• HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
Friday, November 20, 2026
BIOENGINEER.ORG
No Result
View All Result
  • Login
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
No Result
View All Result
Bioengineer.org
No Result
View All Result
Home NEWS Science News Technology

Integer Wavelet DNA Framework Achieves Lossless Encryption for Tactical Imagery

by
November 20, 2026
in Technology
Reading Time: 6 mins read
0
Integer Wavelet DNA Framework Achieves Lossless Encryption for Tactical Imagery

Integer Wavelet DNA Framework Achieves Lossless Encryption for Tactical Imagery

Share on FacebookShare on TwitterShare on LinkedinShare on RedditShare on Telegram

A research team led by Sriroop Byna, Sujith Gunnam Reddy, and Sobitha Ahila has developed an experimental cryptographic framework tailored for high-resolution tactical images, demonstrating mathematically exact, bit-for-bit lossless reconstruction under a compatible numerical environment. Published in the Journal of King Saud University Computer and Information Sciences, the study addresses a persistent challenge in military image protection, where aerial imagery collected by platforms such as unmanned aerial vehicles must remain entirely free of reconstruction distortion while resisting active cryptanalysis. By substituting standard floating-point transforms with an Integer Haar Lifting Scheme and combining a content-sensitive biological DNA algebra stage with an invertible cross-band wave-diffusion mechanism, the authors reported zero mean squared error upon decryption alongside high ciphertext diffusion. However, dedicated empirical benchmarks revealed that the scheme’s adaptive routing design yielded no measured wall-clock execution advantage over uniform processing on current computing hardware, and the unoptimized software prototype operated twelve to twenty-eight times slower than a conventional baseline block cipher.

Tactical reconnaissance imagery, including Synthetic Aperture Radar captures gathered by unmanned aerial systems, carries dense two-dimensional spatial detail that demands robust confidentiality against interception. Military edge deployments operate under severe computational constraints and finite battery reserves, leaving narrow headroom for resource-intensive cryptographic algorithms. Conventional block ciphers such as the Advanced Encryption Standard operate uniformly over serialized visual streams without exploiting spatial context, encrypting empty background voids with identical mathematical overhead to primary structural objects. Prior research explored chaos-driven maps, dynamic substitution boxes, and biological DNA computing to introduce complex non-linear confusion. A recurring structural pattern across this literature, however, is that algorithms often flatten two-dimensional matrices into one-dimensional vectors or apply heavy biological mutations uniformly across all pixels. This spatial blindness forces identical operational burdens onto irrelevant background pixels, accelerating edge battery drain.

An equally severe vulnerability in continuous chaos and frequency-based image encryption stems from floating-point arithmetic. Standard Discrete Wavelet Transforms and fractional chaotic systems rely on double-precision decimal calculations that inherently introduce minute rounding and quantization errors during inverse transformation. In precision targeting, surveillance, and automated edge-detection contexts, decrypted pixel errors as small as fractions of a percent cause visual speckle noise, leading to target classification failures. To eliminate decimal rounding, the investigators constructed an Integer Haar Lifting Scheme operating strictly on signed sixteen-bit integers without floating-point division. The forward transform separates image arrays into column-wise and row-wise components using integer floor arithmetic, yielding four distinct frequency sub-bands designated as Low-Low, Low-High, High-Low, and High-High. Because intermediate wavelet operations produce signed integer differences ranging from negative 32,768 to positive 32,767, the framework treats the values as unsigned patterns and separates each coefficient into high and low eight-bit bytes, preventing overflow and rounding noise.

To encrypt the packed frequency sub-bands, the framework isolates chaotic key generation from discrete biological transformations through a multi-stage pipeline. The first phase employs a plaintext-bound key derivation protocol built on PBKDF2-HMAC-SHA256 executed with 100,000 iterations to resist dictionary and brute-force guessing attacks. An initial SHA-256 hash computed over the complete plaintext image is combined with a random 32-byte master salt and an explicit color-channel index, deriving distinct 256-bit keys for each individual channel. The first 192 bits of the derived channel key are partitioned into three 64-bit unsigned integers to set the initial coordinates of a continuous three-dimensional Lorenz chaotic differential system, while the remaining 64 bits remain reserved. The chaotic trajectory is solved numerically using an adaptive-order, adaptive-step LSODA predictor-corrector integration routine, discarding the initial 2,000 transient iterations. Instead of multiplying decimal states by scaling factors, the framework applies a bit-level mantissa extraction to the raw IEEE-754 double-precision representation, performing multi-layer exclusive-or mixing and bit-rolling to generate integer keystreams that exhibit near 8.0 bits per byte Shannon entropy.

