Lossless Energy Efficient Communication of Temporally Correlated Data using Silent Symbol based Communication
Pratham Majumder, Punyasha Chatterjee, Koushik Sinha · 2023
We propose a lossless energy-efficient communication scheme for transmitting temporally correlated data in UAV-IoT systems. Existing techniques based on data compression or prediction may lead to information loss, which is undesirable in data-critical applications. Our approach ensures lossless transmission while achieving energy savings. Our proposed approach consists of two phases. In the first phase, we exploit temporal redundancy in the data by preprocessing it to create an equivalent dataset with a significantly high percentage of zeros. In the second phase, the data is encoded to have a highly asymmetric symbol frequency distribution. The symbol with the highest frequency, called the null symbol, is not transmitted, resulting in substantial transmitter energy savings. Despite this, the original data can be easily reconstructed at the receiving end, ensuring lossless communication. The transceiver employs an ASK-FSK modulation/demodulation scheme for silent-symbol based communication. Simulation results using real-life datasets demonstrate that our scheme achieves transmitter energy savings of 10.82 times and 13.6 times compared to existing energy-efficient techniques based on Neighborhood Correlation Sequence (NCS) and discrete wavelet transform (DWT), respectively. Additionally, the proposed approach reduces communication time by a factor of two, leading to a corresponding reduction in receiver energy. The lossless nature of our approach makes it particularly suitable for data-critical applications, distinguishing it from existing lossy techniques those are used for eliminating data redundancy.