A hybrid packet scheduling algorithm with QoS provision in HSDPA system

Abstract

Data services are expected to exhibit a significant increase in data rate and system capacity over the next years. The demand for high speed data communications at anytime and anywhere leads to ever evolving area of mobile communications system. Many researchers have delivered a lot of quality technologies to support high rate data service in the area of mobile communications. As a result, the world is now witnessing wide spread wireless communications technology. $\It{High Speed Downlink Packet Access (HSDPA)}$ system is a piece of evidence of the growth. It was first launched in Korea in 2006 and can achieve downlink data rates up to 14.4 Mbps. In addition, scheduling is one of the most important issues in radio resource management (RRM) functions. We investigate scheduling algorithms proposed for HSDPA; also, we propose a simple hybrid downlink packet scheduling algorithm. Generally, a HSDPA packet scheduler should be able to adapt instantaneous channel variations, e.g., sudden degradation or deep fading. Recently, many scheduling algorithms have been proposed for system throughput maximization. User``s traffic types and channel conditions which are two major ingredients in existing scheduling algorithms. Which are called a traffic class-based (i.e., QoS-based) and channel-dependent scheduling algorithm respectively. Traffic class-based scheduling algorithm is designed to guarantee QoS for end users. And, channel-dependent schedulers are designed to adapt instantaneous channel variations. Furthermore, both can be combined together resulting in two-layered scheduling algorithm. However, no work has taken consideration into hybrid automatic repeat request (HARQ) process in which retransmissions cause the waste of resources. Our proposed packet scheduler makes decisions based on HARQ operations as well as channel status and traffic classes. In short, a QoS concept is an important factor in HSDPA system since there exists a fundamental tradeoff between...