Propagation Faults in Real-Time Content Streaming Across Low-Bandwidth Learning Infrastructure

Abstract

Delivering live educational videos in places with thin internet pipes is still tough because the signal breaks often, showing up as lost frames, endless rebuffering, or delays that pile on top of one another. In response, this paper builds a detailed simulation tool that tests how these failures unfold across different hardware layouts and network grades, paying special attention to bandwidth, mixed devices, and cache-based buffers. The experiments find that storing key content at the edge cuts the average stalls by more than 40 per cent versus relying on a distant central server, a gap that widens under the 512-kbps cap. A three-dimensional model of delay spread further shows that both pause rates and picture quality drop rise sharply, not linearly, as bandwidth jitter and the number of viewers climb. The work also pinpoints fault link patterns tied to specific protocols and suggests tuning buffer sizes along with smarter retry timings to dampen these cascades. Taken together, the results give clear design tips for rolling out remote learning where infrastructure is weak.