Author: Min Sung Kim

Publisher:

Published: 2013

Total Pages:

ISBN-13:

Cellular wireless data traffic is now growing at about 80% compound annual growth rate (CAGR), and hence creating a corresponding demand for additional capacity. This traffic growth in the developed world comes mainly from the increasing use of video file transfers or viewings. In the emerging countries, the increasing smart-phone penetration is an additional factor. The demand for capacity can be met by more spectrum, higher spectrum efficiency, and a greater number of cells. However, increasing spectrum and spectrum efficiency has proved to be slow because of both regulatory and technological barriers. Therefore, increasing the number of cells is an effective way to add capacity. Because of the diverse types of mobile data, the use of different tiers of cells has become necessary: high-power and high-capacity large macro cells to serve vehicle based users, medium-power and medium-capacity small pico cells for outdoor pedestrian users, and very low-power and low-capacity femto cells for indoor users. These cells often work together and are known as a heterogeneous network. The cells in a heterogeneous network often share the limited spectrum available to the operator. They reuse the spectrum across tiers as well as within the same tier. This inter-tier reuse creates interference, e.g., femto to macro interference. In addition, intra-tier interference arises from frequency reuse within a tier, e.g., femto to femto interference. Managing interference in heterogeneous networks is hence complex compared to the interference management of earlier single-tier macro cell networks, which is well understood. The research presented in this dissertation studied two interference management problems: inter-tier interference management between a macro cell and a femto cell, and intra-tier interference among femto cells. Chapter 2 presents the downlink femto-macro inter-tier interference management problem. This chapter proposes a cost criteria of maximizing the minimum (worst-case) rate of femto users while constraining the worst-case performance degradation of macro cell users caused by interference from the femto base station. Analysis shows that a combination of frequency (resource-block) allocation for femto users and power control of each resource block can lead to good results. Because the general joint resource block-allocation and power-control problem is very complex (NP-hard), a suboptimal method is developed using a decoupled resource allocation and power control. All these solutions were also verified by simulations. An analytical solution for the single femto user case also appears. Chapter 3 presents the management of downlink intra-tier interference in an enterprise femto network with multiple femto base stations. This chapter discusses admission of users to receive service, allocation of users to femto base stations, and the downlink power control to maximize the service rate while maintaining the QoS criteria, which specifies the worst-case 95th percentile download time of a file. The analysis demonstrates the effectiveness of the proposed algorithm by simulation and also discusses naive (baseline) scheduling, which does not manage intra-tier interference, that leads to poor throughput as the density of femto cells increase. Last, this chapter explores a flexible user-base station association approach wherein the base station associated with a given user can be dynamically changed during a download period at network events. Such flexibility adds a significant performance improvement. Chapter 4 concludes the thesis by summarizing the results, discussing the limitations of the work presented, and making suggestions for future research.