A GPU accelerated framework for monitoring LTE/5G interference to DVB-T systems

This work presents a new computation system to evaluate the interference maps between the radio mobile signal and the DVB-T signal based on a multiprocessor computing architecture, such as GPU systems (Graphics Processing Units) designed for high-performance parallel computing. The goal is to overcome the time complexity of current approaches based on sequential implementations on multicores CPUs that require several hours to generate and update the interference maps of the Italian territory. The new system’s design considers new parallel computing methodologies to consider rapid changes in the GPU hardware and uses a high-performance distributed SQL engine on RAPIDs to create an efficient and scalable framework for processing a massive volume of simulation data radio. Carrying out test on the radio-electric data of the Italian provinces with the most significant territorial extension shows compression of at least a factor of 30 in the execution times, which for the entire Italian territory fall below 20 minutes.