto

NIST-standard ASCON — efficient authenticated encryption for constrained devices.

[4] ECC vs RSA

ECC gives equivalent security with 12× smaller keys — ideal for UAVs.

[5][6][7] ML, Trust & Blockchain

Each effective in isolation. None combined into a unified UAV pipeline.

Gap Identified

No work unifies all five layers into one real-time system for UAV swarms.

Problem Statement & Objectives

Threats Addressed & Project Objectives

UAV Attack Surface

Drone A Legitimate Drone B Legitimate Drone C Legitimate ROGUE Attacker Replay Attack Rogue Drone Cmd Injection Fig 3. UAV swarm attack surface
O1

Lightweight Secure Communication

ECC + ASCON/AES layer ensuring authenticity & confidentiality with minimal compute.

O2

Anomaly Detection & Trust Scoring

Lightweight ML flags rogue drones; dynamic trust isolates threats in real time.

O3

Tamper-Proof Blockchain Logging

Private blockchain with async batching ensures integrity & traceability.

O4

Integrated Scalable Architecture

All modules unified into one modular, real-time-capable system.

Assumptions & Constraints

Assumptions & Constraints

Assumptions

Software Simulation No physical UAV hardware required Private Mesh Net Partial disconnections handled by protocol Known Swarm Size Dynamic join/leave via trust module ECC Pre-Provisioned Key pairs loaded before deployment Python Simulation Env NetworkX · Flask · scikit-learn PyCryptodome · ASCON Fig 4. System assumptions

Constraints

Limited Compute & Memory UAV-class hardware only Real-Time Requirement No latency disrupting ops No Cloud Dependency Fully on-device operation Blockchain Write Overhead Async + batched — non-blocking Fig 5. System constraints chain
Project Execution Plan

System Pipeline — SentinelMesh Flowchart

Drone Message ECC P-256 Key Exchange ASCON-128 Encryption Sentinel Monitor Node ML Anomaly Detection Trust Score? NORMAL Continue ops High Trust MALICIOUS Isolate drone Low Trust Async Queue Event logged Blockchain Immutable Log Auth Encrypt Monitor Detect Decide Queue Log Fig 6. SentinelMesh complete system pipeline flowchart
Project Requirements

Technology Stack

Presentation Layer Flask · React · Matplotlib · NetworkX Blockchain Layer Hyperledger Fabric · SHA-256 · Async Queue ML & Trust Layer scikit-learn · Isolation Forest · NumPy · Pandas Cryptography Layer ASCON-128 · ECC P-256 · PyCryptodome Core Runtime Python 3.10+ · JavaScript · Solidity (opt.) Fig 7. Technology stack layers

Why ASCON?

NIST-standardised for constrained devices. Authenticated encryption in one pass — lower energy than AES-GCM.

Why ECC?

256-bit ECC equals 3072-bit RSA security with 12× smaller keys — critical for UAV memory limits.

Why Isolation Forest?

Unsupervised ML — detects anomalies without labelled data. Low memory footprint for UAV hardware.

Why Private Blockchain?

Permissioned chain avoids public-network latency. Async batching keeps writes off the critical path.

Hardware (Optional)

Raspberry Pi 4 and ESP32 modules for physical validation. Software simulation is the primary deliverable.

Project Outcomes

Expected Deliverables

SentinelMesh Integrated System D1 Secure Comms ECC + ASCON framework D2 Trust System ML anomaly detection D3 Blockchain Log Tamper-proof records D4 Dashboard Real-time monitoring UI D5 Simulation Prototype Attack scenario validation Fig 8. Deliverable map

D1 — Secure Communication Framework

ECC + ASCON/AES ensuring confidentiality, authentication, and real-time performance on constrained hardware.

D2 — Anomaly Detection & Trust System

ML pipeline scoring drones dynamically; rogue & compromised nodes isolated without disrupting the swarm.

D3 — Blockchain Logging Mechanism

Async batched private chain — tamper-proof, traceable records of all events including attacks.

D4 & D5 — Dashboard + Simulation Prototype

Full attack-scenario validation (replay, rogue drone, command injection) + real-time monitoring UI.

Work Plan

Project Timeline — January to December 2026

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Research & System Design Simulation Development Security Implementation Detection & Trust System Blockchain Logging Integration & Demo TODAY Fig 9. Project Gantt chart — Jan to Dec 2026. Diamonds mark phase milestones.
Team Contributions

Individual Roles & Responsibilities

SentinelMesh Integrated System AS Security PP Blockchain AV ML KG Dashboard BB UI + Test Fig 10. Team role-to-module mapping
AS

Arjun Singh — Security

ECC key exchange; ASCON/AES encryption module; inter-drone authentication protocol.

PP

Priyanshu Patel — Blockchain

Private blockchain; async queue logging; SHA-256 block structure; batching optimisation.

AV

Arpita Verma — Machine Learning

Isolation Forest anomaly pipeline; trust score algorithm; malicious node isolation logic.

BB

Bhumika Bandyopadhyay — UI & Testing

Simulation QA; attack scenario scripting (replay, rogue, cmd injection); UI validation.

KG

Krrish Garg — Frontend & Dashboard

React dashboard; real-time trust visualisation; swarm topology display; health monitoring.

References

References

[1] UAV Cybersecurity Survey

Niyonsaba & Konate — Semantic Scholar. Spoofing & replay vulnerabilities in UAV networks.

[2] FANETs Security Survey

IEEE Xplore, doc. 10329156. Security & privacy in Flying Ad Hoc Networks.

[3] NIST ASCON Standard (2023)

nist.gov. ASCON selected as lightweight cryptography standard for constrained devices.

[4] RSA vs ECC Performance

IJNS Vol. 20, No. 4, pp. 625–635. ECC equivalent security with 12× smaller keys.

[5] ML Intrusion Detection for UAVs

MDPI Electronics 10(13). Machine learning enabled IDS for UAV network security.

[6] Trust Management in FANETs

ResearchGate, pub. 381793815. Trust-based flying ad-hoc network security analysis.

[7] Blockchain Security for UAVs

IEEE Xplore, doc. 10855475. Blockchain-based security architecture for UAV systems.

[8] Lightweight Crypto in UAVs

TACS KPI, article 326898. Comparative analysis of lightweight cryptography in UAV systems.

SentinelMesh  |  CPG No. 187  |  TIET Patiala, March 2026

Thank you. Questions are welcome.