TY  - JOUR
AU  - Ng, SX
AU  - Yee, MS
AU  - Hanzó, Lajos
TI  - Coded modulation assisted radial basis function aided turbo equalization for dispersive Rayleigh-fading channels
JF  - IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
J2  - IEEE T WIREL COMMUN
VL  - 3
PY  - 2004
IS  - 6
SP  - 2198
EP  - 2206
PG  - 9
SN  - 1536-1276
DO  - 10.1109/TWC.2004.837410
UR  - https://m2.mtmt.hu/api/publication/2928592
ID  - 2928592
N1  - Cited By :16            
            Export Date: 21 July 2022            
            Correspondence Address: Ng, S.X.; Department of Electrical, , Hampshire SO17 1BJ, United Kingdom            
            Funding details: European Commission, EC            
            Funding details: Engineering and Physical Sciences Research Council, EPSRC            
            Funding text 1: Manuscript received November 4, 2002; revised September 28, 2003; accepted December 1, 2003. The editor coordinating the review of this paper and approving it for publication is C. Xiao. This work is supported by the European Union under the auspices of the SCOUT project; by the Engineering and Physical Sciences Research Council, Swindon, U.K.; and by the Virtual Centre of Excellence (VCE) in Mobile Communications.
AB  - In this contribution a range of coded modulation (CM)-assisted radial basis function (RBF)-based turbo equalization (TEQ) schemes are investigated when communicating over dispersive Rayleigh-fading channels. Specifically, 16 quadrature amplitude modulation-based trellis coded modulation (TCM), turbo TCM (TTCM), bit-interleaved coded modulation (BICM), and iteratively decoded BICM (BICM-ID) are evaluated in the context of an RBF-based TEQ scheme and a reduced-complexity RBF based in-phase/quadrature-phase (I/Q) TEQ scheme. The least mean square (LMS) algorithm was employed for channel estimation, where the initial estimation step-size used was 0.05, which was reduced to 0.01 for the second and the subsequent TEQ iterations. The achievable coding gain of the various CM schemes was significantly increased, when employing the proposed RBF-TEQ or RBF-I/Q-TEQ rather than the conventional noniterative decision feedback equalizer (DFE). Explicitly, the reduced-complexity RBF-I/Q-TEQ-CM achieved a similar performance to the full-complexity RBF-TEQ-CM, while attaining a significant complexity reduction. The best overall performer was the RBF-I/Q-TEQ-TTCM scheme, requiring only 1.88 dB higher signal-to-noise ratio at BER = 10(-5), than the identical throughput 3 b/symbol uncoded 8 PSK scheme communicating over an additive white Gaussian noise channel. The coding gain of the scheme was 16.78 dB.
LA  - English
DB  - MTMT
ER  - 

TY  - JOUR
AU  - Chen, S
AU  - Mulgrew, B
AU  - Hanzó, Lajos
TI  - Least bit error rate adaptive nonlinear equalisers for binary signalling
JF  - IEE PROCEEDINGS-COMMUNICATIONS
J2  - IEE P-COMMUN
VL  - 150
PY  - 2003
IS  - 1
SP  - 29
EP  - 36
PG  - 8
SN  - 1350-2425
DO  - 10.1049/ip-com:20030284
UR  - https://m2.mtmt.hu/api/publication/2928618
ID  - 2928618
N1  - Department of Electronics, University of Southampton, Highfield, Southampton SO17 1BJ, United Kingdom            
            Department of Electronics, University of Edinburgh, King's Buildings, Edinburgh EH9 3JL, United Kingdom            
            Cited By :15            
            Export Date: 21 July 2022            
            CODEN: IPCOE            
            Correspondence Address: Chen, S.; Department of Electronics, , Highfield, Southampton SO17 1BJ, United Kingdom
AB  - The paper considers the problem of constructing adaptive minimum bit error rate (MBER) neural network equalisers for binary signalling. Motivated from a kernel density estimation of the bit error rate (BER) as a smooth function of training data, a stochastic gradient algorithm called the least bit error rate (LBER) is developed for adaptive nonlinear equalisers. This LBER algorithm is applied to adaptive training of a radial basis function (RBF) equaliser in a channel intersymbol interference (ISI) plus co-channel interference setting. A simulation study shows that the proposed algorithm has good convergence speed, and a small-size RBF equaliser trained by the LBER can closely approximate the performance of the optimal Bayesian equaliser. The results also demonstrate that the standard adaptive algorithm, the least mean square (LMS), performs poorly for neural network equalisers because the minimum mean square error (MMSE) is clearly suboptimal in the equalisation setting.
LA  - English
DB  - MTMT
ER  -