School Work

Simulacao de Canal DRM

Description
channel simulation using matlab, DRM, radio digital, Simulacao de Canal DRM, Simulacao de Canal DRM, Simulacao de Canal DRM, Simulacao de Canal DRM, Simulacao de Canal DRM,
Categories
Published
of 3
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  imulaio o igial Radio Modiale ael Model Kui Li GxSOC Research Institute Communication University of China Beijing, China liki _1497@yahoo.com.cn Abstract-Th paper analyzes the channel model of igital Radio Mondiale and discusses the parameters given by the specication. Watterson model is used to simulate the radio channel. We describe the implementation of Watterson channel model in details, including the channel multi-path and timing-variability. Multi-level interpolator is used in our realization scheme to improve calculation eciency. The simulation results are given and indicate that the timing and frequency selective fading effect of the DRM channel are well modeled. Keywords D WSSUS; Watterson Model; Multi path; Doppler Spread; Muliinterpolate . INOON Digita Radio Mondiae (DRM) is a word-wide specication for Amplitde odlation (AM) broadcasting. It is used to repace te traditiona anaogue AM roadcasting for its advantage such as robustness in fading channels, beter adio qality, power sings at the  transmiting side. The eqency of DRM is below 30MHz. t  this roadcasting equency band, te signals are  transmited by sky-wave or ground-wave or both of them. The characteristics of DRM channel is complex becase of te instability of the ionospere and the muti-hop propagation beteen the eart s surface and te ionospere. It is a typical multi-path and time-varying channel that presents deep time and eqency selective fading. The channel modeing and simulation is signicant for the systems design, impementation and performance evauation. The DRM specication gives the channel model [1]. In this paper,  we discss  the model realization in detail based on the given channe caracteristics and parameters. I. DRM CHANNEL M OEL DRM channel given by the specication is described as an adaptable parameters Wide Sense Stationary Uncorrelated Scatering (WSSUS) model.  is a stocastic  time-varying mode wit a stationary statistics and several sets of appropriate parameter vales. The parameters include the nmbers of multi-path, path delay, path gain, oppler shi, and Dopper spread. Tae-I shows one of the six sets parameters given by the specication. The nmber of discrete multi-path, each path delay and gain describes the equency seective fading of the channel. Te Doppler si and spread presents e channe s  time seective fading. In 98--8-8-/$00 ©0 IEEE 35 Van Ming GxSOC Research Institute Communication University of Cina Beijing, China  yanming668@yahoo.com.cn  the WSSUS model, each path of the channel is regarded as independent. TABLE  D HANNL PAATS Path t Path  Path  Path 4 el 0 2 ms 4 ms 6 ms ath Gain, nS 0.5 1 0.25 0.0625 Doppler Shi 0 1.2 Hz 2.4 Hz 3.6 Hz Doppler Spread 0.1 Hz 2.4 Hz 4.8 Hz 7.2 Hz The WSSUS channel model is expressed as: h T    L   T -Tn (1)  P k is the attenuation of the k-t pat, describing te  relative path gain of each path. r k is the relative delay of the k-th path. {e k (t)} are the time-variant tap  weights, and they are zeros mean complex-vaued stationary Gaussian random processes  with nit  variance. The magnitudes Ie k (t)1 are Rayeig-distriuted and te phases (t) are uniformy distribted. Each {e k (t)} is caracterized y its Power Density Spectrum (PDS) that determines the average speed of  variation in time. The wi of the PDS is quantied by Dopper spread wich is specied as 2-side PDS and contains 68% of the power. The PDS has a Gassian shape for the ionosphere path based on real observation. There might be a non-zeros center equency of te PDS, wich is an average equency si dened y Doppler shi. III. MPLEMENTAION OF CHAEL ODEL The WSSUS channel Model can be implemented by Watterson mode as sown in Fig. I. Waterson channe mode is a transversa lter were  taps g k (t) are complex and vary with time [2].  