The heavy dependency of cryptography methods on mathematical theory and computational science necessitates the designing of a computationally secure cryptographic algorithm which is hard to break by an adversary. Matrix methods are intensively used in cryptography for their robustness and adaptability in the encryption of long messages but time complexity is a big concern while dealing with matrices of volumetric size. Further, the key generation function may limit the size of key domain which compromises the perfect security. Our work addresses these hindrances in priority to give an algorithm which is perfectly secure. This work contributes towards the matrix methods in cryptography; the two main attributes of a secured communication namely, true randomness of key selection and a liberty to extend the dimension of key space.