Modeling and simulation of a Tamil language encoder for advanced encryption technologies

•A connection between Hardy-Ramanujan prime factors and Tamil letters is reported•A language encoder proposed to replace the pre-round transformation of AES•Tamil can provide a stronger cryptographic strength than English language•Tamil enhance strengths AES for data science applications The misuse modern technology huge threat privacy data, safety machine learning models, security encryption techniques. With advances in natural processing—one areas artificial intelligence—it possible utilize features other languages (other language) current Every has unique features; however, this paper, (TL) considered utilized develop TL advanced standard (AES). establishment one-to-one mathematical relationship letters, formulation TL-based set that satisfies properties Galois field (finite field), deliver significant impact short- long-term goals (privacy, safety, security) learning. This paper presents benefits technologies such as It defines product vowel consonant sounds reveals its mapping. also combined with digits from 1 9, form 28 over an irreducible polynomial degree 8. Additionally, it implements these two properties, models encoder, replaces model strengths. are measured by runs test scores bit sequences ciphers compared language. modeling simulation conclude enhances at every step flow. alphabet phonetic alignment sentences have been robust confidentiality protection techniques cybersecurity. Confidentiality one essential requirements (AES) algorithm currently standardized widely used many cybersecurity applications. first encodes message hexadecimal states then performs several tasks generate confusion diffusion properties—the major defined based on Shannon’s idea.1Dodis Y. Stam M. Steinberger J. Liu T. Indifferentiability confusion-diffusion networks.in: Annual International Conference Theory Applications Cryptographic Techniques. Springer, 2016: 679-704https://doi.org/10.1007/978-3-662-49896-5_24Crossref Scopus (13) Google Scholar Hence, encoding processes; texts using positional information develops vulnerability step, which may propagate through remaining steps algorithm. alphabet, only 26 generates redundancy; hence, difficult Therefore, important study problem leveraging their scientific properties. perfect candidate purpose; studied article. interesting learn how vowels consonants named Tamil. called Uyir Ezhuthu, meaning soul Mei body letters. One way interpret give context language, while context. However, another interpreting be sounds—when spoken proper pronunciation—give body, which, turn, health enhancing circadian rhythms (this just hypothetical statement). formed combining consonants, UyirMei soul+body language—while providing expressive communication—may bring include processing, security, intelligence. carries hidden computational constituents must understood perspectives adapted For instance, power divine sound vibration word ஓம் (ōm) depth understand healing power. regularly Tamils (in particular Sri Lanka) daily dialogue express agreement provides same “yes” dialogues. well-known Indologist, Asko Parpola, mentioned his book contributes Dravidian languages.2Parpola A. Roots Hinduism: Early Aryans Indus Civilization. Oxford University Press, USA2015https://doi.org/10.1093/acprof:oso/9780190226909.001.0001Crossref More about descriptions discussions found page 170 book.2Parpola if properly spoken, speakers Today, we witness contribution physical mental yoga society; significance still explored detail. occupies 17th position top 200 most world—according statistics available https://www.ethnologue.com/guides/ethnologue200 (accessed online January 2022). website informs us 85 million people world. 20 follows: (1,348 M), Mandarin Chinese (1,120 Hindi (600 Spanish (543 Standard Arabic (274 Bengali (268 French (267 Russian (258 Portuguese Urdu (230 Indonesian (199 German (135 Japanese (126 Marathi (99 Telugu (96 Turkish (88 (85 Yue Wu (82 Korean M). Among languages, number closer 256 meaningfully securely occupy 8-bit processor computing device. key player secure storage computer cryptography network reconfigurable hardware applications.3Gura N. Patel Wander Eberle H. Shantz S.C. Comparing elliptic curve rsa cpus.in: workshop embedded systems. 