Network Utility Maximization in Wireless Networks over Fading Channels with Uncertain Distribution

In this the network utility maximization problem is investigated for wireless networks when only channel mean and variance are known. Due to the randomness of wireless channel, link outage will happen and a traffic flow's rate will gradually decrease along its routing path. When only channel mean and variance are known, channel distribution is uncertain and it is harder to control the rate loss caused by link outage. We take into account this feature of rate loss and target at maximizing the network utility of multiple traffic flows' rates at their destinations. An optimization problem is formulated, which is non-convex. Global optimal solution is achieved with the aid of bilevel optimization method and monotonic optimization method. Numerical results are presented to verify the effectiveness of our proposed method.