No abstract

The evolution of communication networks has made network system increasingly enormous and complex in recent years. Inter-cell interference (ICI) in the denser network will degrade the throughput and affect the Signal-to-Interference-plus-Noise Ratio in the Long Term Evolution (LTE) network. Assigning the Physical Cell ID (PCI) to cells in a proper way is effective to reduce ICI in the downlink network. In this work, the aim is to consider PCI collision, confusion and Reference Signal Collision (CRS) between neighbor cells comprehensively. For the purpose, a local search algorithm is proposed and its time complexity is O(Km 3 n 3 ) in the worst case. Compared with the random plan, numerical results of simulation experiment demonstrate the effectiveness and advantage of local search.

TD-SCDMA is the third-generation mobile communication standard of our country that adopts time slot duplex and code division multiple access (CDMA). Time slots planning, i.e. the division of uplink and downlink time slots, is critical in TD-SCDMA. Most existing algorithms and strategies are greedy algorithm, which can't ensure the optimality or approximate optimality of the result. In this paper, some optimization models on the planning of the time slots for single channel problem and multiple channel problem are proposed. The simulation result show that our method can get a good result.

In this paper, we propose a fine-grained grid-based multi-objective model which aims at optimizing base station antennas' configurations, such as transmit power, antenna tilt and antenna azimuth, in order to upgrading network performance in cellular networks. As the model is non-convex, non-smooth and discrete and computationally expensive, we use decomposition method to solve the MOP problem. We mainly focus on addressing the scalarized sub-problem after decomposition. For the scalarized sub-problem, we propose an enhanced difference method. First, difference of each component is calculated, which provides the guidance of optimization. Then an OPSO is applied to search the optimal step length. The method is applied to GSM network optimization on an area in Beijing. The effect of the application shows that proposed method has a good performance, and is effective/efficient to solve mobile network optimization problems.

Particle swarm optimization (PSO) and differential evolution (DE) algorithm are biologically inspired computing, which simulate the behavior of survival of the fittest from the systems of nature biology, ecology and so on. As heuristic random algorithms, they have characteristics of self-adaptation, self-organization, selflearning, etc. PSO and DE can be used to solve various complex problems that are difficult by using traditional calculation methods. Based on the platform of ANSYS finite element soft, parametric programming of a single-phase cable and a three-phase cable is accomplished in the paper by using ANSYS parameter design language (APDL). Optimization calculation of the maximum electric field strength of cable insulation layer is realized by PSO and DE combined with finite element method (FEM). That biologically inspired computing combined with FEM provides reference and enlightenment for optimizing highvoltage equipment in aspects of electromagnetic field.