is an ideal discrete time delay model and the vaues of deay  refer to Tae-. The taps of each path is independent [3][4][5]. { g k(t)} are complex  valued stationary Gaussian random processes and teir PDS can be descried y:  P f)=   P k  exp[_� f-fDSh k)2]  (2)  k _k 5  _k   P k is te atenuation of te k-t pat, f  DS  P _k  the Dopper spread and h k is te Doppler shi. se l - r t) Figure 1. Watterson hannel Model. The generation of time-varying taps can be divided into  three independent parts:  the path gain, oppler shi simulation and oppler spread simlation, as shown in Fig. 2. r-------------- igure 2. Three Pars of Taps in Waterrson Model Te g k (t) can be expressed as: g k(t) = P k . c k(t)· }J _k (3) P k and Ck(t) are descried in Section I. The path gain directy multipies on each path. Multipy }_k  to eac pat  to reaize  the Dopper shi effect. The key of oppler spread eect simlation is the taps C k (t) generation. Te zero-means, unit varance stochastic processes {Ck (t)} ave Gaussian Probaiity Density Function (PDF) and Gaussian shapes PDS which are controled y  the Doppler spread, as described in Section I. We generate C k (t) trogh  the following three steps: 1  Generate a zero mean and unit variance compexvaued Gaussian wite noise. 2. Filter the Gassian  white noise to forming a color noise with the required PDS. 3. Ierpoate te tered sampes for matc  the samping  rate. This process illustrates in Fig. 3. 36 Figure 3. Generation of Time-varing Taps Te compex-vaued Gaussian white noise is produced by Box-Muler algoritm. In step-2, because ltering is a linear operation,  the color noise has the same PDF with the srcinal wite noise. Te lter is designed based on the PDS  requirements and its coefcients are dened y: r; 22 h(t) = e J { (4)  f  DS  P is the Doppler spread. The pass band width of the Gaussian lter is controlled by oppler spread We can calclate  the taps by a lower speed for efciency ecause te channe  timing seective fading is far lower than  the system sampling eqency, bt it mst be at least 32  times than the Doppler spread  values for accuracy [6]. At the same time, an ierpoator is used for match te sampling equency of taps and signa sampes. F or different values of oppler spread, the taps updating  rate maybe different to dozens of times. At this condition, use the same sets of lter coefcients and through adjustable mlti-level interpolator control the lter pass band can simpi  the simuation design. IV. IULATION RESULTS The simulation of multi-path and Doppler shi of the channel is straightforward. The key point of the model simuation is te  time-varying taps generation. The magnitude and phase of the compex-valued stationary Gaussian random processes sample nction C k (t) is shown in Fig and Fig. 5. Te magnitude is Rayeigdistribted and the phase is uniformly distribted as expected. igure 4. The magnitudes distribution of C k   Figure 5. The angles distribution of C k  The power density spectrum of ck (t) is shown in Fig. 6 and Fig. 7. The magnitude in time-domain is sown in Fig. 8 and Fig. 9. The vale of the Doppler spread decides the timeselective fading degree. Te simulation results show tey matc  well. Figure 6. The PDS of Ck , Doppler spread = 7 2 H  z Figure 7. The S of Ck  (t), oppler sp rea d   = AHz  V. CONLUSION In is paper, we discuss the DM channe model in detail. Te aterson model is used to simuation  the channel. The tapped delay line with time-varying taps simuates te equency and time seective fading. 37 Figure 8. Magnitude of C k  in time domain, Doppler spread = 7 2 H  z Figure 9. Magnitude of C k  in time domain, Doppler spread = 2AHz e generate the taps y ltering Gaussian wite noise to form color noise with required PDS which derives om the channel Doppler spread. In our simulation scheme, mltievel interpolate is used to adapt wide dynamic range of Doppler spread. Te simulation resuts show tat te scheme can simlate  the DRM channel accurately and efciently. REFEENES [I] ETSI ES 201980 V2.1.1 (2004-06), igital adio Mondiale (M) System Specification. [ 2] C C Watterson, J. R Juroshek, W. D. Bensema, Experimental onrmation of an H hannel odel , IEEE Trans. On omm. Tech, Vol. OM-18,No. 6,Dec. 1970. [ 3] Yip, K. W, Ng, T.S., An analytic discrete-time model for a fading dispersive WSSUS channel, IEEE Vehicular Technology onference, 1994 I EEE 44th, 8-10 June 1994 age(s): 180-184 vo!. 1. [ 4] Angling, M l, Davies, N. C An assessment of a new ionospheric channel model driven by measurements of multi-path and Doppler spread, IEEE Frequency Selecltion and Management Techniques for HF ommunications (Re No. 1999/017), EE olloquium on, 29-30 March 1999 Page(s): 41 4/6. [ 5] Behm, C 1 A narrowband high equency channel simulator with delay spread, IEEE HF Radio Systems and Techniques, Seventh Inteational onference on (on Pub . No. 441),7-10 July 1997 Page(s) 388-391. [ 6] W Furman, 1 Nieto, Harris orporation, Understanding HF channel simulator requirements in order to reduce H moderm performance measurement variabilit, Proceedings of HF I, the Nordic HF onference, August 2001.
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks