Study and Analysis of Matrix Operations in RLNC Using Various Computing

TLDR: This work makes use of different types of computing in order to solve the problem of high computation of matrix operations, and RLNC encoding and decoding achieved higher throughputs than already available approaches.

Abstract: Random linear network coding gained its importance in recent days with its greater potential to enhance the performance of the IoT systems. But the challenging issue is the matrix multiplications and inversions involved in it. Nowadays, with increase in multimedia streaming formats, IoT devices like smartphones will try to make full use of heterogeneous multicore architectures, which are drawing everyone’s attention. The approach presented in this paper is the improvement of matrix operations through optimized operations on matrix blocks. We can schedule the operations on matrix blocks in the heterogeneous cores through directed acyclic graph (DAG). The utilization of computer technology to complete the task is known as computing. It is the process of using computer to complete a given goal-oriented task. Here, we make use of different types of computing in order to solve the problem of high computation of matrix operations. RLNC encoding and decoding achieved higher throughputs than already available approaches.