Rajeev Krishnamoorthy

A frequency offset estimation technique for frequency-selective fading channels

Rapid frequency offset estimation is required in burst communication systems. The performance of traditional open loop frequency estimation techniques is severely degraded by multipath propagation. We present a new data-aided technique for frequency offset estimation in the presence of strong multipath. The technique allows estimation of the frequency offset with no knowledge of the channel parameters. Results comparing the new technique with the maximum-likelihood estimation technique in a frequency selective Ricean fading environment are presented.

Efficient implementation of parallel decision feedback decoders for broadband applications

Parallel Decision-Feedback Decoders (PDFD) use a joint equalization and channel decoding scheme that performs decision-feedback equalization based on each survivor path in the channel decoder. In this paper, we propose novel efficient computation methods to implement the PDFD. We show that, contrary to the belief that PDFD is too complex to implement, standard Viterbi decoders can be modified to implement the PDFD and achieve very high bit-rates.

A robust Viterbi algorithm for symbol recovery in the 1900 MHz PCS band

Cellular TDMA-based handsets for the PCS 1900 MHz band will soon become very popular. With the cellular band at 900 MHz being congested by too many users the new band offers the service to twice as many customers. However, the requirements on the symbol recovery become much harder since a Doppler speed twice as high as in the cellular band is expected. Further impairment is given by the hardware realization that introduces timing errors and frequency offsets. This paper describes a Viterbi based equalizer-scheme that performs well under these conditions and satisfies the standards for cellular phones.