Study On Resource Allocation Algorithm In Ad Hoc Network Based On Network Utility Maximization

Ad Hoc network is widely used in many fields such as industry,military and disaster relief because of its characteristics of self-organization,no center and distributed.The host nodes in the Ad Hoc network are mainly battery-powered and have limited energy,so it is important to save the energy of the nodes through an effective power control strategy.At the same time,the wireless channel used in communication between Ad Hoc network nodes is also susceptible to factors such as signal attenuation and interference.In view of these problems,this paper deeply studies the resource allocation problem of Ad Hoc network under limited node energy and channel fading based on the Network Utility Maximization theory,and obtains some research results,Listed below are our main research contents:Firstly,we study the utility maximization problem of Ad Hoc network with link outage probability constraints under channel fading.Because there are multiple distributions of the channels of the wireless network,the evaluation requires a large number of samples,which is difficult.but we can get the information of the upper bound of link outage probability for any distribution subjected to the mean and variance through the simple channel mean and variance information.So,Based on the mean and variance information of the channel known in the Ad Hoc network,this paper considers the situation that the data stream rate is decreasing due to the interruption of the link during the data transmission and models the network utility maximization by using the utility of the effective rate as the objective utility function.The non-convex model is transformed into a convex optimization model by variable substitution,and the model is analyzed by using The Lagrangian duality theory.Meanwhile a distributed algorithm is proposed.The experimental results show that the effectiveness and convergence of the algorithm.Secondly,considering the impact of node transmit power on wireless link capacity,we introduce the power cost function based on the maximization of rate utility,and propose an optimization model of joint power control and rate utility in Ad Hoc network.In contrast to the traditional wireless network utility maximization based on the subgradient-based dual-decomposition algorithm,this paper uses the distributed Newton algorithm with the secondary convergence speed to solve the resource allocation problem of Ad Hoc network and realizes the network utility maximization.And Aiming at the problem that the distributed Newton algorithm based on matrix splitting technology has a slow convergence of dual variables,an improved scheme is proposed.The experimental results show that this model can effectively decrease the transmit power of nodes and achieve a good balance between network rate utility and energy consumption.Moreover,the improved distributed Newton algorithm has faster convergence speed.