Impact of MBOFDM and DSUWB Interference on C Band Receivers

1
Project IEEE P802.15 Working Group for Wireless
Personal Area Networks (WPANs) Submission Title
Impact of MB-OFDM and DS-UWB Interference Part
2 Date Submitted Source Torbjorn
Larsson Company Paradiddle Communications Addre
ss 13141 Via Canyon Drive, San Diego, CA 92129,
USA Voice1 858 538-3434, FAX 1 858
538-2284, E-Mailtlarsson_at_san.rr.com Re
Analysis of the impact of MB-OFDM and DS-UWB
interference on a DTV receiver made in earlier
contributions, in particular 802.15-04/0412r0,
802.15-04/547r0 and 04/451/r2 Abstract The
impact of MB-OFDM and DS-UWB interference on a
C-band DTV receiver is investigated by
simulation Purpose To present an unbiased
comparison of the impact of MB-OFDM and DS-UWB
interference based on a minimal set of
universally accepted assumptions 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.
2
Impact of MB-OFDM and DS-UWB Inteference Part 2

• Torbjorn Larsson
• Paradiddle Communications, Inc.

3
Motivation and Objective

• Motivated by the following three contributions
• 04/0412r0, In-band Interference Properties of
  MB-OFDM, by C. Razell, Philips
• 04/547r0, Responses to In-Band Interference
  Properties of MB-OFDM, by C. Corral, G. Rasor
  and S. Emami, Freescale Semiconductor
• 04/451/r2, Multiband OFDM Interference on
  In-band QPSK Receivers Revisited, by C. Corral,
  S. Emami and G. Rasor, , Freescale Semiconductor
• The above contributions focused on the impact of
  MB-OFDM interference on a DTV victim receiver
• In contrast, the objective here is to quantify
  the difference in the impact of interference from
  MB-OFDM and DS-UWB

4
C-Band DTV Systems

• The C-band downlink spans 3.7 4.2 GHz
• C-band antennas are typically 6 12 feet in
  diameter
• Based on the DVB-S (Digital Video Broadcasting
  Satellite) standard (EN 300 421)
• DVB-S was designed for MPEG-2 broadcasting in the
  Ku-band, but is also used in the C-band
• DVB-S does not specify a unique set of data rates
  or symbol rates However
• Typical transponder bandwidth is 36 MHz (33 MHz
  also used)
• Typical symbol rate 27 29 Msps
• DVB-S2 is the next generation with improved
  bandwidth efficiency and FEC

5
DVB-S
6
Typical C-Band Downlink Channelization
(Telesat satellite Anik F2. Footprint North
America)

• Total of 24 channels
• Each polarization has 12 channels
• Transponder bandwidth is 36 MHz with a 4 MHz
  guard band
• The center frequencies are separated by 40 MHz
• The center frequencies for the two polarizations
  are offset by 20 MHz
• The result is 24 center frequencies separated by
  20 MHz

7
DTV Simulation Model

• Excludes Reed-Solomon coding and interleaving
• Impossible to simulate error rates with RS coding
• Symbol rate 28 Msps
• No quantization (including input to Viterbi
  decoder)
• Ideal pulse shaping/matched filters (0.35
  roll-off)
• No nonlinarity
• No frequency offset
• No phase noise
• Pre-computed phase error and time offset
• Receiver noise figure 4 dB
• Code rates 2/3 and 7/8

8
MB-OFDM Transmitter Model

• Based on the Jan. 2005 release of the MB-OFDM PHY
  spec
• Complete Matlab implementation of the
  specifications
• System operating in band-hopping mode
• Includes (5-bit) DAC and realistic filter
  characteristics
• Spectral pre-shaping to compensate for non-ideal
  filter characteristics (gt worst-case in this
  context!)
• Channel number 9 (Band group 1, TFC 1)
• Data rate 110 Mbps (106.7 Mbps)

9
DS-UWB Transmitter Model

• Based on the July 2004 release of the DS-UWB PHY
  specifications (P802.15-04/0137r3)
• Complete Matlab implementation of the
  specifications
• No DAC
• Ideal RRC pulse shaping filter truncated to 12
  chip periods (gt worst-case)
• Channel number 1 (chip rate 1313 Mcps)
• Data rate 110 Mbps (109.417 Mbps)
• BPSK modulation
• Spreading code for preamble and header (PAC) -1
  0 1 -1 -1 -1 1 1 0 1 1 1 1 -1 1 -1 1 1
  1 1 -1 -1 1
• Spreading code for frame body 1 0 0 0 0 0

10
Interference Spectra
Resolution 10 kHz PSD averaged over 10 packets
(roughly 0.9 ms)

• Transmit power is set so as to push each spectrum
  as close as possible to the FCC limit (worst-case
  condition)
• MB-OFDM transmit power is -10.3 dBm
• DS-UWB transmit power is -10.8 dBm (data rate
  dependent)

11
Interference Spectra Close Up
DTV center frequencies

• Both spectra exhibit substantial variations
• Solution run simulation for multiple DTV center
  frequencies

12
Amplitude Histogram Wideband (without Multipath)
Data rate 110 kbps
PAR 12.0 dB
PAR 11.9 dB
13
Amplitude Histogram Wideband (with Multipath)
PAR 14.1 dB
PAR 14.2 dB
Data rate 110 kbps
100 multipath channel realizations
14
Amplitude Histogram Output of DTV Matched Filter
(with Multipath)
Center frequency 4 GHz
100 multipath channel realizations
15
Output of Matched Filter (Close-Up)
16
Changes Since November 2004

• All simulations carried out with center
  frequencies according to channelization plan on
  slide 6
• 3.72 GHz to 4.18 GHz in steps of 20 MHz
• Added multipath (CM3, no shadowing)
• Increased symbol rate from 27 to 28 Msps

17
Simulation Block Diagram

• Attenuation 1 is set so that the received DTV
  power is 3 dB above sensitivity
• Each simulation is performed with all 24 DTV
  center frequencies
• Simulation results are plotted as a function of
  center frequency and attenuation 2

18
DTV Sensitivity (NF 4 dB)
Defines sensitivity
19
DTV Sensitivity (NF 4 dB)
(Symbol rate 28 Msps)
20
BER versus Center Frequency
Code Rate 1/2, without multipath
Interference attenuation 67 dB
21
BER versus Center Frequency
Code Rate 1/2, with multipath
Interference attenuation 71 dB
22
Two Performance Metrics

• Compute average error rate over the 24 center
  freqeuencies
• Select maximum error rate from the 24 center
  frequencies

23
Average BER
Code Rate 1/2, with multipath
24
Maximum BER
Code Rate 1/2, with multipath
25
BER versus Center Frequency
Code Rate 7/8, with multipath
Interference attenuation 71 dB
26
Average BER
Code Rate 7/8, with multipath
27
Worst-Case BER
Code Rate 7/8, with multipath
28
Conclusions

• For DTV code rate 1/2, there is practically no
  difference between the two interfering waveforms
• The difference increases with code rate
• MB-OFDM interference is more bursty in nature and
  thus has more impact for higher code rates
• For DTV code rate 7/8, MB-OFDM is 1.3 dB worse
  than DS-UWB _at_ BER 210-4