Over the past two decades, the rapid and exponential development of technological innovation has deeply integrated human life with the digital world, largely due to the widespread adoption of the Internet of Things (IoT) and Cyber Physical Systems (CPS). In the most promising applications of the Internet of Things and CPS, wearable sensors have become a transformative technology, fixed on the human body (whether it is soft skin, rigid joints, or various clothing), converting human stimuli (such as strain, pressure, humidity, temperature) into electrical signals, and reflecting key information such as respiratory rate, heart rate, body temperature, and physical exercise through the Internet of Things and CPSs to determine respiratory diseases, life stress and emotions, physical exercise levels, etc. So far, a considerable amount of research has focused on the advancement of resistive strain sensors, and many basic sensing mechanisms that reveal changes in the resistance of strain sensing materials have been elucidated. Representative mechanisms include geometric changes in bulk materials, nano/micro crack propagation in conductive films/layers, disconnection/sliding of nanoscale conductive networks or microstructure contacts, and tunneling effects between conductive fillers. Linearity is usually quantified by the coefficient of determination (R2) obtained from linear regression analysis and is a key parameter of strain sensors. High linearity indicates that strain sensors can more accurately reflect the proportional relationship between electrical signals and applied strain. However, resistive strain sensors typically exhibit nonlinear electrical response or piecewise linear behavior over a wide strain range (such as 0-100% or wider), due to microstructural changes in the conductive path and contact failures, as the sensing material undergoes a morphological transition from a uniform state to a non-uniform state during the stretching process. Nonlinear signals can lead to inaccurate measurements, increasing design costs and the complexity of signal processing circuits, as additional data calibration and compensation are required. In fact, pursuing excellent linearity means expanding the linear response range of sensing materials.

In order to improve the linear range of sensors, efforts have been made to construct stable and high-density conductive networks to ensure continuous conductive paths. For example, P ö tschke and colleagues developed a blend matrix composed of different crystalline phases of polyvinylidene fluoride and polybutylene succinate, as well as their mixed amorphous phases. The amorphous phase localized multi walled carbon nanotubes (MWCNTs) form an efficient conductive network that remains stable near the yield point of the entire conductive polymer composite material (CPC). This strategy defines the linear range of relative resistance variation (Δ R/R0, defined as [R-R0]/R0, where R represents instantaneous resistance (Ω) and R0 corresponds to initial resistance (Ω). Liu and colleagues designed a high-performance strain sensor composed of Ti3C2Tx MXene, Ag nanowires (AgNWs), and liquid metal (LM), which exhibits excellent linearity (R2=0.98157) and a GF (strain factor, defined as Δ R/(R0 ε), where ε is strain) value of 3.22 over a wide strain range of 0-100%. The characteristic of this architecture design is that AgNW forms an interconnected network between MXene sheets, while LM phase maintains a low resistance, serving as a conductive bridge for overlapping MXene/AgNW composites and providing a stable interface for the entire conductive network, effectively improving the stretchability and linear detection range of the sensor. Liu and his colleagues designed a novel dual strain layered structure in three material systems: multi walled carbon nanotube/polyurethane (PU) film, multi walled carbon nanotube/gelatin polyvinyl alcohol (PVA) composite film, and multi walled carbon nanotube/silicone rubber core-shell fiber. During the stretching deformation process, as microcracks gradually initiate within a conductive layer, its resistance significantly increases. However, the complete layer connected in parallel maintains a continuous conductive path, effectively compensating for the damaged conductivity. This innovative structural design ensures a stable conductive network within an extended strain range, significantly reducing the rapid resistance rise typically associated with crack propagation in strain sensors.

Fiber based materials (such as filaments, yarns, fabrics) have good flexibility and excellent mechanical deformation ability, such as stretching, bending, and twisting, which allows them to be easily integrated and attached to the surface of the human body, clothing, and other uneven objects. The addition of conductive components endows fiber based materials with unique electrical response characteristics, significantly expanding their potential applications as wearable sensors in various fields, including but not limited to real-time health monitoring, precise motion detection, and advanced human-machine interface technology. Similar to other CPCs, conductive fibers also face fundamental challenges in balancing sensitivity (characterized by resistance changes) and signal linearity. It is worth noting that fiber based materials with inherent flexibility have unique advantages as they can be designed into different hierarchical structures through advanced textile processing techniques such as weaving, knitting, weaving, and thermal bonding, enabling customization of their electromechanical properties and functional characteristics. For example, Hu and colleagues designed a double threaded structure by winding and heat bonding nylon yarn on spandex, then modifying it with PVA, impregnating it with CNT six times, and encapsulating it with Ecoflex. The obtained strain sensor exhibits a wide linear range (0-100%) with an R2 of 0.991.