AI/ML-Based Sensing-Assisted Edge Computing in Next-Generation Mobile Networks

Integrated sensing and communications (ISAC) has received increased attention in light of the high-frequency bands to be employed by next-generation mobile networks; such waveform technologies natively support high-speed communications and high-resolution sensing, the latter of which, thus far, has been reserved exclusively for radar sensing platforms. To advance this integration, we propose two sensing parameters and corresponding analytics to be incorporated into two sequential optimization problems to minimize the latency in an end-to-end edge computing network. It is shown that without vital sensing functionalities, the network consistently executes poor offloading decisions. However, when equipped with crucial sensing-analytics, the best latency performance is guaranteed.