Pipelined Message Authentication Code Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cipher block chaining algorithms that produce message authentication codes are not suitable for pipeline processing due to data dependencies between stages, leading to algorithm latency and potential bottlenecks in data authentication.
Innovation Solution
Divide the cipher block chaining algorithm into segments where each segment produces an intermediate cipher block dependent on a subset of data blocks, allowing pipelining, and then apply XOR and encryption steps to obtain a message authentication code dependent on all data blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a cipher block chaining algorithm is implemented with sequential stage dependencies, then data authentication accuracy is improved, but processing speed and pipeline efficiency deteriorate
Solution Approach 1:
The cipher block chaining algorithm is divided into multiple independent segments or lanes, where each segment processes a subset of data blocks independently. This segmentation breaks the sequential dependency chain while maintaining authentication integrity through final combination of segment results.
Solution Approach 2:
The algorithm transitions from a single-dimensional sequential processing chain to a multi-dimensional parallel structure. Multiple processing lanes operate simultaneously in parallel, with each lane handling a portion of the data blocks, thereby increasing throughput while preserving security properties.
2Productivity
If pipeline processing is implemented for cipher block chaining, then processing throughput is improved, but algorithm complexity and implementation difficulty worsen
Solution Approach 1:
The algorithm is segmented into standardized processing lanes with identical structures, each handling a fixed subset of data blocks. This regular segmentation simplifies pipeline design and control logic compared to arbitrary complex parallelization schemes.
Solution Approach 2:
Data blocks are pre-organized and distributed to specific processing lanes before execution begins. This preliminary arrangement simplifies the pipeline control logic by eliminating dynamic routing decisions during execution, reducing control complexity while maintaining high throughput.
Data Source
AI summary
Methods and apparatus provide for: dividing a quantity of data values into a plurality of blocks; dividing the plurality of blocks into respective sub-sets of blocks; computing an intermediate message authentication code block for each sub-set of blocks by performing a first cipher block chaining algorithm on the blocks of each sub-set of blocks; and computing a message authentication code block for the plurality of blocks by performing a second cipher block chaining algorithm on the intermediate message authentication code blocks.