The central cryptographic engine couples these pseudo-random sequences to a Master-Slave adaptive biological algebra stage. Binary bit pairs are dynamically mapped to the four biological nucleotide bases, designated adenine, cytosine, guanine, and thymine, governed by eight fully reversible biological encoding rules. Algebraic addition, subtraction, and exclusive-or transformations are accelerated across array dimensions using precomputed three-dimensional lookup tables. The Low-Low sub-band carries dominant low-frequency structural approximations and serves as the master control layer for the slave detail bands. To avoid recursive circular dependencies that would prevent decryption, the Low-Low sub-band is encrypted using lightweight reversible DNA exclusive-or operations. An intensity threshold calculated dynamically via Otsu’s method on the unpacked sixteen-bit Low-Low matrix partitions the scene into two distinct spatial domains: structural target regions where pixel intensities exceed the threshold receive heavy non-linear DNA addition to maximize algebraic diffusion, whereas background regions below the threshold are routed through lightweight DNA exclusive-or.

Initial differential evaluations revealed an unexpected cryptographic limitation in the isolated Master-Slave substitution stage. Standard differential cryptanalysis assesses whether altering a single plaintext pixel triggers extensive, unpredictable modifications throughout the resulting ciphertext, quantified by the Number of Pixels Change Rate and the Unified Average Changing Intensity. Under fixed-key conditions, where the encryption key was held completely static rather than re-derived from the altered plaintext, a single flipped pixel bit produced an initial NPCR of roughly 0.0003 percent and a UACI of 0.0001 percent. Because the Master-Slave stage applies per-position substitutions without cross-pixel propagation, the impressive headline differential metrics originally observed stemmed almost entirely from plaintext-bound key derivation recalculating the master salt rather than intrinsic cipher diffusion. Furthermore, minimal chosen-plaintext tests on 64-by-64 images revealed that constant plaintexts caused Otsu’s threshold to collapse, forcing the entire image through the lightweight exclusive-or path, while known structural regions allowed algebraic inversion of the DNA lookup tables to expose the underlying keystream completely.

To resolve this structural vulnerability and provide genuine fixed-key resistance, the authors added an invertible cross-band wave-diffusion mechanism immediately following the Master-Slave stage. The four encrypted sub-bands are stacked vertically into a single two-dimensional array, allowing localized perturbations to propagate across all frequency components. The wave-diffusion process applies a linear cumulative-sum chain combined with non-linear substitution using the standard bijective Advanced Encryption Standard S-box, executing once in forward raster order and once in reverse order. To ensure cryptographic independence, the row and column mixing keys are generated through a dedicated domain-separated PBKDF2 branch and HKDF-style HMAC stream expansion rather than recycling the chaotic keystream. Incorporating this wave-diffusion round elevated fixed-key differential metrics on an urban Synthetic Aperture Radar test scene from 0.0003 percent to 99.71 percent NPCR and 34.01 percent UACI, with twenty-trial averages reaching 99.79 percent NPCR and 33.61 percent UACI, closely matching theoretical cryptographic reference expectations without degrading exact losslessness.

Across experimental evaluations, the proposed framework verified mathematically complete reconstruction alongside strong statistical masking. On the evaluated radar scene, the ciphertext achieved an information entropy of 7.9998 bits, closely approaching the theoretical maximum of 8.0 bits for eight-bit data. Ciphertext adjacent-pixel correlation dropped from strong positive plaintext associations to near-zero values across horizontal, vertical, and diagonal directions, remaining below sample-size reference thresholds. Decryption accurately recovered the original imagery with a Mean Squared Error of exactly 0.0, an infinite Peak Signal-to-Noise Ratio, a Structural Similarity Index of 1.0, and a maximum absolute pixel error of zero. This precision carries an inherent transmission cost: packing sixteen-bit wavelet coefficients into pairs of eight-bit bytes exactly doubles the raw ciphertext payload relative to the original image dimensions, producing an incompressible encrypted byte stream that the authors explicitly accepted as a necessary engineering trade-off for zero distortion.

Despite its mathematical success, empirical investigations disproved the hypothesis that content-aware adaptive DNA routing delivers wall-clock execution savings on modern computing architectures. In a dedicated runtime ablation conducted across three distinct computing systems, routing pixels adaptively using Otsu thresholding failed to accelerate processing over uniform routing. On a Windows 11 desktop, an all-addition pipeline required 6.829 seconds per 256-by-256 test image, whereas adaptive routing took 8.771 seconds—representing an approximately 28 percent slowdown. Isolating the cross-band DNA routine revealed an 11 percent delay, while two independent Linux systems observed no statistically significant execution differences among uniform exclusive-or, adaptive, and uniform addition modes. Vectorized NumPy implementations evaluate precomputed lookup tables across entire arrays simultaneously, meaning subset indexing does not diminish execution time. Instead, the computational overhead of computing Otsu thresholds, generating boolean masks, and maintaining branching structures introduced net latency that outweighed theoretical algebraic reductions.

Overall computational complexity benchmarks further highlighted substantial processing overhead compared with traditional symmetric ciphers. When measured on a 256-by-256 image without hardware cryptographic acceleration, a pure-Python implementation of AES-256 in cipher-block chaining mode completed encryption in an average of 0.672 seconds, achieving a throughput of 95.2 kilobytes per second. In contrast, the unoptimized prototype of the proposed framework averaged 8.100 seconds per image, corresponding to a throughput of 7.9 kilobytes per second—approximately twelve times slower than the baseline cipher on Windows, and twenty-eight times slower in an independent Linux test environment. The dominant runtime consumers were fixed per-call setup routines, specifically the 100,000 PBKDF2 hashing iterations, LSODA numerical trajectory integration, and three-dimensional lookup-table synthesis, which accounted for several seconds before pixel-level DNA operations began.

