Adaptive PSP-MLSE receiver for GSM/EDGE system in fast multipath fading channels

Zhenhong Li, Aarne Mämmelä

    Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

    5 Citations (Scopus)

    Abstract

    An adaptive MLSE receiver based on per-survivor processing (PSP-MLSE) for the GSM/EDGE system with 8PSK modulation is presented. The initial channel impulse response is estimated by the LS algorithm with a training midamble sequence. The PSP-MLSE is in conjunction with the LMS algorithm to track the fast time-varying channel. Based on the EDGE burst structure, the second part of the data stream is tracked forwards while the first part is tracked backwards in order to improve the performance in rapidly changing channels. Computer simulations indicate that the PSP-MLSE is clearly superior to the conventional adaptive MLSE (CA-MLSE), especially, at high velocities. Moreover, the loss due to the estimation error is also presented by measuring the ideal system performance. Furthermore, the termination effect of the Viterbi algorithm is also discussed.
    Original languageEnglish
    Title of host publication2000 IEEE International Conference on Personal Wireless Communications
    PublisherIEEE Institute of Electrical and Electronic Engineers
    Pages184-188
    ISBN (Print)0-7803-5893-7
    DOIs
    Publication statusPublished - 6 Aug 2002
    MoE publication typeA4 Article in a conference publication
    EventIEEE International Conference on Personal Wireless Communications -
    Duration: 17 Dec 200020 Dec 2000

    Conference

    ConferenceIEEE International Conference on Personal Wireless Communications
    Period17/12/0020/12/00

    Keywords

    • GSM
    • Fading
    • Computer simulation
    • Least squares approximation
    • Maxium likehood estimation
    • Delay estimation
    • Phase shift keying
    • Viterbi algorithm

    Fingerprint

    Dive into the research topics of 'Adaptive PSP-MLSE receiver for GSM/EDGE system in fast multipath fading channels'. Together they form a unique fingerprint.

    Cite this