2004: 119-132https://doi.org/10.1007/978-3-540-28632-5_9Crossref (699) Scholar,4Suthaharan S. Calculating primitive: suitable wireless sensor 2008 Australasian Telecommunication Networks Conference. IEEE, 2008: 51-56https://doi.org/10.1109/ATNAC.2008.4783294Crossref (0) Scholar,5Rodriguez-Henriquez F. Perez A.D. Saqib N.A. Koc C.K. A brief introduction cryptography.in: Algorithms Reconfigurable Hardware. 2007: 7-33https://doi.org/10.1007/978-0-387-36682-1_2Google paper6Bharatharaj Alyami Henning M.A. Krägeloh C.U. version fear covid-19 scale.Int. Ment. Health Addiction. 2022; 20: 2448-2459https://doi.org/10.1007/s11469-021-00525-yCrossref PubMed (3) large countries Fiji, India, Malaysia, Mauritius, Singapore, South Africa, Lanka speak In last three decades, migrated Australia, Europe, North America; expected could higher. advancement research benefit globalization. recent years, there growing interest cryptography,7Rajendiran Ibrahim B.S. Pratheesh R. Babu C.N.K. Multilanguage SMS techniques.in: Proceedings Advances Computing. 2013: 455-460https://doi.org/10.1007/978-81-322-0740-5_55Crossref Scholar,8Geetha Padmavathy Thilagam Lallithasree Tamilian cryptography: efficient hybrid symmetric algorithm.Wireless Pers. Commun. 2020; 112: 21-36https://doi.org/10.1007/s11277-019-07013-6Crossref (11) speech recognition,9Lokesh Malarvizhi Kumar P. Ramya Devi Parthasarathy Gokulnath C. An automatic recognition system bidirectional recurrent neural self-organizing map.Neural Comput. Appl. 2019; 31: 1521-1531https://doi.org/10.1007/s00521-018-3466-5Crossref (71) character recognition,10Ulaganathan Rohith Abhinav Vijayakumar V. Ramanathan L. Isolated handwritten convolutional 2020 3rd Intelligent Sustainable Systems (ICISS). 2020: 383-390https://doi.org/10.1109/ICISS49785.2020.9315945Crossref (2) Scholar,11Sigappi Palanivel Spoken strategy language.International Journal Computer Science Issues (IJCSI). 2012; 9: 227-233Google semantics analysis.12Selvam Natarajan Thangarajan Structural parsing text phrase structure model.Int. Inf. Syst. Sci. Eng. 2008; 2: 737-743https://publications.waset.org/vol/15Google Most recently, research, related pandemic, performed population, where indirect influence study,6Bharatharaj Scholar,13Premarajan K. Ashok Rajkumar Naik B. Mental status flood-affected adults rural nadu: cross-sectional study.CHRISMED Res. 6: 97-101https://www.cjhr.org/text.asp?2019/6/2/97/258972Crossref addition cultural geographical contributors. focus article application cryptography. patterns might useful reduction redundancy—caused alphabet—the when applied encrypt messages share them Internet. Geetha et al.,8Geetha example, stated 247 make crack cipher. They mapped (encoded) translated randomly selected 2-bit combination algorithms encoded texts. limited studies connections interpretable environment. helps protect block applying modules.14Heron Advanced (AES).Netw. Secur. 2009; 2009: 8-12https://doi.org/10.1016/S1353-4858(10)70006-4Crossref (123) As standard, require protection. adapts concept takes characters encrypts sequentially established transformations adding keys so plaintext ciphertext cannot recovered adversary. essence, uses multiple rounds each round Sub-Bytes, ShiftRows, MixColumns, AddRoundKey achieve goals. These operated polynomial, multiplication operators, multiplicative inverse, hence delivering algorithm, texts, since much larger English. adds flow Modern leverage intrinsic finite idea,.1Dodis Scholar,15Stallings W. Cryptography Network Security Principles Practices.4th edition. Pearson Education, Inc, 2006Google words corresponding integer representations triggered use GF (28) 8.14Heron Scholar,16Nasser Y.A. Bazzoun Abdul-Nabi implementation simple low cost portable microcontroller.Sixth Digital Information Processing Communications (ICDIPC). 