Effects of Dielectric Properties of Human Body on Communication Link Margins and Specific Absorption Rate of Implanted Antenna System

Abstract

Highlights: This is the first work where the carrier link margin and the data link margin of a muscle-implanted antenna system with an external monitoring device are studied. The variation in the specific absorption rate (SAR) was investigated for a wide range of variations in the effective relative permittivity and conductivity of the human body. What are the main findings? The carrier and data links margins of a biocompatible gelatin-encapsulated implantable medical device have been demonstrated. The performance matrices [including signal radiation (gain) and reflection (S11)] within the body model are comparable with the leading implantable antennas. The SAR value (220.26 W/kg) for a 1 g tissue at 1 W input power ensures the patient’s safety after implantation compared with the existing works. ANN modelling was performed to analyze the accuracy of the dependence. The innovative use of ANN in this field further underlines the pioneering nature of this research. What is the implication of the main finding? The reported research findings will aid the innovative development of 5G/6G IoMT devices for existing and/or emerging wearable technology use cases and/or applications. This study examines how the effective dielectric characteristics of the human torso affect the carrier-link-margin (CLM) and data-link-margin (DLM) of a biocompatible gelatin-encapsulated implantable medical device (IMD) that consists of a small implantable antenna, battery, printed circuit board (PCB), camera, and sensor operating at 2.5 GHz. The specific absorption rate (SAR) and the radio frequency (RF) link performances of the IMD are tested for ±20% changes in reference to the mean values of the effective relative permittivity, ɛeff, and the effective conductivity, σeff, of the human body model. An artificial neural network (ANN) with two inputs (ɛeff, σeff) and five outputs (SAR_1 g, SAR_10 g, fractional bandwidth, CLM, and DLM) is trained by 80% of the total scenarios and tested by 20% of them in order to provide reliable dependent analyses. The highest changes in 1 g SAR value, 10 g SAR value, fractional bandwidth, CLM, and DLM at a 4 m distance for 100 Kbps are 63%, 41.6%, 17.97%, 26.79%, and 5.89%, respectively, when compared to the reference effective electrical properties of the homogeneous human body model. This work is the first to accurately depend on the electrical analyses of the human body for the link margins of an implantable antenna system. Furthermore, the work’s uniqueness is distinguished by the application of the CLM and DLM principles in the sphere of IMD communication.