project: ieee p802.15 working group for wireless personal area networks (wpans)
DESCRIPTION
Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [ Impact of MB-OFDM and DS-UWB Interference on C Band Receivers ] Date Submitted: [] Source: [ Torbjorn Larsson ] Company [ Paradiddle Communications ] - PowerPoint PPT PresentationTRANSCRIPT
Nov 2004
Torbjorn LarssonSlide 1
doc.: IEEE 802.15-04/0609r2
Submission
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: [Impact of MB-OFDM and DS-UWB Interference on C Band Receivers]Date Submitted: []Source: [Torbjorn Larsson] Company [Paradiddle Communications]Address [13141 Via Canyon Drive, San Diego, CA 92129, USA]Voice:[+1 858 538-3434], FAX: [+1 858 538-2284], E-Mail:[[email protected]]
Re: [Analysis of the impact of MB-OFDM and DS-UWB interference on a DTV receiver made in earlier contributions, in particular 802.15-04/547r0 and 802.15-04/0412r0]
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.
Nov 2004
Torbjorn LarssonSlide 2
doc.: IEEE 802.15-04/0609r2
Submission
Impact of MB-OFDM and DS-UWB Inteference on C-Band Receivers
Torbjorn Larsson
Paradiddle Communications, Inc.
Nov 2004
Torbjorn LarssonSlide 3
doc.: IEEE 802.15-04/0609r2
Submission
Motivation and Objective• Motivated by two contributions:
1) 04/0412r0, In-band Interference Properties of MB-OFDM, by C. Razell, Philips
2) 04/547r0, Responses to “In-Band Interference Properties of MB-OFDM”, by C. Corral, G. Rasor, S. Emami, Freescale Semiconductor
• The emphasis in the above contributions is on qualitative analysis
• In contrast, the approach here is “brute force” simulation
• Our hope is that the assumptions made are universal enough to be accaptable to the entire 802.15.3a task group
• The author is an independent consultant, not affiliated with any UWB company. This work was not carried out under any consulting contract
Nov 2004
Torbjorn LarssonSlide 4
doc.: IEEE 802.15-04/0609r2
Submission
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
Nov 2004
Torbjorn LarssonSlide 5
doc.: IEEE 802.15-04/0609r2
Submission
DVB-S
RSEncoder
Inter-leaver
204 bytes
188 bytes
Rate-1/2Conv
EncoderPuncturing
QPSKModulation
RRCPulse
Shaping
RadioTX
Code rates: 1/2, 2/3, 3/4, 5/6, 7/8
RSDecoder
De-Interleaver
De-Puncturing
Soft-InputViterbi
Matched Filter
RadioRX
188 bytes
204 bytes
Required BER = 2·10-4
Nov 2004
Torbjorn LarssonSlide 6
doc.: IEEE 802.15-04/0609r2
Submission
Typical C-Band Downlink Channelization
Horizontal Polarization Vertical Polarization
Channel Center Frequency (GHZ) Channel Center Frequency (GHz)
1A 3.720 1B 3.740
2A 3.760 2B 3.780
3A 3.800 3B 3.820
4A 3.840 4B 3.860
5A 3.880 5B 3.900
6A 3.920 6B 3.940
7A 3.960 7B 3.980
8A 4.000 8B 4.020
9A 4.040 9B 4.060
10A 4.080 10B 4.100
11A 4.120 11B 4.140
12A 4.160 12B 4.180
(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
Nov 2004
Torbjorn LarssonSlide 7
doc.: IEEE 802.15-04/0609r2
Submission
DTV Simulation Model• Excludes Reed-Solomon coding and interleaving
– Impossible to simulate error rates with RS coding– Will probably favor DS-UWB
• Symbol rate: 27 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
• Intend to run simulations for all code rates – Results presented only include rate 1/2 and 2/3
Nov 2004
Torbjorn LarssonSlide 8
doc.: IEEE 802.15-04/0609r2
Submission
MB-OFDM Transmitter Model
• Based on the Sep. 2004 release of the MB-OFDM PHY Specifications (P802.15-04/0493r1)
• 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 (=> worst-case in this context!)
• Channel number 9 (Band group 1, TFC 1)
• Data rate “110” Mbps (106.7 Mbps)
Nov 2004
Torbjorn LarssonSlide 9
doc.: IEEE 802.15-04/0609r2
Submission
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 (=> 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
Nov 2004
Torbjorn LarssonSlide 10
doc.: IEEE 802.15-04/0609r2
Submission
Interference Spectra
• 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)
2 2.5 3 3.5 4 4.5 5 5.5 6-90
-80
-70
-60
-50
-40
-30
GHz
dB
m
Power over 1 MHz Bandwidth
MB-OFDMDS-UWBFCC Mask
Resolution: 10 kHz
PSD averaged over 10 packets (roughly 0.9 ms)
Nov 2004
Torbjorn LarssonSlide 11
doc.: IEEE 802.15-04/0609r2
Submission
Interference Spectra – Close Up
• Both spectra exhibit substantial variations• Solution: run simulation for multiple DTV center frequencies
3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5
-45
-44
-43
-42
-41
-40
GHz
dB
m
Power over 1 MHz Bandwidth
MB-OFDMDS-UWBFCC Mask
DTV center frequencies
Nov 2004
Torbjorn LarssonSlide 12
doc.: IEEE 802.15-04/0609r2
Submission
Simulated DTV Center Frequencies
• Rate 1/2 simulations: 3.8 – 4.3 GHz in steps of 10 MHz– Arbitrary choice across 500 MHz bandwidth
• Rate 2/3 simulations: 3.72 – 4.18 GHz in steps of 20 MHz– According to channelization plan on slide 6
Nov 2004
Torbjorn LarssonSlide 13
doc.: IEEE 802.15-04/0609r2
Submission
Simulation Block Diagram
• Attenuation 1 is set so that the received DTV power is 3 dB above sensitivity
• Each simulation is performed with multiple DTC center frequencies
• Simulation results are plotted as a function of center frequency and attenuation 2
• No multipath!