203-207https://doi.org/10.1109/ICDIPC.2016.7470819Crossref (7) contributed via operations, add module, n-bit (n>0) integers together operation additions n following axioms field15Stallings Scholar: closure under addition, associative law commutative existence additive identity, multiplication, distributive law, no zero dividers. operations generally represented tables along inverses satisfying aforementioned axioms. readers encouraged refer materials Dodis al.,1Dodis Stallings,15Stallings Forouzan17Forouzan B.A. Security. McGraw-Hill, 2008Google obtain detailed information. well element appears times (except table), property makes tool words, notably because probability occurrence same, equal 1/2n, n=8 AES. value n, very small, consequently, becomes stronger. popular industry government agencies. represents consists (i.e., 28) 0,1,2,…,255 arithmetic (addition multiplication) pair numbers (00, 01, 02, …, FF) R(x):x8+x4+x3+x+1 8.18Daemen Rijmen Reijndael: standard.Dr. Dobb's Softw. Tools Prof. Program. 2001; 26: 137-139Google There 30 polynomials attached integer.19Stallings standard.Cryptologia. 2002; 165-188Crossref (43) developers list construct inverses. earlier article, assumes byte (8-bits) input bits (plaintext encrypted) evident design algorithm14Heron added inverse highly resistant adversarial attacks attempt recover given ciphertext. our goal not technique but rather show support become fewer intermediate encryption. experimental analysis, framework, effects plaintexts final without keys. analysis presented Figure systematically perform experiments. We know against experiments, interested effect (English Tamil) developing randomness ciphertexts modules. 2 used.Figure 2The examples experimentsShow full captionNote example shows multiples ways writing sentence 7, 8, 9 fill holes plaintext, they will BA, D9, F8 GFT-Grid 5.View Large Image ViewerDownload Hi-res image Download (PPT) Note 5. validated English-based approach. measure statistical process determine sequences.20Alwan L.C. Statistical Process Analysis. McGraw-Hill/Irwin, 2000Google testing hypothesis (H0: sequence random) alternative (H1: ρ follows20Alwan Scholar,21Runs N.I.S.T. Test Detecting Non-randomness.2021https://www.itl.nist.gov/div898/handbook/eda/section3/eda35d.htmGoogle Scholar:μ=2n1n2n1+n2+1σ=2n1n2(2n1n2−n1−n2))(n1+n2)2(n1+n2−1),and (Eqaution 1) ρ=|r−μ|σ,(Eqaution 2) μ sequence, σ deviation runs, n1 n2 zeros ones (1) variable r sequence. NIST,21Runs reject null hypothesis, H0: random 95% confidence ρ>1.96, provided greater 10. 1.96 compare performance encryptions. non-random expose deterministic patterns, adversaries brute force attack, example. Two coding (Python) validate Figures S1 S2. outputs Python modules bottom listing comments. randomness, diffusion, seen placement 0s 1s Tamil-encoded text. “I am going school” experiment stream 1. traditionally location profiles alphabet. 3 information, 4 representations. values text, form, analyzed level score ρ. results row Table 3.102 above 1.96) indicates resulting school.” (without keys) MixColumns calculating previously performed. are, respectively, Tables 2, 3, 4. observe, Sub-Bytes achieves 1.045, ShiftRows 1.559. increased encryption.Figure 4An (decimal hex)17Forouzan ScholarView (PPT)Table 1Comparison after transformationPlaintextMean, μSD, σRuns score, ρRandom (Y/N)English sentence51.8594.4683.102NTamil 164.6095.6000.287YTamil 264.4385.5850.078YTamil 364.6095.6000.109YTamil 463.4385.4960.079YTamil 563.7345.2220.048Y Open table new tab 2Comparison sentence62.3595.4001.045YTamil 165.0005.6350.000YTamil 265.0005.6350.355YTamil 364.9845.6330.003YTamil 564.9845.6330.003Y 3Comparison 165.0005.6350.355YTamil 265.0005.6350.887YTamil 364.9845.6331.068YTamil 464.9385.4960.366YTamil 564.9845.6330.358Y 4Comparison sentence64.7505.6121.559YTamil 164.0005.5461.082YTamil 264.9385.6290.189YTamil 464.0005.5460.180YTamil 564.9845.6330.713Y second versions language—by grid (GFT-Grid) 5—are shown 6. All forms, analyzed, sixth rows five pass tests lower range 0.048 0.287 average 0.1202, significantly (3.102). comparing all 1, say carry help stage algorithm.Figure 6Five school”Show captionThis figure (28). APT-Grid empirically padding texts; padded authenticate or confirm integrity