The authors reported several methodological constraints and defined concrete priorities for ongoing verification. The primary statistical dataset relied on ten synthetic tactical scenes mimicking radar target-to-background distributions rather than downloaded public benchmark suites such as real MSTAR collections. The LSODA differential equation solver exhibits internal step adaptivity and IEEE-754 floating-point dependencies that can yield subtle variations in keystream generation across differing operating systems, compilers, and linear algebra libraries, limiting guaranteed exact reconstruction to mutually compatible numerical environments until platform-independent discretization is finalized. Furthermore, the serialized metadata transmits the SHA-256 plaintext digest to facilitate key derivation, which does not provide cryptographic message authentication and could allow dictionary guessing on low-entropy images. Future investigations aim to port the architecture to compiled C, C++, or FPGA hardware, evaluate authenticated modes against AES-GCM, perform full-resolution multi-query cryptanalysis, and directly measure battery consumption during wireless tactical transmissions.

Subject of Research: Lossless context-aware tactical image encryption using integer wavelets and biological DNA algebra

Article Title: Context-aware tactical image encryption: toward exact (0.0 MSE) reconstruction via integer wavelets and biological DNA algebra

Article References: Byna, S., Gunnam Reddy, S., & Ahila, S. (2026). Context-aware tactical image encryption: toward exact (0.0 MSE) reconstruction via integer wavelets and biological DNA algebra. Journal of King Saud University Computer and Information Sciences, 38(10), Article 1027. https://doi.org/10.1007/s44443-026-01422-x

Image Credits: AI Generated

DOI: 10.1007/s44443-026-01422-x

Keywords: Image Encryption, Synthetic Aperture Radar, Integer Wavelet Transform, Haar Lifting Scheme, DNA Cryptography, Lorenz Chaotic System, Differential Cryptanalysis, PBKDF2, Wave Diffusion, Otsu Thresholding, Tactical Imagery, Unmanned Aerial Vehicles, Ciphertext Entropy, NPCR, UACI, Lossless Reconstruction

News Source: Denise Maddox. (November 20, 2026). Integer Wavelet DNA Framework Achieves Lossless Encryption for Tactical Imagery. Scienmag.

Tags: Ciphertext Entropydifferential cryptanalysisDNA cryptographyHaar Lifting Schemeimage encryptionInteger Wavelet TransformLorenz Chaotic SystemLossless ReconstructionNPCROtsu thresholdingPBKDF2synthetic aperture radarTactical ImageryUACIunmanned aerial vehiclesWave Diffusion
Share12Tweet7Share2ShareShareShare1

Related Posts

Nanostructured Material Designs Drive Performance Advances in Batteries and Supercapacitors

Nanostructured Material Designs Drive Performance Advances in Batteries and Supercapacitors

November 20, 2026
Federated Dynamic Graph Neural Network Framework Improves Epileptic Seizure Prediction Efficacy

Federated Dynamic Graph Neural Network Framework Improves Epileptic Seizure Prediction Efficacy

November 20, 2026

Recurrent Neural Network Framework Detects Serverless Digital Wallet Attacks With High Accuracy

November 20, 2026

Compact Multimodal Transformer Boosts Collaborative Robot Manipulation Success to 87 Percent

November 20, 2026

POPULAR NEWS

  • Study Uncovers Stark Regional and Social Disparities in Mozambique HIV Testing

    Study Uncovers Stark Regional and Social Disparities in Mozambique HIV Testing

    29 shares
    Share 12 Tweet 7
  • Optimized Machine Learning Predicts Sugarcane Bagasse Ash Concrete Compressive Strength Accurately

    29 shares
    Share 12 Tweet 7
  • Stowers Institute Appoints Brandon Mayer as Chief Technology Officer

    29 shares
    Share 12 Tweet 7
  • Activating MAC-1 Receptors Alleviates Brain Inflammation Following Intracerebral Hemorrhage in Preclinical Models

    29 shares
    Share 12 Tweet 7

About

We bring you the latest biotechnology news from best research centers and universities around the world. Check our website.

Follow us

Recent News

Gene Knockouts and Buffer Systems Boost Bacillus subtilis Spore Display Yields

Narrative Review Integrates Imaging, Biomechanics, and Movement in Lumbar Spondylolysis

Bifrontal Brain Stimulation Improves Working Memory in Schizophrenia Without Sex Differences

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 89 other subscribers
  • Contact Us

Bioengineer.org © Copyright 2023 All Rights Reserved.

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In
No Result
View All Result
  • Homepages
    • Home Page 1
    • Home Page 2
  • News
  • National
  • Business
  • Health
  • Lifestyle
  • Science

Bioengineer.org © Copyright 2023 All Rights Reserved.