Research On Wireless Resource Optimization Strategy Based On Chaotic Neural Network

With the rapid development of broadband wireless technology and the wide spread of the multimedia mobile business, all kinds of broadband mobile multimedia businesses become fashionable. However, the limited wireless resource cannot meet the requirements of quality of service (QoS) of these multimedia services, which has become one of the major bottlenecks for the development of broadband wireless communication technology. In order to break through this bottleneck, the techniques about wireless resource allocation and optimization are constantly emerging and try to improve the utilization of wireless resource. These algorithms have achieved some successes, but, it’s very hard to obtain the optimal solution due to the extreme complexity of this problem, there is still a certain gap between the obtained solution by these algorithms and the optimal solution. This paper focuses on further optimizing the wireless resource while allocating the wireless resource using chaotic neural network (CNN). The improved chaotic neural network is applied into three typical broadband wireless communication systems, the goal of further optimizating wireless resource is achieved by successfully searching the optimal solution using rich chaotic neurodynamics. The contributions made in this paper include:(1) An improved chaotic neural network scheme is proposed. Since the Travelling Salesman Problem (TSP) was solved successfully by Hopfield neural network (HNN), the HNN series has been intensively studied and applied in various kinds of optimization problems. However, its exact parameters choice is exquisitely sensitive and complicated due to the complexity of chaotic theory. In order to improve the system performance, the characteristics of its parameters are investigated in detail. Through a large quantity of analyses and numerical simulations, we present the improved scheme with reasonable parameter settings method and an effective parameter set which gives1) less steep sigmoid function,2) stronger synaptic weights, and3) higher initial temperature for annealing. Compared with traditional algorithms, the improved scheme has much faster convergence speed with a small amount of accuracy loss, which is more suitable to be used for practical application scenarios. The improved algorithm is successfully applied to three typical wireless resource optimization problems in the latter part of this paper.(2) An adaptive resource optimization algorithm for multiuser OFDMA system is proposed based on the transiently chaotic neural network (TCNN). The algorithm investigates the problem of wireless resource optimization in multi-user OFDMA system and explores the margin adaptive (MA) optimization criterion again. The objective of resource optimization is to minimize the total transmission power while guaranteeing the subcarrier requirement of every user. Subcarriers are allocated to every user adaptively according to their channel state information (CSI), the suitable modulation scheme is selected to determine the number of bits of each OFDM symbol on each subcarrier. Finally, the transmission power of each subcarrier is calculated. The simulation results show that TCNN can find the optimal solution using the chaotic neurodynamics generated transiently. The problem of wireless resource optimization in multi-user OFDMA system is solved effectively.(3) A joint wireless resource optimization algorithm for unicast and multicast services is proposed based on noisy chaotic neural network (NCNN). To support the multicast and unicast services in OFDMA system simultaneously, a channel-aware adaptive resource allocation algorithm is proposed to maximize the total throughput of the unicast service while guaranteeing the required QoS for the multicast service. The two-step optimisation scheme is developed to solve the problem:firstly, subcarriers are allocated to the multicast and the unicast services under the assumption that power is divided equally to every subcarrier. Especially, an improved NCNN is applied to allocate the subcarriers to the unicast service. Secondly, the power averagely allocated to the unicast service is reallocated quickly in a linear water-filling fashion. The optimal solution is found successfully through its rich neurodynamics (including flexible chaos and stochastic noise) so as to exploit the multiuser diversity gain fully. The proposed algorithm achieves higher spectrum efficiency and better bit error rate (BER) for the multicast service, also higher throughput for the unicast service.(4) A channel allocation optimization algorithm is proposed for Multimedia Broadcast/Multicast Service Single Frequency Networks (MBSFN) based on NCNN. Three types of electromagnetic compatibility constraints are discussed, and these interference constraints are redefined as four different interference constraints according to the topology of MBSFN areas. The four interferences are avoided effectively by elaborately constructing the energy function, the total channel number is minimized by the gradient descent searching algorithm. Simulation results show that the proposed algorithm can avoid successfully the four electromagnetic interferences in MBSFN areas, and can improve the spectrum utilization ratio by frequency reuse technology, and can minimize total channel number. The channel allocation optimization algorithm for MBSFN is solved effectively.In conclusion, this paper investigates the problem of wireless resource optimization technology based on chatic neural network. In order to further improve wireless resource efficiency, this paper carries out appropriate physical abstraction, mathematical modeling, and typical case simulations using chatic neural network. Through theoretical analyses and simulations, the proposed algorithms can make full use of rich neurodynamics to search the optimal solution, and achieve the purpose of further optimization wireless resources. Meanwhile, the research presented in this paper sets a good example for developing multi-disciplinary association theory.