Offloading Methodologies for Air-Ground Edge Intelligent Computing Systems

Nowadays, a new domain for next Edge Intelligent computing systems, i.e., systems that combine edge computing with artificial intelligence (AI) capabilities, is that based on a functional integration of Unmanned Aerial Vehicles (UAVs) as flying computing nodes with terrestrial networks. In particular, this chapter presents a UAV-Aided Mobile Edge Intelligent Computing system, in which heterogeneous traffic flows with different quality of service constraints, have to be offloaded on processing nodes consisting of terrestrial and flying edge computing nodes. According to this, we illustrate here the use of a matching algorithm to perform proper offloading strategies. The matching algorithm provides decisions on the basis of per-flow end-to-end delay bounds formulated by resorting to the combined application of stochastic network calculus and martingale envelopes theory. Furthermore, this chapter offers a theoretical discussion of the matching stability and provides some numerical results to highlight the validity of the stochastic framework proposed in fitting the actual network behavior, considering also different state-of-the-art offloading alternatives.