A Comprehensive Survey of Multiplier and Adder-Based DCT Architectures for Image Compression

Abstract

The Discrete Cosine Transform (DCT) is a fundamental signal-processing technique widely employed in image compression systems due to its ability to transform spatial-domain image data into frequency-domain coefficients and concentrate most of the signal energy into a small number of coefficients. The computational efficiency of DCT is strongly influenced by the implementation of arithmetic components, particularly multipliers and adders, which directly affect hardware area, power consumption, processing speed, and overall system performance. This survey presents a comprehensive review of multiplier and adder-based DCT architectures for image compression, with particular emphasis on hardware-efficient and high-speed implementations. Various DCT architectures, including conventional, fast, pipelined, parallel, and approximate architectures, are reviewed and compared. Different multiplier techniques, such as conventional, Booth, modified Booth, shift-and-add, and constant-coefficient multipliers, are analyzed along with adder architectures including Ripple Carry Adder (RCA), Carry Look-Ahead Adder (CLA), Carry Save Adder (CSA), and Carry Select Adder (CSLA). The survey evaluates existing approaches based on area, power consumption, propagation delay, throughput, hardware complexity, and image quality metrics such as Peak Signal-to-Noise Ratio (PSNR). Furthermore, the trade-offs between computational accuracy and hardware efficiency are discussed to identify suitable architectures for real-time image compression applications. Finally, the survey highlights existing research gaps and emerging opportunities for developing low-power, area-efficient, and high-speed DCT architectures using optimized multiplier and adder structures for FPGA- and ASIC-based implementations.