This study represents an exploration of advantages of encryption of data for security and confidentiality. The significance of encrypting data is of more relevance in light of the mushrooming applications and globalization of communication. Encryption of data is of particular significance or an imperative for applications like email, electronic transactions, digital cash, electronic voting and so on. The study traces the development of Encryption algorithms in terms of their diversity of applications that we feel in our daily lives. Some Encryption algorithms have been developed to make transmission and storage of data more secured and confidential.
The study discusses the distinguishing features of two encryption algorithms and their models in terms of their applications. The first algorithm considers a random matrix key which is multiplied with a ternary vector. A sign function is applied on the product to generate a sequence. This sequence is used to build three different encryption models. Each model can be used for encryption of data. Two models like Model1 & Model2 are based on block cipher technique and Model3 is a stream cipher. In Model 1, the sequence generated is used as subkey to be mapped to alphanumerical equivalent of Plaintext to generate cipher text. The computational power of developed model is low and the model is using sign function which makes it relatively free from Differential cryptoanalysis. In model 2, the sequence generated is divided to basins based on equal values. These basins contain unequal values and based on chosen value corresponding basin is used as subkey. Since theses basins can be considered as variable length keys, the model is relatively free from Linear cryptoanalysis also. In particular, Model3 is based on probabilistic encryption scheme where multiple cipher texts will be generated for one plain text and any one cipher text can be used for transmission of data. Since the model is probabilistic in nature, it is free not only from Differential and Linear cryptoanalysis but also free from Chosen Ciphertext attack. The model is also free from public key attacks. The second algorithm considers not only the key but also time stamps and Initialization Vector to generate a sequence which is used as subkey to generate cipher text. The study outlines the components of these algorithms and their strength in a real time environment.
A similar model is used which identifies any garbled keys while transmission from Key Distribution Centre. This mechanism uses the same algorithm which is used for generation of sub keys, to generate a hash value. The generated hash value is appended to the key and both will be encrypted by either private keys of the participating parties or the master key shared by participating parties and this encrypted message will be transmitted to the participating parties. At the receiver’s side the received hash value and the generated hash value are compared. This provides information about any garbling of key values during transmission.
The core of the study is the analysis and discussion of Developed encryption models. All the models are studied for their performance in terms of computational power, avalanche effect, complexity of the models in terms of their construction and strength. The models are also studied for their improved performance against crypto analysis.
The conclusions drawn above, reflecting the overall security and confidentiality rates of transmitting data, confirm the improvement in the efficiency of transmitting data. The methodology used in the dissertation can be used for evaluating new encryption algorithms in terms of multiple parameters. Further, the quantitative data indicates relationship between Random key considered, sub keys (Basins) generated, computational power needed by the first algorithm and the strength and security of the algorithm. It also identifies the importance of multiple parameters like keys, time stamps and nonce values used in second algorithm in terms of its security & strength.
In the case of Substitution encryption algorithms, the gain by using DES algorithm is its low computational power, which will be very much gained by using the developed models. The developed models are giving almost equal security at low computational overhead (Computing power). As the security of encrypting models is directly related to the key length, the more the key length the more will be the security of the algorithm. But this parameter increases the computational overhead (computing power) of the encryption algorithms. The security of the developed models is relatively free from the key length which gives more flexibility regarding computing power.
The results show that the security to data provided by the developed models can be compared with standard algorithms like RC4 & DES and there is a substantial amount of computing power gain with the developed algorithms.
The methodology used in the dissertation can be used for evaluating the performance of developed Models and their applications in a real time environment. The Models can also be studied for their strengths and limitations.