A Lossless Compression Technique for Huffman-based Differential Encoding in IoT for Smart Agriculture

Abstract

Agriculture faces several uncertain problems in terms of making the best use of its natural resources. As a result, and in light of the growing threat of changing weather conditions, we must closely track local soil conditions and meteorological data to expedite the adoption of culture-friendly decisions. In the Internet of Things (IoT) era, deploying Wireless Sensor Networks (WSN) as a low-cost remote monitoring and management system for these types of features is a viable choice. However, the WSN is hampered by the motes’ insufficient energy sources, which reduces the network’s overall lifespan. Each mote collects the tracked feature regularly and sends the data to the sink for further analysis. This method of transmitting large amounts of data requires the sensor node to use a lot of energy and a lot of network bandwidth. We propose a lightweight lossless compression algorithm based on Differential Encoding (DE) and Huffman techniques in this paper, which is especially useful for IoT sensor nodes that track environmental features, especially those with limited computing and memory resources. Rather than attempting to create novel ad hoc algorithms, we show that, given a general understanding of the features to be monitored, classical Huffman coding can be used to effectively represent the same features that measure at different times and locations. Even though the proposed system does not achieve the theoretical limit, results using temperature measurements show that it outperforms standard methods built specifically for WSNs.