The main goal of this research is to use the algebraic methods used in the network coding process to deal with destructive attacks and enemy contamination to the information transmission network. Random linear network coding is used to increase the throughput of a data transmission network by allowing random linear combination of packets in a network. In this research, in order to increase the operational capacity of the information transmission network and also to deal with the threats of eavesdropping, destruction and contamination of the network, we study a mathematical model of random coding of the network called subspace codes and by using the algebraic methods used in this process, we obtained new structures of subspace codes with optimized parameters, which, in addition to improving the previous results, increase network throughput, reduce information transfer time, increase network safety factor, and reduce energy. And the cost of bandwidth consumption of the network. In particular, we obtain new categories of codes by using specific matrix subgroups.