Waveform Data Compression Using Truncated Entropy Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Implantable medical devices (IMDs) face challenges in efficiently collecting, storing, and transmitting data due to limited resources such as memory and power, necessitating effective data compression methods to maximize benefit while minimizing power consumption and memory usage.
Innovation Solution
The implementation of a data compression module within IMDs that employs linear predictive encoding and truncated entropy encoding to compress cardiac data, using a lookup table for Huffman encoding and a rule processor to determine optimal codeword lengths, ensuring lossless compression and efficient storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data compression is implemented in IMDs, then memory usage and power consumption are reduced, but data processing complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing entropy values for all possible 12-bit data values in a lookup table during device manufacturing. This pre-computed entropy information is embedded in the device memory, allowing the compressor to quickly retrieve and apply optimal Huffman codeword lengths without performing complex real-time entropy calculations, thus reducing processing complexity while achieving efficient compression
Solution Approach 2:
The patent introduces an intermediary lookup table that maps data values to their pre-calculated entropy values and corresponding optimal codeword lengths. This intermediary structure acts as a bridge between the raw data and the compression algorithm, eliminating the need for complex real-time entropy computations and simplifying the compression process to table lookups and codeword selections
2Measurement precision
If more memory is allocated for data storage, then data resolution and integrity are maintained, but device size and power consumption increase
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the codeword length parameter based on the entropy of each data value. Instead of using a fixed bit depth for all data points, the system varies the number of bits used to encode each value according to its probability of occurrence, allowing frequent values to use shorter codewords and rare values to use longer codewords, thus optimizing the balance between data fidelity and memory usage
Solution Approach 2:
The patent employs asymmetry in the codeword length distribution, where the majority of data values (those with higher entropy and frequency) are assigned shorter codewords, while a minority of values are assigned longer codewords. This asymmetric encoding strategy ensures that the most common data patterns consume the least memory, achieving efficient compression while maintaining the ability to represent all possible input values with sufficient precision
3Productivity
If compression algorithms are made more complex, then compression ratios improve, but processing time and power consumption increase
Solution Approach 1:
The patent achieves high compression ratios without complex real-time processing by performing the computationally intensive entropy calculations in advance during device manufacturing. The lookup table containing pre-computed entropy values and optimal codeword assignments is stored in the device, enabling the compression algorithm to operate simply by looking up pre-determined values and applying corresponding Huffman codes, thus achieving excellent compression ratios with minimal processing time and power consumption during actual data compression
Data Source
AI summary
Waveforms are digitally sampled and compressed for storage in memory. The compression of the data includes generating a truncated entropy encoding map and using the values within the map to obtain good compression. An encoder further sub-selects values to be encoded and values to remain unencoded to provide an overall compression of the data.