DTVTransmitter
3 dB above sensitivity
Atten 2
PDTV
MB-OFDMTransmitter
DS-UWBTransmitter
Atten 1 + DTVReceiver
BERCounter
RandomData
CenterFrequency
CenterFrequency
PMB-OFDM
PDS-UWB
Re-Sample
fs = 21.6 GHz
204 byte packets
Nov 2004
Torbjorn LarssonSlide 14
doc.: IEEE 802.15-04/0609r2
Submission
-95 -94 -93 -92 -91 -90 -89 -8810
-6
10-5
10-4
10-3
10-2
10-1
100
BER After Viterbi Decoder (NF = 4 dB)
PDTV
[dBm]
BE
R
Code Rate 1/2Code Rate 2/3
BER Performance without Interference
• Sensitivity for rate 1/2 is -92.5 dBm (Eb/No = 3.2 dB)
• Sensitivity for rate 2/3 is -90.7 dBm (Eb/No = 3.7 dB)
Noise Figure = 4 dB
Defines sensitivity
Nov 2004
Torbjorn LarssonSlide 15
doc.: IEEE 802.15-04/0609r2
Submission
BER versus Center Frequency (Code Rate 1/2)
Interference attenuation = 67 dB
3.8 3.85 3.9 3.95 4 4.05 4.1 4.15 4.2 4.25 4.3
10-4
10-3
10-2
10-1
BER After Viterbi Decoder versus Center Frequency
Center Frequency [GHz]
BE
RMB-OFDMDS-UWB
Center frequencies separated by 10MHz
Nov 2004
Torbjorn LarssonSlide 16
doc.: IEEE 802.15-04/0609r2
Submission
Average BER (Code Rate 1/2)
65 65.5 66 66.5 67 67.5 68 68.5 69 69.5 70
10-4
10-3
10-2
10-1
Average BER After Viterbi Decoder versus Interference Attenuation
Interference Attenuation [dB]
BE
R
MB-OFDMDS-UWB
Nov 2004
Torbjorn LarssonSlide 17
doc.: IEEE 802.15-04/0609r2
Submission
Worst-Case BER (Code Rate 1/2)
65 65.5 66 66.5 67 67.5 68 68.5 69 69.5 70
10-4
10-3
10-2
10-1
Maximum BER After Viterbi Decoder versus Interference Attenuation
Interference Attenuation [dB]
BE
R
MB-OFDMDS-UWB
Nov 2004
Torbjorn LarssonSlide 18
doc.: IEEE 802.15-04/0609r2
Submission
BER versus Center Frequency (Code Rate 2/3)
3.7 3.75 3.8 3.85 3.9 3.95 4 4.05 4.1 4.15 4.210
-4
10-3
10-2
10-1
BER After Viterbi Decoder versus Center Frequency
Center Frequency [GHz]
BE
RMB-OFDMDS-UWBInterference
attenuation = 67 dB
Center frequencies separated by 20MHz
Nov 2004
Torbjorn LarssonSlide 19
doc.: IEEE 802.15-04/0609r2
Submission
Average BER (Code Rate 2/3)
65 65.5 66 66.5 67 67.5 68 68.5 69 69.5 70
10-4
10-3
10-2
10-1
Average BER After Viterbi Decoder versus Interference Attenuation
Interference Attenuation [dB]
BE
RMB-OFDMDS-UWB
Nov 2004
Torbjorn LarssonSlide 20
doc.: IEEE 802.15-04/0609r2
Submission
Worst-Case BER (Code Rate 2/3)
65 65.5 66 66.5 67 67.5 68 68.5 69 69.5 70
10-4
10-3
10-2
10-1
Maximum BER After Viterbi Decoder versus Interference Attenuation
Interference Attenuation [dB]
BE
R
MB-OFDMDS-UWB
Nov 2004
Torbjorn LarssonSlide 21
doc.: IEEE 802.15-04/0609r2
Submission
Conclusions
• For the two simulated cases (rate 1/2 and 2/3), the difference in average BER across the C-band is 1 dB or less
• The difference in worst-case BER is less than 0.5 dB
• More general conclusions should be postponed until all code rates have been simulated
Nov 2004
Torbjorn LarssonSlide 22
doc.: IEEE 802.15-04/0609r2
Submission
Onward…
• Run simulations for code rates 3/4, 5/6, 7/8• Run simulations for TFC 3 or 4• Include multipath• Suggestions?