Source-linked AI summary

Symmetric Algorithm Survey: A Comparative Analysis

Mansoor Ebrahim, Shujaat Khan, Umer Bin Khalid

arXiv:1405.0398v1cs.CR

TL;DR

The paper addresses the need to assess symmetric algorithms’ strengths and limitations for application-specific selection. It compares algorithms across several design and performance criteria, finding AES (Rijndael) best on security, flexibility, memory use, and encryption performance, while noting tradeoffs and algorithm-specific limitations.

  • Problem

    Choosing an algorithm for a particular application requires an assessment of its strengths and limitations across relevant parameters.

    Source: §2 · p. 2
  • Method

    The paper comparatively evaluates symmetric algorithms across architecture, security, scalability, flexibility, reliability, and limitations.

    Source: §5 · p. 7
  • Results

    AES (Rijndael) was found best in security, flexibility, memory usage, and encryption performance, while many other algorithms trade off memory use and encryption performance.

    Source: §5 · p. 7
  • Takeaways & Limitations

    The assessment is intended to make each algorithm’s strengths and limitations transparent for application selection.

    Source: §5 · p. 7

    The paper reports DES vulnerability to linear cryptanalysis and brute force, 3DES exposure to differential and related-key attacks, and difficulty implementing MARS in hardware.

    Source: §3.6.1 · p. 7, §3.6.2 · p. 7, §3.6.11 · p. 7

Abstract

from arXiv · show

Information Security has become an important issue in modern world as the popularity and infiltration of internet commerce and communication technologies has emerged, making them a prospective medium to the security threats. To surmount these security threats modern data communications uses cryptography an effective, efficient and essential component for secure transmission of information by implementing security parameter counting Confidentiality, Authentication, accountability, and accuracy. To achieve data security different cryptographic algorithms (Symmetric & Asymmetric) are used that jumbles data in to scribbled format that can only be reversed by the user that have to desire key. This paper presents a comprehensive comparative analysis of different existing cryptographic algorithms (symmetric) based on their Architecture, Scalability, Flexibility, Reliability, Security and Limitation that are essential for secure communication (Wired or Wireless).

1. INTRODUCTION

The introduction distinguishes symmetric encryption, which uses a shared secret key, from asymmetric encryption, which uses different public and private keys. The paper focuses on evaluating existing symmetric algorithms.

  • Cryptography applies encryption and decryption algorithms to support data security and authentication.The passage describes cryptography as having been designed for military and diplomatic communications and later expanding alongside information technology.Source: §1 · p. 1
  • Symmetric algorithms use a shared secret key with a public algorithm, whereas asymmetric algorithms use separate public and private keys.The paper surveys existing symmetric cryptographic algorithms.Source: §1 · p. 1

2. SYMMETRIC ALGORITHMS

The paper evaluates symmetric algorithms because application choices require understanding their strengths and limitations. Its criteria span architecture, security, flexibility, scalability, and known limitations, assessed through study of algorithm materials and research.

2. SYMMETRIC ALGORITHMS

  • The assessment considers architecture, security, scalability, limitations, and flexibility to inform algorithm choice for particular applications.Scalability includes encryption rate, memory use, software and hardware performance, and computational time.Source: §2 · p. 2
  • The paper argues that scalability must be traded off against other design features and is difficult to compare across ciphers and platforms.Source: §2 · p. 2

2.1.1 Architecture

  • Architecture describes an algorithm’s structure, operations, implementation characteristics, and whether it uses a secret or public key.Source: §2.1.1 · p. 2

2.1.2 Security

  • Security is framed as resistance to attacks, with key size identified as a factor in encryption strength.The text says larger key sizes generally mean stronger encryption.Source: §2.1.2 · p. 2

2.1.3 Flexibility

  • Flexibility concerns whether an algorithm can be modified to meet requirements.Source: §2.1.3 · p. 2

2.1.4 Scalability

  • Scalability is assessed through memory use, encryption rate, software and hardware performance, and computational efficiency.Authors’ scalability claims are discussed alongside simulations reported in prior work [1-3].Source: §2.1.4 · p. 2, §2.2 · p. 2

2.1.5 Limitations (Known Attacks)

  • Known limitations are considered in terms of resource use and vulnerability to attacks.Source: §2.1.5 · p. 2

2.2 Assessment Methodology

  • The methodology studies algorithm code, manuals, source code, and research papers, evaluating each algorithm against the stated parameters.The authors report that no algorithm fully met the evaluation criteria.Source: §2.2 · p. 2

3.1 Overview

The overview sketches the history, use, and distinguishing attributes of several symmetric algorithms, including the AES finalists. It notes the practical strengths and limits of individual designs.

