Browse Prior Art Database

METHOD FOR IMPROVED TIMING AND FREQUENCY OFFSET ESTIMATION IN A DIGITAL RADIO RECEIVER

IP.com Disclosure Number: IPCOM000008200D
Original Publication Date: 1997-Jun-01
Included in the Prior Art Database: 2002-May-27
Document File: 4 page(s) / 224K

Publishing Venue

Motorola

Related People

David Gurney: AUTHOR [+2]

Abstract

In a wireless digital communication system, a digital radio receiver may derive several demodu- lation parameters directly from the received signal. For example a symbol timing estimate is computed to determine the optimal received symbol sampling or decimation point and a frequency offset estimate is computed to enable the demodulator to compen- sate for receiver oscillator inaccuracies and Doppler shifts. In a typical digital receiver for a system such as the Personal Access Communication System (PACS), the received signal is oversampled and an entire slot of samples is stored until timing and frequency estimates have been calculated for that slot. This "block processing" approach enables the receiver to acquire and demodulate slots immediately, as is required for signaling bursts in many commu- nication systems.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 36% of the total text.

Page 1 of 4

m M-LA Technical Developments

METHOD FOR IMPROVED TIMING AND FREQUENCY OFFSET ESTIMATION IN A DIGITAL RADIO RECEIVER

by David Gurney and Kevin Baum

BACKGROUND

  In a wireless digital communication system, a digital radio receiver may derive several demodu- lation parameters directly from the received signal. For example a symbol timing estimate is computed to determine the optimal received symbol sampling or decimation point and a frequency offset estimate is computed to enable the demodulator to compen- sate for receiver oscillator inaccuracies and Doppler shifts. In a typical digital receiver for a system such as the Personal Access Communication System (PACS), the received signal is oversampled and an entire slot of samples is stored until timing and frequency estimates have been calculated for that slot. This "block processing" approach enables the receiver to acquire and demodulate slots immediately, as is required for signaling bursts in many commu- nication systems.

  The accuracy of symbol timing and frequency offset estimates is a critical factor in the perfor- mance of a digital TDM receiver. Prior block processing estimation methods rely on received data from only the desired timeslot to derive timing and frequency offset information. The present estima- tion method significantly improves symbol timing and frequency offset estimation in a digital TDM receiver, thereby increasing overall radio perfor- mance. The method enables the receiver to operate at a lower bit error rate or under worse channel conditions than is possible with the prior block estimation methods.

SOLUTION

  The described method achieves improved estimation of both timing and frequency by using information from prior slots to aid in the block timing and frequency offset estimation calculation. It retains the primary benefit of block estimation algorithms (namely fast acquisition) while simulta-

neously improving estimate quality. It is ideally suited for a TDMlTDMA receiver employing switched antenna selection diversity (such as in the PACS system) because it can use information from the antenna diversity selection slots to improve the quality of the timing and frequency offset estimates. The method may also utilize information from previous frames to further improve performance. The method could be applied to other digital radio systems which transmit on a continuous or semi- continuous basis, where prior channel information is available.

  Typically, both symbol timing and frequency offset are estimated from the data in the desired received slot. In hardware, once the optimal sampling point and the frequency offset are determined, the received data is decimated down to the symbol rate at the optimal symbol timing point. As mentioned above, the received data is stored in a buffer while this calculation is taking place since subsequent processing relies on the knowledge of the optimal symbol timing for decimation and frequency offset for coherent detection. Th...