1 / 58

Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs)

Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [ CM MATLAB Release 0.85 Support Document ] Date Submitted: [ 10 Nov 2006 ] Source: [ Rick Roberts ] Company [ Intel, Corp ] Address [Intel, Oregon]

carla-wong
Download Presentation

Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [CM MATLAB Release 0.85 Support Document] Date Submitted: [10 Nov 2006] Source: [Rick Roberts] Company [Intel, Corp] Address [Intel, Oregon] Voice:[503-929-5624], FAX: [Add FAX number], E-Mail:[richard.d.roberts@intel.com] Re: [] Abstract: [This document supports release 0.85 of the Matlab CM code.] Purpose: [Discussion and clarification of lingering questions.] Notice: This document has been prepared to assist the IEEE P802.15. It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor acknowledges and accepts that this contribution becomes the property of IEEE and may be made publicly available by P802.15. Rick Roberts, Intel

  2. This document “documents” the version 0.85 release of the MATLAB CM code. It does not reflect any changes made to the channel model parameters after about 5 November 2006. 06/0459r1 was modified by Su-Khiong Yong with additional information Rick Roberts, Intel

  3. Note 1: CM10 can be derived from the CM9. Rick Roberts, Intel

  4. Overloaded Channel Models Question: Is this approach acceptable to the committee? In addition to the original 8 channel environments, the overloaded CMs now give us 15 channel models. At 100 realizations per environment (typical) this gives the potential for 1,500 channel impulse responses. Just FYI. Rick Roberts, Intel

  5. Pertinent Definitions source: 15-06-0483-00-003c Rick Roberts, Intel

  6. Fig 1: Graphical representation of the CIR as a function of TOA and AOA. Source: 15-06-0195-03-003c Rick Roberts, Intel

  7. Small Scale Parameterization Rick Roberts, Intel

  8. Small Scale Parameterization (2) Rick Roberts, Intel

  9. Small Scale Parameterization (3) Rick Roberts, Intel

  10. Circular Polarized (1) Rick Roberts, Intel

  11. Circular Polarized (2) Rick Roberts, Intel

  12. } These first 3 parameters are stored in the data base but not used in the simulation. Is shadowing part of the link budget or should it be included in the simulation? Parameters in #12 are currently not used in the MATLAB code. All clusters are generated as relative energy and not absolute energy. d, Tx-Rx separation, h1, Tx height, h2 Rx height, GT, Tx gain, GR, Rx gain, k, ray Rician factor, , average power of the first ray of the first cluster (for combined two path and S-V model) Source: 15-06-0195-04-003c Rick Roberts, Intel

  13. Channel Model Parameters Blue = Provided Red = Assumed (missing value) Ref. 15-06-0400-01 & 15-06-0375-01 for CP Rick Roberts, Intel

  14. Rick Roberts, Intel

  15. Good agreement on Cluster Statistics between theory and actual. Rick Roberts, Intel

  16. Good agreement on Cluster Statistics between theory and actual (cont.). Rick Roberts, Intel

  17. Calling all mathematicians … are these correct? Getting these two functions correct is fundamental! Log Normal Poisson Rick Roberts, Intel

  18. Determining the number of clusters and the number of rays per cluster Rick Roberts, Intel

  19. Cluster Generation Ray Generation Definition of Variables Rick Roberts, Intel

  20. Putting it All Together – Composite Cluster/Ray Generation Rick Roberts, Intel

  21. Cluster Definition Rick Roberts, Intel

  22. Rick Roberts, Intel

  23. Ray Definition Rick Roberts, Intel

  24. Rick Roberts, Intel

  25. Combined Cluster + Ray Definition Rick Roberts, Intel

  26. Rick Roberts, Intel

  27. 3-D Representation Rick Roberts, Intel

  28. Rick Roberts, Intel

  29. Apply the Spatial Filtering to form IR Rick Roberts, Intel

  30. Rick Roberts, Intel

  31. Creating Continuous Time Impulse Response Rick Roberts, Intel

  32. Convert Continuous Time to Discrete Time Rick Roberts, Intel

  33. One Impulse Response out of 100 Realizations Note: there is a mistake in release 0.85 of the code that gave the real part of the impulse response for the imaginary part. Rick Roberts, Intel