3.1.1 DES

  • DES was initially considered strong, but its short key length and the large volume of data limit its present use.Source: §3.1.1 · p. 2

3.1.2 Triple-DES

  • 3DES applies DES three times with three unrelated keys and was approved for US governmental systems, though AES has replaced it.Source: §3.1.2 · p. 2

3.1.3 Blowfish

  • Blowfish is a fast, compact block cipher with a variable-length key that permits a speed–security tradeoff.It is public domain and used in SSL and other programs.Source: §3.1.3 · p. 2

3.1.4 IDEA

  • IDEA is described as fairly fast, secure, and resistant to linear and differential analysis.Source: §3.1.4 · p. 2

3.1.5 TEA

  • TEA is noted for its simple structure and easy implementation, typically requiring only a few lines of code.Source: §3.1.5 · p. 2

3.1.6 CAST 5

  • CAST-128 is a block cipher used in GPG and PGP editions and by the Canadian Communications Security Establishment.Source: §3.1.6 · p. 2

3.1.7 AES (Rijndael)

  • Rijndael became the AES standard in 2000; the paper describes it as fast, compact, flexible, and adaptable to several processor types.The standard is named AES, while Rijndael is the algorithm described; in practice the algorithm is also called AES.Source: §3.1.7 · p. 2

3.1.8–3.1.11 AES finalists

  • RC6 was an AES finalist and is described as offering security and compatibility, while Serpent placed second to Rijndael and used a conventional security approach.RC6 is patented by RSA Security; Serpent is public and unpatented. Twofish was an AES finalist but was not selected for standardization and is unpatented. MARS was also a finalist, designed with a layered, compartmentalized approach and intended to resist future cryptographic advances.Source: §3.1.8 · p. 3, §3.1.9 · p. 3, §3.1.10 · p. 3

3.2 Architecture

The architecture section compares block sizes, structures, rounds, key lengths, and S-boxes across symmetric algorithms. Most descriptions identify the same algorithm for decryption as for encryption, used in reverse.

3.2.1 DES

  • DES uses a Feistel structure, 64-bit blocks, 16 rounds, and a 56-bit effective key, divided into sixteen 48-bit subkeys.It has eight S-boxes and uses the same algorithm in reverse for decryption.Source: §3.2.1 · p. 3

3.2.2 Triple DES

  • 3DES applies three DES iterations, using 64-bit blocks, 48 rounds, and a 168-bit key.Its description retains the Feistel structure, eight S-boxes, and reverse use of the algorithm for decryption.Source: §3.2.2 · p. 3

3.2.3 Blowfish

  • Blowfish is a Feistel block cipher with 64-bit blocks, 16 rounds, and variable keys up to 448 bits.It has two parts, key expansion and data encryption, and the passage lists four S-boxes and 18 subkeys of 32 bits each.Source: §3.2.3 · p. 3

3.2.4 IDEA

  • IDEA uses a substitution-permutation structure with 64-bit blocks, eight rounds, and a 128-bit key.The description lists 52 subkeys of 128 bits each, no S-boxes, and reverse use of the algorithm for decryption.Source: §3.2.4 · p. 3

3.2.5 TEA

  • TEA is described as a Feistel cipher using 64-bit blocks and a 128-bit key, with 64 rounds and a recommendation of 64 Feistel rounds.It has no S-boxes and uses the same algorithm in reverse for decryption.Source: §3.2.5 · p. 3

3.2.6 CAST

  • CAST uses a Feistel structure, 64-bit blocks, 12 or 16 rounds, and variable 40–128-bit keys.It has four S-boxes and uses the same algorithm in reverse for decryption.Source: §3.2.6 · p. 3

3.2.7 AES (Rijndael)

  • AES uses 128-bit blocks, 10, 12, or 14 rounds, and 128-, 192-, or 256-bit keys; the passage characterizes it as Feistel-based.It describes a single S-box and the same algorithm in reverse for decryption.Source: §3.2.7 · p. 3

3.2.8–3.2.9 AES (RC6 and Serpent)

  • RC6 has 128-bit blocks, 20 rounds, and variable 128-, 192-, or 256-bit keys; Serpent uses 128-bit blocks, 32 rounds, and a substitution-permutation network.RC6 has no S-boxes; Serpent has eight. Both descriptions specify reverse use of the algorithm for decryption.Source: §3.2.8 · p. 3, §3.2.9 · p. 3

3.3 Security

The security discussion reports different strengths and attack boundaries across the surveyed ciphers, while noting DES's practical weakness and several algorithms' resistance to specified attacks.

