<p>The proliferation of image data in the Internet of Things applications has introduced significant security challenges. Traditional text-based algorithms fall short of effectively securing intricate image data. In contrast, lower-dimensional chaotic maps, being less resource-intensive, offer a promising security solution. However, these maps are constrained in their chaotic behavior and nonlinear dynamics. This paper utilizes an infinitely chaotic one-dimensional enhanced Logistic map and several lightweight techniques to devise a novel image encryption scheme for safeguarding low-end devices. The methods employed include region of interest selection, lossy compression, adaptive key generation, and a novel diffusion–confusion–diffusion encryption algorithm. The proposed image encryption scheme has been evaluated in MATLAB against various performance and security metrics, including compression ratio, PSNR, SSIM, MSE, NPCR, UACI, encryption, decryption, compression times, correlation, histogram analysis, histogram variance, and entropy. The proposed scheme has also been assessed for resilience against cropping, noise, histogram equalization, low-pass filtering, and similar attacks. The scheme demonstrates strong cryptographic performance, achieving an average encryption time of <InlineEquation ID="IEq1"> <EquationSource Format="TEX">$0.078$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>0.078</mn> </mrow> </math></EquationSource> </InlineEquation>s and decryption time of <InlineEquation ID="IEq2"> <EquationSource Format="TEX">$0.074$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>0.074</mn> </mrow> </math></EquationSource> </InlineEquation>s, with average UACI<InlineEquation ID="IEq3"> <EquationSource Format="TEX">$= 32.01\%$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mo>=</mo> <mn>32.01</mn> <mo>%</mo> </mrow> </math></EquationSource> </InlineEquation>, NPCR<InlineEquation ID="IEq4"> <EquationSource Format="TEX">$= 99.50\%,$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mo>=</mo> <mn>99.50</mn> <mo>%</mo> <mo>,</mo> </mrow> </math></EquationSource> </InlineEquation> and image entropy <InlineEquation ID="IEq5"> <EquationSource Format="TEX">$= 7.937$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mo>=</mo> <mn>7.937</mn> </mrow> </math></EquationSource> </InlineEquation>, together with average MSE <InlineEquation ID="IEq6"> <EquationSource Format="TEX">$= 1.027 \times 10^{4}$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mo>=</mo> <mn>1.027</mn> <mo>×</mo> <msup> <mn>10</mn> <mn>4</mn> </msup> </mrow> </math></EquationSource> </InlineEquation>, SSIM <InlineEquation ID="IEq7"> <EquationSource Format="TEX">$= 0.0075$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mo>=</mo> <mn>0.0075</mn> </mrow> </math></EquationSource> </InlineEquation>, and PSNR <InlineEquation ID="IEq8"> <EquationSource Format="TEX">$= 8.14$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mo>=</mo> <mn>8.14</mn> </mrow> </math></EquationSource> </InlineEquation>dB, all indicating effective encryption. Its resilience to attacks is evident from consistently high PSNR and SSIM values and minimal MSE, confirming robust security. The results highlight high sensitivity, strong randomness, fast execution, and resistance to attacks, making the scheme well suited for secure, real-time IoT environments. The scheme proves effective and resilient, offering a viable solution for securing image data in resource-limited, time-sensitive Internet of Things applications.</p>

Securing IoT applications with enhanced logistic map-based adaptive lightweight image compression and encryption

期刊论文
发表日期：

2026年3月2日

查看全文

Mir Nazish,
Munika Javid,
M. Tariq Banday

摘要

The proliferation of image data in the Internet of Things applications has introduced significant security challenges. Traditional text-based algorithms fall short of effectively securing intricate image data. In contrast, lower-dimensional chaotic maps, being less resource-intensive, offer a promising security solution. However, these maps are constrained in their chaotic behavior and nonlinear dynamics. This paper utilizes an infinitely chaotic one-dimensional enhanced Logistic map and several lightweight techniques to devise a novel image encryption scheme for safeguarding low-end devices. The methods employed include region of interest selection, lossy compression, adaptive key generation, and a novel diffusion–confusion–diffusion encryption algorithm. The proposed image encryption scheme has been evaluated in MATLAB against various performance and security metrics, including compression ratio, PSNR, SSIM, MSE, NPCR, UACI, encryption, decryption, compression times, correlation, histogram analysis, histogram variance, and entropy. The proposed scheme has also been assessed for resilience against cropping, noise, histogram equalization, low-pass filtering, and similar attacks. The scheme demonstrates strong cryptographic performance, achieving an average encryption time of $0.078$ $0.078$ s and decryption time of $0.074$ $0.074$ s, with average UACI $= 32.01\%$ $= 32.01 %$ , NPCR $= 99.50\%,$ $= 99.50 %,$ and image entropy $= 7.937$ $= 7.937$ , together with average MSE $= 1.027 \times 10^{4}$ $= 1.027 \times 10^{4}$ , SSIM $= 0.0075$ $= 0.0075$ , and PSNR $= 8.14$ $= 8.14$ dB, all indicating effective encryption. Its resilience to attacks is evident from consistently high PSNR and SSIM values and minimal MSE, confirming robust security. The results highlight high sensitivity, strong randomness, fast execution, and resistance to attacks, making the scheme well suited for secure, real-time IoT environments. The scheme proves effective and resilient, offering a viable solution for securing image data in resource-limited, time-sensitive Internet of Things applications.