Development of a Theoretical Model for Physiological Signal Degradation in Wearable Sensor Networks Under Sub-Saharan African Deployment Conditions

Large-scale wearable sensor network deployments for population health surveillance and chronic disease monitoring in sub-Saharan African settings face signal degradation mechanisms not adequately modelled by existing theoretical frameworks, creating a gap between system design assumptions and real-world performance that prevents reliable health data collection. This dissertation develops an original theoretical model of physiological signal degradation in wearable sensor networks specifically accounting for the combined effects of environmental, physiological, and infrastructure factors characteristic of sub-Saharan African deployment contexts. The theoretical model is constructed from first principles for three physiological signal domains: photoplethysmographic signals, skin-surface electrochemical biosensor signals, and bioimpedance signals. For each domain, a degradation transfer function is formulated that maps environmental and physiological state variables, including ambient temperature, relative humidity, physical activity intensity, sweat rate, skin tone index, and sensor-skin contact pressure, to signal quality degradation components including noise floor elevation, baseline drift amplitude, artifact event frequency, and sensor sensitivity drift. Model parameters are identified from a structured empirical characterization study conducted across three sites in Nigeria, representing South West, North Central, and North East environmental conditions, with forty wearable sensor node deployments per site over six-month periods. Experimental data from one thousand and forty sensor-months of deployment were used for model identification and cross-site validation. The validated degradation model is then applied to develop deployment-condition-aware data quality flagging and adaptive signal processing specifications for a population health monitoring network architecture. Original theoretical contributions include the unified multi-domain degradation transfer function formulation and the empirically identified sub-Saharan African deployment condition parameter set. Keywords: physiological signal degradation, wearable sensor networks, theoretical model, sub-Saharan Africa, deployment conditions.