3.3.1 DES; 3.3.2 TDES

  • DES's 56-bit key was recovered in an EFF demonstration in under a day; TDES uses a 168-bit key and is reported more secure than DES.Despite DES's exposure risk, the paper says it may still be preferred for backward compatibility and upgrade costs.Source: §3.3.1 · p. 4, §3.3.2 · p. 4

3.3.3 Blowfish

  • Blowfish uses 128–448-bit keys; the paper reports fewer cryptanalysis attempts than for other algorithms and resistance to differential related-key attacks.Its key-derived round keys are described as independent, making such attacks complicated or infeasible.Source: §3.3.3 · p. 4

3.3.4 IDEA; 3.3.5 TEA

  • IDEA resists differential cryptanalysis under stated hypotheses, while its best reported all-key attack applies to a reduced six-round version.TEA is described as having IDEA-level security with a simple structure and easy implementation.Source: §3.3.4 · p. 4, §3.3.5 · p. 4

3.3.6 CAST; 3.3.10 AES (Twofish)

  • CAST uses variable key sizes and is reported resistant to linear and differential attacks; Twofish is described as resistant to related-key attacks, with no weak keys.The Twofish discussion specifically includes slide and related-key differential attacks.Source: §3.3.6 · p. 4, §3.3.10 · p. 5

3.3.7 AES (Rijndael)

  • Rijndael supports keys up to 256 bits, and reported attacks on concentrated-round versions were not practically possible.The cited attacks include Square, Improved Square, Impossible Differential, and Reversed Key Schedule attacks.Source: §3.3.7 · p. 4

3.3.8 AES (RC6)

  • A 16-round RC6 linear attack would require 2^119 known plaintexts, while differential cryptanalysis is reported to work with more than 12 rounds.The paper also associates RC6 security with random-looking output bits at 15 rounds or fewer.Source: §3.3.8 · p. 4

3.3.9 AES (Serpent)

  • Serpent extends 16 rounds to 32 for protection against future cryptanalytic discoveries and is usually changed before 2^64 blocks are encrypted.The paper says Serpent with half its rounds is still as secure as three-key triple DES.Source: §3.3.9 · p. 5

3.3.11 AES (MARS)

  • MARS's security is attributed to data rotations and complex middle rounds; the paper reports resistance to related-key, differential, and timing attacks.It is also described as offering enhanced security and speed compared with DES and triple DES.Source: §3.3.11 · p. 5

3.4 Scalability

The section compares scalability through memory use and encryption performance, reporting TEA as strongest on those measures while emphasizing trade-offs with security and other design features.

3.4 Scalability

  • The paper defines memory use by the number of functions performed and encryption rate as processing time for a given data size.It says encryption rate depends on processor speed and algorithm complexity, and that smaller values are desired.Source: §3.4 · p. 5
  • The comparison draws on prior research analyzing performance, memory use, and hardware implementation issues across common platforms.The paper describes a summary of scalability performance for 11 popular symmetric algorithms.Source: §3.4 · p. 5
  • Scalability must be traded off against security, architecture, flexibility, and robustness, and comparing or designing ciphers that scale across platforms is difficult.The section evaluates memory usage and encryption performance, including encryption and key scheduling.Source: §3.4 · p. 5
  • TEA is reported best for encryption performance and memory use across the compared algorithms, but its security is compromised and it is currently obsolete.Figure 2a covers pre-AES algorithms; Figure 2b covers the five AES finalists.Source: Figure 2a · p. 5, §3.4 · p. 5, Figure 2b · p. 6

3.5 Flexibility

Flexibility is treated as an algorithm's ability to accept modifications according to requirements, and Table 2 summarizes symmetric algorithms on that dimension.