  34. Generate Impulse Response Metrics by CM Rick Roberts, Intel

  35. Rick Roberts, Intel

  36. Rick Roberts, Intel

  37. Example of Extreme Spatial Filtering Rick Roberts, Intel

  38. An example of Spatial Filtering out the Signal (CM10) Model Characteristics Mean delays: excess (tau_m) = 1.24 ns, RMS (tau_rms) = 0.15 ns # paths: NP_10dB = 1.1, NP_85% = 1.0 Channel energy: mean = -0.0 dB, std deviation = 0.0 dB Channels Spatially Nulled: 24.0 , Remaining Channels: 76.0 Rick Roberts, Intel

  39. Cluster Structure Definitions Rick Roberts, Intel

  40. Four Different Types of Cluster Structure • Regular LOS (CM5, 7, 8, 9, 10) • TSV LOS (CM1, 3, 9) • NLOS (CM2, CM4) • Synthesized NLOS (6) Rick Roberts, Intel

  41. Regular LOS Clusters First cluster contains both LOS impulse and multipath energy Rick Roberts, Intel

  42. TSV LOS Clusters First cluster is just the LOS impulse Rick Roberts, Intel

  43. Regular NLOS Clusters All clusters are present with no consistent dominate path Rick Roberts, Intel

  44. Synthesized NLOS Clusters First cluster (LOS) is nulled out Rick Roberts, Intel

  45. Impulse Response Truncation Rick Roberts, Intel

  46. Discrete Time Impulse Response Truncation Routine – prevents excessively long impulse responses containing little energy % truncate impulse response to the -40 dB point z_max=max(max(abs(ImpDt))); for index_cn=1:NumChannels IM_done=0; for index=length(ImpDt):-1:1 % work backwards thru vector if IM_done==0 if abs(ImpDt(index,index_cn))>z_max/1e2 index_max(index_cn)=index; % search for largest index that gives -40 dB IM_done=1; end end end end ImpDtTrunc=ImpDt(1:max(index_max),:); % truncate by using the largest index Rick Roberts, Intel

  47. Questions to be Resolved Rick Roberts, Intel

  48. Questions for the Committee • In document 15-06-0400-01, there are currently 5 parametric representations of CM1, 2 representations of CM3, and 3 representations of CM9.  For this code release, these multiple representations are denoted as CM1.1, CM1.2, CM1.3, CM1.4, CM1.5, CM3.1, CM3.2, CM9.1, CM9.2, CM9.3.  The committee needs to tell me how to deal with multiple parametric representations of the same channel model environment. Should I leave it as I have it now? • Bring to 06/483r0 under “unresolved issues” • 2. I’ve derived the missing NLOS models from existing LOS models.  Specifically I‘ve generated CM2 from CM1 and CM6 from CM5. My question is: “When generating CM2, which of the 5 representations of CM1 should I use”? • Bring to 06/483r0 under “unresolved issues” • 3. Each ray is generated with a corresponding AoA. This AoA can be used for either spatial filtering, antenna pointing, or both. I need guidance as to what the committee wants to do. An option is just to leave it the way I currently have it. • Bring to technical requirement document • 4. Currently 3 things needed to be input before an impulse response can be generated. These are the antenna pattern, the pointing direction and the sample rate. Are these going to be specified by the committee or by the proposer. • Bring to technical requirement document Rick Roberts, Intel

  49. Questions for the Committee (continued) 5.There are a number of S-V parameters that are currently missing. What can be done about these? Bring to 06/483r0 under “unresolved issues” 6. Traditionally, standards committees generate a “golden set” of impulse responses for all proposers to use. We have the option of generating a “golden set” of clusters for all proposers to use. The clusters would be loaded and the spatial filtering and sample rate would be entered by the user. Does the committee want to do it this way? Generate one set without spatial filtering. And possibly additional sets with spatial filtering 7. Currently I do not having shadowing in the code. TG3a typically applied several dB of random shadowing. Does the committee want to include this? Should it be part of the link budget? Is it really necessary? The group does not consider shadowing effect in the Matlab code 8. Currently all LOS impulse responses start at T=0. TG3a actually had a random dither on the LOS starting time. The code is already set up to accommodate a LOS time dither but it is not currently being implied. What is the opinion of the committee? Does this matter? Should I leave it as is? The group does not need to reflect this. 9. Does the committee concur with truncating the impulse response to the -40 dBr point? -30 dB shall be used. Rick Roberts, Intel

  50. Appendix Menu Options Rick Roberts, Intel

More Related