Intercellular coordination schemes for improving spectrum utilization in wireless cellular networks

Junyi Li, Purdue University

Abstract

Wireless networking is fast becoming one of the dominant areas in communication systems. The problem of spectrum scarcity is a major bottleneck for wireless network systems. Therefore, spectrum reuse, such as in cellular systems, is essential for efficient spectrum utilization. We categorize various reuse schemes into two intrinsically different methods: one based on worst-case signal-to-interference-ratio (SIR) and the other based on capture. In this thesis, we propose several intercellular coordination schemes to improve spectrum utilization for these two spectrum reuse methods. We first present a novel channel carrying scheme to address the problem of call handoffs for the worst-case SIR based spectrum reuse. Our basic idea is to allow mobile users to carry their current channels into new cells under certain conditions. In order to avoid co-channel interference due to channel movement, we develop the (r + 1)-channel assignment scheme for linear cellular networks. This affords us channel mobility at the expense of some capacity. We then build on the channel carrying idea to develop a meta-cell based channel sharing scheme, in which channels are assigned to meta-cells instead of cells (as in traditional channel assignment schemes). The fundamental advantages include: (1) channels can be shared within meta-cells without coordination with other cells; and (2) handoff processing within meta-cells is simplified. Finally, we propose a novel static coordinated power control for capture based spectrum reuse. Our basic idea is to reduce intercellular interference and improve the capture probability by coordinating transmission powers of users in different cells. This coordination is determined beforehand and no real-time intercellular coordination is required. Power control is static and fixed, for simplicity of implementation. We formulate and solve a generic optimal scheduling problem with our coordination scheme. We find that the optimal scheduling policy is in a simple form of bang-bang control. We evaluate, via numerical analysis and simulation, both throughput and delay of the coordination scheme, and compared them with other schemes. We find that the coordination scheme can achieve significant performance improvement over a wide range of capture ratio values.

Degree

Ph.D.

Advisors

Chong, Purdue University.

Subject Area

Electrical engineering|Systems design

COinS