3.5 Flexibility

  • The section defines flexibility as an algorithm's ability to accept modifications according to requirements.The paper uses this as the basis for analyzing the algorithms in Table 2.Source: §3.5 · p. 6
  • Flexibility is framed as a comparison dimension rather than a specific modification or performance result.The section's stated criterion is modification according to requirements.Source: §3.5 · p. 6
  • Table 2 is presented as a summary of symmetric algorithms' flexibility.The supplied passage identifies the table's scope but gives no specific comparison outcomes.Source: Table 2 · p. 6, §3.5 · p. 6

3.6 Limitation

The section lists attack exposures and implementation constraints across symmetric algorithms, from DES and 3DES to the AES finalists. The limitations range from weak-key and reduced-round vulnerabilities to speed, complexity, and hardware concerns.

3.6 Limitation

  • DES is vulnerable to linear cryptanalysis, weak keys, and brute-force attacks, while 3DES is exposed to differential, related-key, and meet-in-the-middle attacks.The section also notes certain variations of meet-in-the-middle attacks against 3DES.Source: §3.6.1 · p. 7, §3.6.2 · p. 7

3.6.3 Blowfish

  • Blowfish has weak-key classes, and its 4-round version is exposed to second-order differential attacks.The authors question Blowfish reliability because of its large number of weak keys.Source: §3.6.3 · p. 7

3.6.4 IDEA

  • IDEA has weak-key and reduced-round susceptibilities; its first three rounds are especially exposed to key attacks.The passage also reports exposure to collision attacks and key-schedule and related-key differential timing attacks.Source: §3.6.4 · p. 7

3.6.5 TEA

  • TEA’s equivalent keys reduce its effective key size to at least 126 bits, and it is also exposed to a related-key attack.The passage specifies 2^23 chosen plaintexts and complexity 2^32 for the related-key attack.Source: §3.6.5 · p. 7

3.6.6 CAST 5

  • CAST5’s reduced 64-bit-key version is susceptible to a differential related-key attack.The passage gives 2^17 chosen plaintexts, one related-key query, and offline work of 2^48 for the attack.Source: §3.6.6 · p. 7

3.6.7 AES (Rijndael); 3.6.8 AES (RC6)

  • AES (Rijndael) has no serious weakness reported, but its inverse-cipher implementation is inappropriate on smart cards; RC6 has a weak-key class with incomplete arbitrariness up to 17 rounds.The AES passage mentions a mathematical property that might be vulnerable to an attack, but says it is not itself an attack.Source: §3.6.7 · p. 7, §3.6.8 · p. 7

3.6.9 AES (Serpent)

  • Serpent has no reported limitation, but its 32 rounds make it slower and more complex to implement on small blocks.Source: §3.6.9 · p. 7

3.6.10 AES (Twofish); 3.6.11 AES (MARS)

  • Twofish may be susceptible to chosen-key attacks in certain implementations, such as hash functions, while MARS is complex to analyze and difficult to implement in hardware.The section reports no significant limitations for MARS.Source: §3.6.10 · p. 7, §3.6.11 · p. 7

4. COMPARATIVE ANALYSIS

The comparison reports that AES (Rijndael) is the most secure, faster, and better-performing algorithm overall, while other algorithms show tradeoffs and weaknesses. DES and IDEA are slow in software, and the AES finalists have distinct potential weaknesses or implementation concerns.

4. COMPARATIVE ANALYSIS

  • AES (Rijndael) was found the most secure, faster, and better among the algorithms, with no serious weaknesses reported.The comparison also describes it as best in security, while noting a numerical property that might be exposed to attack.Source: §4 · p. 7
  • DES and IDEA are slow in software, while Triple DES is described as potentially safest but slowest and Blowfish as faster than DES and IDEA.The passage says DES is best suited to hardware implementation and notes weak-key concerns for DES and IDEA.Source: §4 · p. 7
  • The AES finalists had no serious weakness identified, but the comparison notes concerns involving Rijndael, RC6, Serpent, Twofish, and MARS.The cited concerns include a possible Rijndael numerical-property attack, incomplete RC6 arbitrariness, Serpent speed and implementation complexity, possible Twofish chosen-key attacks, and MARS analysis complexity.Source: §4 · p. 7

5. CONCLUSION

The paper concludes that AES (Rijndael) performed best in security, flexibility, memory usage, and encryption performance. Other algorithms were competent, but most involved tradeoffs between memory usage and encryption performance.

5. CONCLUSION

  • AES (Rijndael) was best in security, flexibility, memory usage, and encryption performance among the analyzed symmetric block algorithms.Source: §5 · p. 7
  • Most other algorithms showed a tradeoff between memory usage and encryption performance, while some were compromised.Source: §5 · p. 7
Loading 1405.0398v1…