IRZ0140 Signals and Signal Processing

Transcription

1 IRZ040 Sgnals and Sgnal Processng Lecture 4 Jula Berdnkova Contact data Jula Berdnkova e-mal: julad@lr.ttu.ee Home page of the course: 2 Sgnal processng applcatons Sgnal processng applcatons Communcaton systems Data transmsson Speech, Audo, Televson, Broadcastng, Multmeda Communcaton system Data acquston system (data measurng, sensng systems) Navgaton Space applcaton Data collectng systems rtr tr Dgtal/ Analog user Sgnal processng Transmtter Recever Automatc Control Bomedcal systems Robotcs Mltary applcatons jtc. 3 uqn Xong. Dgtal modulaton technques. Boston (Mass.) ; London : Artech House, 2006 Sawhney, G.S, undamentals of Bomedcal Engneerng 4

2 Data acquston systems Passve (wthout transmttng) Actve (sgnal generaton) Man sgnal processng steps for actve systems: sgnal generaton recepton Recepton types for communcaton and sensng system Coherent Noncoherent Coherent detecton the phase of the sgnal to be known exactly. processng (detecton, estmaton), measured data representaton, data storage data post-processng 5 6 Dgtal modulatons Dgtal modulatons N noncoherently D dfferentally 7 8

3 Dgtal modulatons Dgtal modulatons ASK (Ampltude-Shft Keyng ) ( 2 f t), t T s ( t) = A p( t)cos π 0 =,2, KM c SK (requency-shft Keyng) s ( 2 f t+ Φ ), 2, 3, KM = Acos π = PSK ( Phase-Shft Keyng) ( 2πf t+ ), t T s ( t) = Acos θ 0 =,2, K,M c ( 2 ) π θ= M ASK SK PSK 9 0 Crtera of Choosng Modulaton Scemes for Communcaton Systems Sgnal processng algorthms Sgnal processng algorthm depends on the applcaton area and type power effcency (bt error rate or bt error probablty) bandwdth effcency, system complexty. Wdely used: spectral analyss flterng statstcal processng 2

4 Bomedcal Sgnals (Example) Heart electrcal conducton at lmb surfaces Electrocardogram (ECG) Bomedcal Sgnals (Example) Man medcal magng sgnals: Surface CNS electrcal actvty Electroencephalogram (EEG) a) x-ray transmsson b) Gamma-ray transmsson Magnetc felds of neural actvty Magnetoencephalogram (MEG) c) Nuclear magnetc resonance nducton d) Ultrasound echoes Muscle electrcal actvty Electromyogram (EMG) Thes, aban J. Meyer-Bäse, Anke, Bomedcal Sgnal Analyss : Contemporary Methods and Applcatons Amne Naīt-Al, Advanced bosgnal processng 3 4 Actve sensng systems The range to a target, target velocty and target propertes could be measured usng a transmtted sgnal. Bomedcal Sgnals (Example) Ultrasound Mnkoff J., Sgnal Processng undamentals and Applcatons for Communcatons and Sensng Systems 5 6

5 Dgtal sgnal generaton Perodc sgnal generaton (example) s(0) Sne wave generaton sne_table Table generaton s() s(2). s(n-) N N 0x00 0x8 0x or 0xE7 N ormula generaton (recursve algorthms, step value) s(t) s(t) t d t 7 Table addresses are from 0 up to N- Next address could be calculated by ADDR= ADDR-+C s = A sne_table[addr] d out = C d / N 8 If samplng frequency s the output sgnal frequency wll be Perodc sgnal generaton (example) Sawtooth sgnal generaton Perodc sgnal generaton (example) Sne wave generaton, recursve algorthm d - samplng frequency, = d out = T - output sgnal frequency d - samplng frequency, t d=, d Output sgnal frequency should be out, d N- number of level between mn. and max. of the sgnal < T 2 td out = N = N T d 9 out 6-bts xed pont N=65536 = out Sgnal phase value: ϕ =2π out, ϕ - =2π out (-), ϕ =2π out =2π out / d. Sne sgnal: Out_S =A sn(ϕ )= A sn(ϕ - + ϕ )=A sn(ϕ - ) cos(ϕ )+A cos(ϕ - ) sn(ϕ ) Cosne sgnal: Out_C =A cos(ϕ )= A cos(ϕ - + ϕ )=A cos(ϕ - ) cos(ϕ )-A sn(ϕ - ) sn(ϕ ) C_s= sn(ϕ )=sn(2π out /d), C_c= cos(ϕ )=cos(2π out / d ) Recursve output for sne Out_S= Out_S - C c +Out_C - C s and cosne sgnals Out_C= Out_C - C c -Out_S - C s 20

6 MATLAB Demos NI LabVIEW 2