Research On Wireless Resource Allocation And Coverage Control For Cellular Mobile Communication Network Optimization

Network optimization is an essential work to maintain cellular mobile communication networks operating with good performance all the time, as it adjusts network parameters to match the current demand and operation environment effectively.In this thesis, radio resource allocation and coverage control technologies are studied deeply for cellular mobile communication network optimization.Joint allocation of frequency and scrambling code is investigated for TD-SCDMA (Time-division Synchronous Code-division Multiple Access) systems. TD-SCDMA has very limited numbers of frequencies and scrambling codes, and the cross-correlation between the spreading codes are very poor. The purpose of joint allocation is to take full use of the freedom degrees of these two resources. Interference to signal ratio (ISR) is adopted as the measure to evaluate allocation scheme for its interference suppression effect. A complete model of inter-cell ISR is established by taking into account the radio propagation condition, traffic distribution, frequency orthogonality and code cross-correlation. The optimization problem of joint frequency and scrambling code allocation is formed to minimize network-wide ISR. And then a two-step allocation algorithm JFSA (Joint Frequency and Scrambling-code Allocation) is proposed to solve the problem efficiently, that is, joint frequency and scrambling code family allocation first, and scrambling code selection then. Such a two-step processing reduces computational complexity dramatically while maintains the performance excellently. Numerical results show that the proposed algorithm can effectively decrease interference severity locally and network-wide in varios scenarios, as compared with the frequency-scrambling code sequential allocation algorithm.With the rapid development of urbanization and transportation, wireless traffic is extended to distribute in the three-dimensional space.Therefore, a new demand of three-dimensional coverage of common control signals (CCSs) is raised. In such a background, three-dimensional coverage control technologies of CCSs are studied, based on planar array antenna and three-dimensional beamforming. In order to reduce the complexity of coverage control, we propose an idea of a sector-by-sector three-dimensional coverage control technology. Namely, the sectors are decoupled through interference control on the sector boundaries, so that the coverage control of a single sector can be performed individually, and then the coverage control of the whole network can be realized by going through all the sectors one by one. For a single sector, two three-dimensional coverage control algorithms are proposed, either for whole-area receive power constraint, or for boundary receive power constraint. In the former, an optimization problem is established, in which transmission power is minimized while zero weak coverage ratio (WCR) and zero interference level is guaranteed. Semi-definite relaxation, randomization and power adjustment are employed to solve the problem and thus obtain a sub-optimal excitation weights effectively. In the latter three-dimensional coverage control method, a desired array pattern reflecting coverage requirement is subtracted from the target coverage distances and signal propagation losses in all directions. To achieve excellent performance, an appropriate array pattern synthesis method is investigated to minimize of array pattern error and transmission power synchronously. Numerical results validate that the two proposed algorithms outperform the JEPD (Joint exhaustion of power and downtilt) algorithm in coverage hole elimination, inter-cell interference suppression and energy saving obviously.Moreover, a cluster-by-cluster three dimensional coverage control of CCSs is studied. Based on the analysis on the interference relationship between sectors, we point out that the whole network can be divided into a series of sector clusters which are mutually interference independent, and the three-dimensional coverage control of the entire network can be realized by exhausting every cluster. For a specific sector cluster, an optimization model to improve receiving signal qualities is built with the objective of approximately average SIR (Signal to interference ratio) maximization and the constraint of zero WCR. An efficient algorithm JCCSC (Joint Coverage control for Sector Cluster) is proposed to obtain the excitation weights through SDR solving, excitation weights recovery and power adjustment. Numerical results shows that JCCSC algorithm can significantly improve the SIR level on the whole target coverage area under the WCR constraint.