Title

Vector Quantization Based Qos Classification For Admission Control In Cdma Systems

Keywords

Admission control; Clustering; Outage probability; Power control; QoS index; QoS space; SINR; Vector quantization

Abstract

Optimal power control is of great importance for CDMA systems and it can be controlled to provide the desired quality of service (QoS) to mobile hosts in a cellular radio system. The power levels of all the mobile hosts are determined and constantly tuned in order to achieve the required SINR (signal to interference and noise ratio) which changes dynamically. The SINR of all the K mobiles in a cell can be expressed in the form of a k-dimensional vector. It helps determine the operating point of the system and hence it is constantly monitored and updated due to the variability in the wireless channel conditions and user mobility. We view this continuously changing vector as the motion of a point in a higher dimensional Euclidean space, called the QoS space. We apply vector quantization technique to shrink the infinite-point space to a finite-point space by partitioning the former into N regions such that the points within a region reflect almost similar system performance and are identified by what we call a QoS index. We show how the system operating point can be mapped to one of the QoS indices. The location of the point or the region of operability in the QoS space conveys the system status in terms of the current load and the QoS being delivered. The dynamism in the system's input conditions due to wireless link characteristics and user mobility acts like an opposing force against which the system has to operate. The system reacts to all such changes preventing it from going into a region with an undesirable QoS index. We show how the apriori knowledge of the operating region helps in decision making pertaining to call admission and resource allocation in CDMA systems. © 2005 Springer Science + Business Media, Inc.

Publication Date

11-1-2005

Publication Title

Wireless Networks

Volume

11

Issue

6

Number of Pages

709-718

Document Type

Article

Personal Identifier

scopus

DOI Link

https://doi.org/10.1007/s11276-005-3525-6

Socpus ID

26644447189 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/26644447189

This document is currently not available here.

Share

COinS