Adaptive compression of operational commands for remote network management over LPWA
Kodai Tanabe, Go Hasegawa, Gen Kitagata · 2023
When LPWA is employed as a management network in out-of-band management of network equipment, the low power consumption and long communication distance of LPWA provide low operational cost. However, the low datarate of LPWA degrades the performance of remote network management. Therefore, in our work, we aim to improve the effective data rate of LPWA by applying highly efficient compression using command history. We introduce a command tree to efficiently maintain the appearance frequency of each command and its encoded codes. We employ the Huffman coding-based compression method, exploiting biases of used commands. In the proposed method, we dynamically construct the command tree according to the command input from the administrator, which allows the proposed method to be applied to various network equipment regardless of its command grammar. Through experimental evaluation, we show that the compression ratio is roughly 3.5 %, which is significantly better than the traditional compression method, while it has enough small memory usage and computation overhead. Also, the proposed method can provide roughly 10 kbps for the effective datarate over LPWA, which is almost equivalent to standard datarate for management using serial communication.