A simplified EM algorithm for detection of CPM signals in a fading multipath channel

Authors:
Linda M. Zeger;Hisashi Kobayashi
Affiliations:
Lucent Technologies, Whippany, NJ;Princeton University, Princeton, NJ
Venue:
Wireless Networks
Year:
2002

Citing 3
Cited 2

A Sequential Algorithm for Joint Parameter Estimation and Multiuser Detection in DS/CDMA Systems with Multipath Propagation

Wireless Personal Communications: An International Journal
Is blind channel estimation feasible in mobile communication systems? A study based on GSM

IEEE Journal on Selected Areas in Communications
Turbo equalization for GMSK signaling over multipath channels based on the Gibbs sampler

IEEE Journal on Selected Areas in Communications

EM-based channel estimation algorithms for OFDM

EURASIP Journal on Applied Signal Processing
EM-based noise plus interference estimation for OFDM-based cognitive radio systems

Wireless Communications & Mobile Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Application of the EM (Expectation-Maximization) algorithm to sequence estimation in an unknown channel can in principle produce MLSE (maximum likelihood sequence estimates) that are not dependent on a particular channel estimate. The Expectation step of this algorithm cannot be directly performed for continuous phase modulated (CPM) signals transmitted in a time varying multipath channel. We therefore derive a simplification of the EM algorithm for CPM signals in this channel. Simulations applied to the Global System for Mobile Communications (GSM) show that the simplified EM algorithm significantly decreases the amount of training data needed for the channel model considered, and removes the majority of the bit errors that are due to imperfect knowledge of the channel.