Engineering advancement over the last decade has significantly benefited the medical field, facilitating personalised and accessible healthcare. Various portable systems have been developed to obtain diagnostic parameters without the necessity for sedation or immobilisation of the patient, and even their presence at the medical facility. This can be especially important for populations that are at greater medical risk and are unable to undergo sedation, as well as for overall screening of the underlying conditions by continuous monitoring. This thesis provides engineering solutions aimed at improving the reliability of one of the modern medical diagnostic techniques – Wireless Capsule Endoscopy (WCE). It is a non-invasive approach for gastrointestinal (GI) tract examination that involves the natural propagation of a capsule through the entire tract of the patient while recording images of the lining. The video data are transmitted to a receiving unit outside the body, which is then accessed by medical practitioners for appropriate diagnostics. The Wireless Body Area Network (WBAN) technology describes the acquisition and transmission of the signal. One of the major challenges associated with WCE is the accurate localisation of the capsule due to the transit time being different for each individual. Localisation methods based on various physical principles are still under investigation by researchers. In this study, radiofrequency (RF) signal propagation analysis was used to provide accurate received signal strength (RSSI)-based localisation of the wireless endoscopy capsule. RSSI-based methods are widely used in indoor and outdoor positioning systems, allowing the estimation of the radial distance between reference sensors and an unknown transmitter position. Due to the significantly different electromagnetic (EM) properties of the human body as a propagation medium compared to air, one of the main objectives of this work was to develop an appropriate path loss propagation model. The advantage of the proposed solution is that it is based on an analytical approach and includes the attenuation constant defined by the EM properties of the soft tissues in the abdominal area. The theoretical basis of the developed attenuation path loss model (APLM) can be used to generalise and implement it for various In-to-On-Body communication systems at different operational frequencies. The APLM was numerically validated using CST Studio Software©, as well as by experiments on ex-vivo porcine tissues and in-vivo measurements on anesthetised living pigs. The experiments also served as the iii performance validation of a receiving inward cavity-backed slot antenna designed specifically for In-to-On-Body communications at 2.45 GHz, ISM band. In-vivo trials included implanting a wireless transmitter at several abdominal positions, which were then used for the 2-D and 3-D localisation accuracy assessment. Two trials were conducted separately at the medical facilities of The University of Southern Denmark (Odense, Denmark) and at the Herston Medical Research Centre, The University of Queensland (Brisbane, QLD, Australia). The research presents new knowledge for WBAN channel modelling for propagation media similar to the human body; confirms the reliability of the slot antenna performance for surface field measurements in WCE applications at 2.45 GHz; validates the application of the attenuation path loss model for In-to-On-Body communication channels; and it demonstrates high localisation accuracy when the proposed attenuation path loss model is used as an inverse solution for finding radial distances between the reference sensor and the unknown implant position.