Diseño de un mecanismo de generación de actualizaciones dinámico utilizando aprendizaje reforzado en comunicaciones IOT celular para optimizar la frescura de la información

The present research work aimed to design “A DYNAMIC EFFICIENT MECHANISM IN ORDER TO REPORT DATA UPDATES GENERATION IN CELLULAR IOT COMMUNICATIONS” using reinforcement learning to optimize the age of performance metric information. A scientific methodology was applied, the same has allowed us to apply techniques such as Bellman's algorithm to fulfill with the control of the age of information and thus optimize the IoT network communication and guarantee on-time delivery information. The development was carried out in MATLAB mathematical software. Three scenarios have been studied to evaluate the maximum traffic load supported by the network; in each of them, the mechanism was designed, so that, the control proportional system directly to the AoI by keeping the reliability level, in terms of successful access probability ratio, greater than 97%. The results were tested using Neural Fitting for predicted value and the alternative hypothesis were accepted through statistical processes control. It is recommended continuing testing and modeling data networks using machine learning technology for greater accuracy future predictions.

• 1. Verfasser: Casanova Rivera, Hugo Fernando (author)
• Format: bachelorThesis
• Sprache: spa
• Veröffentlicht: 2022
• Schlagworte: ACTUALIZACIONES; APRENDIZAJE REFORZARDO
• Online Zugang: http://dspace.unach.edu.ec/handle/51000/9942