Function Approximation Using Analog Building Blocks in Flexible Electronics

2025-02-03

AI Summary

To address the energy-efficiency bottleneck in function approximation for resource-constrained flexible electronics, this work proposes an analog circuit primitive–based continuous function approximation architecture: it constructs spline functions using analog adders, multipliers, and squarers, and further implements Kolmogorov–Arnold networks (KANs). This is the first integration of analog hardware primitives with KAN theory, enabled by parasitic-aware modeling and analog spline synthesis, achieving low-overhead continuous mapping on a single chip. Compared to an 8-bit digital spline implementation, the proposed design reduces silicon area by 125× and cuts power consumption by 10.59%, while maintaining approximation error ≤7.58%. The work overcomes fundamental area and energy constraints of conventional digital approximation methods in flexible electronics, and experimentally validates the feasibility and superiority of analog KANs for ultra-low-power neuromorphic edge computing.

Technology Category

• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Analogy
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Function approximation is crucial in Flexible Electronics (FE), where applications demand efficient computational techniques within strict constraints on size, power, and performance. Devices like wearables and compact sensors are constrained by their limited physical dimensions and energy capacity, making traditional digital function approximation challenging and hardware-demanding. This paper addresses function approximation in FE by proposing a systematic and generic approach using a combination of Analog Building Blocks (ABBs) that perform basic mathematical operations such as addition, multiplication, and squaring. These ABBs serve as the foundation for constructing splines, which are then employed in the creation of Kolmogorov-Arnold Networks (KANs), improving the approximation. The analog realization of KAN offers a promising alternative to digital solutions, providing significant hardware benefits, particularly in terms of area and power consumption. Our design achieves a 125x reduction in area and a 10.59% power saving compared to a digital spline with 8-bit precision. Results also show that the analog design introduces an approximation error of up to 7.58% due to both the design and parasitic elements. Nevertheless, KANs are shown to be a viable candidate for function approximation in FE, with potential for further optimization to address the challenges of error reduction and hardware cost.

Problem

• Flexible electronics
• Function approximation
• Energy efficiency

Innovation

• Kolmogorov-Arnold Networks
• Analog Components
• Function Approximation