Markov Approximation Based Resource Allocation Algorithm for Multi - Frequency Cooperative Networks

Multi-frequency cooperative network (MCN) which integrated millimeter wave (mmWave) and sub-6GHz is envisioned to provide high data rate service whilst guaranteeing reliable coverage. While mmWave signals suffer severe path loss and sudden blockage in the non-line-of-sight (NLOS) environment, sub-6GHz signals can provide reliable connectivity and ensure basic traffic demands. In this paper, a resource allocation problem over both frequency bands is discussed and the users with severe large-scale fading mmWave channels are supposed to be assigned more sub-6GHz bands. A combinatorial nonconvex optimization expression is adopted to maximize the sum data rates of all users over both bands under a series of constraints. With the help of penalty function, the difficult problem is transferred to an equivalent problem whose near optimal solution can be obtained preliminary by log-sum-exp method. Then a Markov chain is employed to iterate over all feasible solutions and converge to the near optimal solution. The numerical results are shown to verify the performance of the proposed algorithm.