Efficient sensor data compression and restoration method and apparatus for low-speed wireless communication

Low-computation IoT sensor data compression using overwrite method for difference values relative to segment start values

This technology stores the start value of sensor data, groups periods where the absolute difference value increases or remains constant, and records only the absolute difference, sign, and overwrite count relative to the start value per segment to compress and restore data according to packet size.

IoT wireless systems are low-power and low-bandwidth, limiting packet sizes to 128–250 bytes, with a one-second latency for transmission and acknowledgment. Sensors with rapidly changing values, such as vibration, acceleration, and gyro sensors, generate large data volumes, leading to increased transmission frequency and latency, while low-performance processors struggle with complex compression operations.

This technology reduces data size through simple operations—overwriting difference values in the same storage space and counting occurrences—minimizing compression overhead and enabling real-time compressed transmission. It can be applied to low-speed wireless sensor networks like equipment vibration monitoring, structural acceleration measurement, and wearable motion sensing, reducing both transmission frequency and latency for battery-powered devices.

‍

‍

Key Features:

• A step of storing the start value of sensor data and determining segments based on the start value and subsequent sensor data values
• A step of storing the absolute value of the difference between the sensor data and the segment start value, the sign of the difference, and the overwrite count for each determined segment
• A configuration that overwrites the absolute value of the newly calculated difference in the storage space of the previous absolute difference until a packet of a predetermined size is formed
• A configuration where the start value is stored in 2 bytes, the sign is stored in the most significant bit of the next 2 bytes, the overwrite count in the remaining 7 bits, and the absolute difference in the remaining 1 byte

‍

This invention was developed with research support from the National Research Foundation of Korea for a real-time condensation detection and prediction system considering user residential behavior patterns.

‍