Adaptive Linear Filtering Compression on realtime sensor networks

Lightweight Data Compression in Wireless Sensor Networks Using Huffman Coding

“…In this work, however, we consider approaches that attempt to achieve efficient lossless data compression by leveraging solely on the temporal correlation of the data collected by each sensor node and performing all the computations locally, without relying on information from other nodes. Two of the most recent and effective approaches in this category are Marcelloni and Vecchio's lossless entropy compression (LEC) [2,3] and Kiely et al 's adaptive linear filtering compression (ALFC) [4].…”

“…We compare the performance of our approach to that of ALFC using the same set of parameters employed in the experimental evaluation presented in [4]. That is, we used the quantization parameters = 15, = 8, and = 14 and the order of the filter was defined to be = 3.…”

“…As for the number of bits required to represent the measurements without compression, we used = 6, since this is the minimum number of bits required to represent the measurements in our datasets. As suggested in [4], for each packet we computed the optimal value of the variable length coding parameter , which has a major impact on the length of the corresponding coding symbols (ALFC uses th Golomb codes [26], with the restriction that = 2 ). Figure 4 shows the average symbol length ( ) after compression using ALFC and our proposed approach.…”

“…We then describe the proposed lightweight compression method in Section 3. In Section 4, we show the results obtained with our approach and compare them with those obtained when employing the methods in [2][3][4]. Finally, Section 5 concludes the paper.…”

WIP: Towards optimal online approximation of data streams

“…Elmeleegy et al define slide filters as disjoint piecewise approximations that are sequentially adjusted in order to minimize residual error, and this method is most similar to our greedy approximation method. Kiely et al propose an "Adaptive Linear Filtering Compression" algorithm as a lossless compression algorithm for sensor networks, in which the filter aspect is used to predict sample values which are corrected in later transmissions if wrong [9]. In our work, we chose to keep all data segments connected without reverse correction or data prediction.…”