Michigan Technological University
Department of Mathematical Sciences

Applied Math Seminar Announcement


Speaker:

Seyed A. Zekavat1 and Fatemeh Emdad2
1Department of Electrical and Computer Engineering, MTU
2 Department of Mathematics, Colorado State University

Title:

Smart antenna beam forming via maximum noise fraction for multi-carrier CDMA scheme

Date:

Tuesday, March 11, 2003

Time:

1:05 pm - 1:55 pm

Location:

Fisher Hall 325

Abstract:

There is a growing market for voice, data and video via wireless communication systems. The goal in this market is to increase the network capacity (numbers of users and bit rates), increase quality of service (probability-of-error performance) and reduce cost. Smart antenna technology as well as Multi-carrier Code Division Multiple Access (MC-CDMA) are key methods which help meet the capacity, quality, and cost requirements of future generation wireless communication systems.

Smart antenna arrays are smart, since they can detect the direction of the desired user and direct the antenna pattern toward that user. This technique creates Spatial Division Multiple Access (SDMA), and, as a result, enhances the capacity of wireless system via directionality. One key problem in smart antenna technology is to shape the beam pattern of antenna arrays, null the interfering users and receive only the desired user signal. This is possible via a perfect antenna array beam forming.

MC-CDMA is emerging as a powerful multiple access protocol. In MC-CDMA, each user's information symbol is transmitted over N carriers simultaneously. To ensure separablity of users' data streams at the receiver, each user assigns a unique spreading sequence to the N carriers, typically a set of values each either +1 or 1. MC-CDMA enhances the probability-of-error performance in wireless communication systems via generating frequency diversity at the receiver.

In this seminar, we introduce maximum noise fraction (MNF) transformation to perfectly generate smart antenna beam forming. The technique MNF is based on an optimization problem which maximizes the signal to noise content. This optimization problem leads to a Generalized Singular Value Decomposition (GSVD), a powerful tool in numerical linear algebra. Considering the received MC-CDMA signals (from all users, including the desired user) at the input of the antenna array elements as a summation of the desired data (signal) plus non-desired data (interference and noise), the goal of MNF transformation is to maximize the signal content assuming the interference and noise covariance matrix is well estimated.


Back to . . .
 
Applied Math Seminars

Center for Applied Mathematics

Department of Mathematical Sciences

Michigan Technological University