Multilevel Content Addressable Memory for Pattern Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Content addressable memories (CAMs) are limited in their ability to perform pattern matching and search operations, as they typically compare binary data (0 or 1) without considering valid ranges, which restricts their functionality in multilevel data scenarios.
Innovation Solution
A multilevel content addressable memory (MCAM) system that converts multilevel-bit binary data into digital string data, allowing for valid range definitions and comparisons, using an input encoder, memory array with NAND strings, and an output encoder to generate match signals based on overlapping digital string data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If binary format comparison is used in traditional CAM, then the comparison operation is simple, but the functionality is limited and cannot handle multilevel data scenarios
Solution Approach 1:
The patent changes the parameter representation from binary (0/1) to multilevel digital values (0-15 or higher), allowing each data element to encode multiple binary bits as a single digit. This enables the CAM to handle multilevel data scenarios by comparing digits directly rather than bit-by-bit, significantly improving adaptability while managing complexity through systematic encoding schemes.
2Productivity
If multilevel-bit binary data is converted into digital string data with valid ranges, then pattern matching capability is enhanced, but the coding and searching complexity increases
Solution Approach 1:
The patent segments multilevel binary data into individual digits (0-15 ranges) that can be independently compared. Each digit represents a specific range of binary values, allowing pattern matching to be performed by comparing individual digits rather than processing entire multilevel words sequentially. This segmentation improves pattern matching efficiency by enabling parallel comparison operations while reducing overall complexity through modular processing.
Solution Approach 2:
The patent applies preliminary action by pre-defining valid ranges for each digit position before performing comparisons. The encoding scheme establishes predetermined digit ranges (e.g., 0-15 for 4-bit groups) that are set in advance, allowing the search operation to directly compare digits within these predefined boundaries without complex runtime calculations, thus improving efficiency while managing complexity through pre-computation.
3Productivity
If traditional binary CAM is used, then the device structure is simple, but the search throughput is limited due to inability to compare multiple instances simultaneously
Solution Approach 1:
The patent applies universality by designing the multilevel CAM to perform multiple functions: it can compare exact digit values, perform range comparisons, and handle wildcard patterns all within the same memory array structure. The input encoder and memory array are designed to accommodate various search types simultaneously, enabling the device to process multiple search instances in parallel while managing complexity through a unified multi-functional architecture rather than separate specialized circuits.
4Measurement precision
If digital string data with indicating bits is used to represent exact values, then search accuracy is improved, but the data conversion complexity increases
Solution Approach 1:
The patent extracts the exact value indication function into separate indicating bits that are distinct from the data storage digits. Each digit position has dedicated indicating bits that signal whether the stored value is an exact match candidate or a range value. This extraction improves search accuracy by clearly distinguishing exact values from range values, while managing encoder complexity by separating the indication function from the data storage function, allowing independent optimization of each component.
Data Source
AI summary
A multilevel content addressable memory, a multilevel coding method and a multilevel searching method are provided. The multilevel coding method includes the following steps. A highest decimal value of a multilevel-bit binary data is obtained. A length of a digital string data is set as being the highest decimal value of the multilevel-bit binary data. The multilevel-bit binary data is converted into the digital string data. If a content of the multilevel-bit binary data is an exact value, a number of an indicating bit in the digital string data is the exact value.