Multi-Rate Block Transmission Over Wideband Multi-Scale Multi-Lag Channels

• Published in: IEEE transactions on signal processing Vol. 61; no. 4; pp. 964 - 979
• Main Authors: Tao Xu, Zijian Tang, Leus, G., Mitra, U.
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01.02.2013; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Applied sciences; Bandwidth; Baseband; Blocking; Channels; Data models; Detection, estimation, filtering, equalization, prediction; Discretization; Equalization; Exact sciences and technology; Information, signal and communications theory; Mathematical models; multi-rate transmission; Multilayers; OFDM; Passband; Receivers; Signal and communications theory; Signal, noise; Studies; Telecommunications and information theory; underwater communications; Wideband; wideband time-varying channels; Performance evaluation; Time variable channel; Discrete channel; Channel structure; underwater communications; Information rate; Transmitter; Wide band transmission; Base band; Multirate system; Equalizer; Equalization; Signalling; wideband time-varying channels; Information transmission; Discretization; Carrier frequency; Wide band; Diagonal matrix; Multiscale method; Signal processing; Submarine telecommunication; multi-rate transmission
• ISSN: 1053-587X; 1941-0476
• DOI: 10.1109/TSP.2012.2230169

Many linear time-varying (LTV) channels of interest can be well-modeled by a multi-scale multi-lag (MSML) assumption, especially when wideband transmissions with a large bandwidth/carrier-frequency ratio are employed. In this paper, communications over wideband MSML channels are considered, where conventional communication methods often fail. A new parameterized data model is proposed, where the continuous MSML channel is approximated by discrete channel coefficients. It is argued that this parameterized data model is always subject to discretization errors in the baseband. However, by designing the transmit/receive pulse intelligently and imposing a multi-branch structure on the receiver, one can eliminate the impact of the discretization errors on equalization. In addition, to enhance the bandwidth efficiency, a novel multi-layer transmit signaling is proposed, which is characterized by multiple data rates on different layers. The inter-layer interference, induced by the multi-layer transmitter, can also be minimized by the same design of the transmit/receive pulse. As a result, the effective channel experienced by the receiver can be described by a block diagonal matrix, with each diagonal block being strictly banded. Such a channel matrix structure enables the use of several existing low-complexity equalizers viable.