Network Coded Parallel Networks and Their Applications

Abstract

This study asserts that parallel network coding is a new communication paradigm that takes advantage of the broadcast using the characteristic of network coding for parallel architectures. Network coding has recently developed as an innovative paradigm for optimization problems of high-scale computation between several nodes of parallel architectures. This study has proposed broadcasting problems in parallel networks using network coding approach to make this communication more efficient. We have proposed a decentralized approach by optimizing parallel communication with the use of network coding to overcome the problem of high communicational and computational time complexity in parallel architectures. To reduce the complexity of such communication, Linear Network Coding (LNC) is implemented in the parallel environment. Parallel architectures involve parallel communication with the aim of receiving the complete information fast with higher information rate at each node. To reduce these complexities of parallel communication author has considered Multi-Mesh of Trees (MMT) architecture for parallel communication. Linear Network Coding (LNC) has been implemented on MMT by proposing Linear-Code Multicast with Parallel Algorithms (LCM-PA) to achieve max-flow from the source to each receiving nodes. We have minimized the communication steps and time complexity involved in transfer of data from one processor to the other processor during parallel communication while achieving the desired throughput in a multicast scenario. This work presents a novel distributed parallel network approach, which is enriched by LNC.