LPRLC: Linear predictive run length coding to improve energy consumption of wireless body area networks

1. Linear Predictive Run Length Coding ( LPRLC) comprises lossy and lossless techniques adjusted to sensor datasets, using the First- Linear Predictive Coding (LPC) to predict at the sensor layer and attain influential accuracy. 2. In the compression phase, because of the data duplication in sensors the Run Length Encoding (RLE) is used to detract data redundancy. 3. LPRLC implemented on the sensor layer has little computational complexity and is appropriate for limited sensor resources. 4. Considering the energy consumption, prediction, and compression that occur in the sensor layer LPRLChas saved up to 98% of the energy by maintaining a appropriate compression rate. A Wireless Body Area Network (WBAN) is a collection of sensors that play a significant role in healthcare. These sensors are designed in small sizes for sending and receiving data. With the increasing volume of these data, energy consumption for processing and transmission rises, leading to energy depletion. One of the most effective ways to reduce data volume is to compress it to save more energy. However, compression methods in WBAN face these problems: 1- They are not compatible with the dataset received from body sensors. 2- They are not suitable for environments with limited energy sources and have computational complexity. 3- They do not consider repetitive and unimportant data. To solve these problems and reduce energy consumption, the Linear Predictive Run Length Coding (LPRLC) is provided, which has three phases. First- Linear Predictive Coding (LPC) for data prediction, Second- Error calculation, Third- Error coding with Run Length Encoding (RLE) for data compression. In the first phase, the LPC method is adapted to the data set received from the body sensors. This method also removes unimportant data using its filter. In the second phase, it significantly reduces the data volume by using a mathematical operation and reducing the computational complexity. Finally, the third phase reduces the data volume in proportion to the repetitive data that the body sensors recorded under the same conditions. The signals attained from sensors include Blood Pressure Systolic (BPsys), Blood Pressure Diastolic (BPdias), Respiration, Oxygen, and Heart Rate, which are measured as a time series. In the first phase, the received signal is predicted continuously, and in the second phase the error resulting is calculated from the actual signal and the expected signal. Eventually, the RLE algorithm compresses the resulting error in the third phase and sends it to the destination. To compare the criteria of Energy Consumption (EC) and Compression Rate (CR), Huffman, Arithmetic, and Lempel-Ziv-Welch (LZW) algorithms are placed instead of RLE. The results show that the RLE algorithm has an average of 98% energy saving and up to 70 times reduction of data volume compared to other algorithms, which has improved 6% in energy consumption and 9 times reduction of data volume.