Throughput Optimization for Cognitive UAV Networks: A Three-Dimensional-Location-Aware Approach

An effective throughput optimization problem is studied in cognitive unmanned aerial vehicle networks (CUAVNs), where a leading unmanned aerial vehicle (LUAV) serving as an aerial mobile access point and sensing node receives data from a group of following unmanned aerial vehicles (FUAVs). For the case of single FUAV, the joint optimizing of three-dimensional location and spectrum sensing duration of LUAV is formulated to maximize FUAV’s throughput. The original non-convex problem is tackled by iteratively optimizing two convex sub-problems. For the case of multiple FUAVs, the preferable LUAV location of each FUAV can be individually calculated by using the algorithm developed for the single FUAV case, and then the final LUAV location is obtained by solving a well-formulated weighted sum problem. Numerical results show that our proposed single FUAV algorithm and throughput weighted sum based on multiple FUAVs algorithm can approach the optimal one with only 2% performance loss. Moreover, compared to the only time-dimension spectrum sensing scheme, the effective throughput of FUAV rises about 500% with the proposed schemes.