Title :
Soft detection with synchronization and channel estimation from hard quantized inputs in impulsive UWB Power Line Communications
Author :
Tonello, Andrea M. ; Palermo, Nicolangelo
Author_Institution :
DIEGM, Univ. di Udine, Udine, Italy
Abstract :
We consider an ultra wide band power line communication system that deploys binary impulse modulation. We propose a simple matched filter receiver that uses one-bit quantized input samples for frame synchronization and soft channel estimation with data training. Data detection can still use unquantized (soft) inputs. The proposed algorithm has low complexity and it requires less memory and computational power than the unquantized maximum likelihood synchronization algorithm. It produces small excess noise w.r.t. the unquantized algorithm for low signal-to-noise ratios. Simulation results over highly frequency selective PLC channels confirm that also the bit-error- rate performance penalty is small.
Keywords :
carrier transmission on power lines; channel estimation; error statistics; matched filters; maximum likelihood estimation; synchronisation; ultra wideband communication; binary impulse modulation; bit error rate performance; channel estimation; data detection; data training; frame synchronization; hard quantized inputs; impulsive UWB power line communications; matched filter receiver; maximum likelihood synchronization algorithm; one-bit quantized input samples; signal-to-noise ratio; soft detection; ultrawideband system; Channel estimation; Computational modeling; Frequency synchronization; Matched filters; Maximum likelihood detection; Maximum likelihood estimation; Power line communications; Programmable control; Signal to noise ratio; Ultra wideband communication; Channel estimation; power line communications; synchronization; ultra wide band;
Conference_Titel :
Ultra-Wideband, 2009. ICUWB 2009. IEEE International Conference on
Conference_Location :
Vancouver, BC
Print_ISBN :
978-1-4244-2930-1
Electronic_ISBN :
978-1-4244-2931-8
DOI :
10.1109/ICUWB.2009.5288765