International Journal of Electrical and Computer Engineering (IJECE) Vol. 12, No. 1, February 2022, pp. 669~683 ISSN: 2088-8708, DOI: 10.11591/ijece.v12i1.pp669-683  669 Journal homepage: http://ijece.iaescore.com Word-based encryption algorithm using dictionary indexing with variable encryption key length Ahmad Al-Jarrah 1 , Amer Albsharat 2 , Mohammad Al-Jarrah 2 1 Department of Applied Science Department, Ajloun University College, Al-Balqa Applied University, Salt, Jordan 2 Department of Computer Engineering, Yarmouk University, Irbid, Jordan Article Info ABSTRACT Article history: Received Jan 5, 2021 Revised Jul 17, 2021 Accepted Aug 4, 2021 This paper proposes a new algorithm for text encryption utilizing English words as a unit of encoding. The algorithm vanishes any feature that could be used to reveal the encrypted text through adopting variable code lengths for the English words, utilizing a variable-length encryption key, applying two-dimensional binary shuffling techniques at the bit level, and utilizing four binary logical operations with randomized shuffling inputs. English words that alphabetically sorted are divided into four lookup tables where each word has assigned an index. The strength of the proposed algorithm concluded from having two major components. Firstly, each lookup table utilizes different index sizes, and all index sizes are not multiples of bytes. Secondly, the shuffling operations are conducted on a two-dimensional binary matrix with variable length. Lastly, the parameters of the shuffling operation are randomized based on a randomly selected encryption key with varying size. Thus, the shuffling operations move adjacent bits away in a randomized fashion. Definitively, the proposed algorithm vanishes any signature or any statistical features of the original message. Moreover, the proposed algorithm reduces the size of the encrypted message as an additive advantage which is achieved through utilizing the smallest possible index size for each lookup table. Keywords: Encryption algorithm word- base Encryption/decryption algorithm English text encoding Two-dimensional shuffling operations This is an open access article under the CC BY-SA license. Corresponding Author: Mohammad Al-Jarrah Department of Computer Engineering, Yarmouk University Irbid, Jordan Email: jarrah@yu.edu.jo 1. INTRODUCTION According to the enormous need for confidentiality in text-data communications, many researchers tackled and developed various techniques for data encryption such as advanced encryption standards (AES) [1], rivest-shamir-aldehman (RSA) [2], triple data encryption standard (TripleDES) [3], blowfish [4], twofish [5], where, each technique has its own merits [6], [7]. Nevertheless, most of these text-encryption techniques depend on having an encryption algorithm implemented on the character level through applying public or private keys [8]. Others used different encryption concepts on the binary level. This paper proposes an encryption technique applied on indices for english words obtained from a lookup table for English words instead of encrypting the words themselves. In algorithms that use an encryption key technique, the encryption keys act as the parameters for mathematical formulas. For example, the well-known RSA algorithm, which is a deterministic one, uses public and private key encryption to encrypt a binary document [3], [9]. Many approaches extended RSA algorithm to encrypt text documents through representing text data in a binary manner [10]. Moreover, some algorithms utilize the concept of a shared dictionary which replaces some repeated characters with certain