Cyclic Properties and Pipeline Implementation of the Fletcher Checksum
Zhichang Lin, Sian-Jheng Lin
Abstract
Zhichang Lin, Sian-Jheng Lin
Abstract
The Fletcher checksum (FC) is a means of error detection that is analogous to cyclic redundancy checking but easier to implement in software. Now FC actually has two versions, one is Fletcher's original paper version termed as FC and the other is RFC1146 version termed as FC-RFC. This paper shows that the FC is a cyclic code. This is done by introducing a cyclic code known as the cyclic Fletcher code (CFC) and then showing that the CFC is equivalent to the FC. The clear algebraic structure of cyclic codes is used to analyze the code block length and code distance, and then the single error correction, double error detection, and burst error detection of the CFC are analyzed. Also given is the mathematical relationship between the CFC and a modified FC. Finally, the issue of implementation is considered, and a new scheduling for the CFC is proposed to prevent data hazards. A simulation shows that the proposed scheduling offers improvements of at least 46% in throughput for the CFC and FC-RFC.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The Fletcher checksum (FC) is a means of error detection that is analogous to cyclic redundancy checking but easier to implement in software. Now FC actually has two versions, one is Fletcher's original paper version termed as FC and the other is RFC1146 version termed as FC-RFC. This paper shows that the FC is a cyclic code. This is done by introducing a cyclic code known as the cyclic Fletcher code (CFC) and then showing that the CFC is equivalent to the FC. The clear algebraic structure of cyclic codes is used to analyze the code block length and code distance, and then the single error correction, double error detection, and burst error detection of the CFC are analyzed. Also given is the mathematical relationship between the CFC and a modified FC. Finally, the issue of implementation is considered, and a new scheduling for the CFC is proposed to prevent data hazards. A simulation shows that the proposed scheduling offers improvements of at least 46% in throughput for the CFC and FC-RFC.
Key concepts: Cyclic redundancy check, Checksum, Error detection and correction, Computer science, Cyclic code, Algorithm, Redundancy (engineering), Low-density parity-check code