Protecting data compression: arithmetic coding

Abstract

Data-compression implementations are particularly sensitive to internal faults because most inherent redundancy in the input data is minimised by the source-coding process. Fault-tolerance techniques are presented for protecting a lossless compression algorithm, arithmetic coding, that is vulnerable to temporary hardware failures. The fundamental arithmetic operations are protected by low-cost residue codes, employing new fault-tolerance methods for multiplications and additions, as recently reported. However, additional fault-tolerant design techniques are developed to protect critical steps such as normalisation and rounding, bit stuffing and index selection. These approaches integrate well with residue codes. Normalisation and rounding after multiplication are protected by efficiently modifying the multiplier to produce residue segments. The decoding step that selects the next symbol is checked by comparing local values with estimates already calculated in other parts of the decoding structure, whereas bit stuffing, a procedure for limiting very long carry propagations, is checked by modified residue values. Overhead complexity issues are discussed as rough estimates.