36
ITC/USA 2001 October 22-25, 2001, Las Vegas, NV 1 www.nova-eng.com 513-642-3000 Performance Of SOQPSK And Multi-h CPM In The Presence Of Adjacent Channel Interference Terrance J. Hill Nova Engineering, Inc., Cincinnati, OH

SOQPSK and Multi-h CPM in the Presence of …...• Multi-h CPM nearly doubles the bps / Hz of SOQPSK -B, with the same detection efficiency. Title SOQPSK and Multi-h CPM in the Presence

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV1

    www.nova-eng.com513-642-3000

    Performance Of SOQPSK And Multi-h CPM In The Presence Of

    Adjacent Channel Interference

    Terrance J. HillNova Engineering, Inc.,

    Cincinnati, OH

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV2

    www.nova-eng.com513-642-3000

    Outline• Waveform definitions (briefly!)• Spectral characterizations of telemetry waveforms• Effects of IF filtering on link performance• Aggregate effect of adjacent channel interference

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV3

    www.nova-eng.com513-642-3000

    Why These Waveforms?• Advanced Range Telemetry (ARTM) Program

    – Before ARTM (Tier 0?)• PCM/FM• “Legacy” waveform for telemetry

    – ARTM Tier 1• Proprietary Feher-patented FQPSK-B, Revision A1• SOQPSK is

    – Comparable in performance to FQPSK-B, Revision A1– Non-proprietary

    – ARTM Tier 2 • Multi-h CPM (M=4, L=3RC, h1 = 4/16, h2 = 5/16)

    • SOQPSK and Multi-h CPM are continuous phase modulations

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV4

    www.nova-eng.com513-642-3000

    CPM Waveforms

    • Constant envelope, information is carried entirely in the phase• PSD and BER can be “traded” by

    – Varying h, modulation index– Changing g(t), the frequency pulse shape

    • FQPSK-B (Revision A1) is not truly a CPM waveform because of its AM component

    -∞< t

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV5

    www.nova-eng.com513-642-3000

    Offset QPSK Variations

    • Unshaped

    • Rectangular– MIL-STD-188-18x

    • “Shaped”– Arbitrary g(t)

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV6

    www.nova-eng.com513-642-3000

    SOQPSK• MIL-STD-188-181, -182 • Continuous phase modulation, h = 1/4, g(t) = rectangle• Constant envelope• Three rates of phase change: 0, +∆f, -∆f

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV7

    www.nova-eng.com513-642-3000

    Definition of SOQPSK-A, -B

    g(t) = n(t) * w(t), where

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV8

    www.nova-eng.com513-642-3000

    Frequency Pulse Shape, g(t)

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV9

    www.nova-eng.com513-642-3000

    Power Spectral Density

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV10

    www.nova-eng.com513-642-3000

    SOQPSK-A Eye Pattern

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV11

    www.nova-eng.com513-642-3000

    SOQPSK-B Eye Pattern

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV12

    www.nova-eng.com513-642-3000

    Definition of Multi-h CPM

    -∞< t

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV13

    www.nova-eng.com513-642-3000

    Key CPM Parameters

    • M – Order of Modulation (2-ary, 4-ary, etc.)• g(t) - Frequency Pulse (Rectangular, Raised Cosine, etc.)• L – Length of Frequency Pulse• h – Modulation Index

    • Results here are for M=4, L=3RC, h1 = 4/16, h2 = 5/16

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV14

    www.nova-eng.com513-642-3000

    Frequency Pulse and Phase Tree

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV15

    www.nova-eng.com513-642-3000

    PSD Comparison

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV16

    www.nova-eng.com513-642-3000

    PSD is Half the Story• Overall spectral efficiency is determined by spacing

    between channels• Receiver selectivity affects channel spacing• A valid comparison must account for both transmitted

    spectrum and “tolerable” receiver filtering• Not all modulations are equally “tolerant” of IF filtering

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV17

    www.nova-eng.com513-642-3000

    SOQPSK-AWide IF

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV18

    www.nova-eng.com513-642-3000

    SOQPSK-ANarrow IF

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV19

    www.nova-eng.com513-642-3000

    SOQPSK-BWide IF

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV20

    www.nova-eng.com513-642-3000

    SOQPSK-BNarrow IF

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV21

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV22

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV23

    www.nova-eng.com513-642-3000

    Experimental Setup

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV24

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV25

    www.nova-eng.com513-642-3000

    9 Mbps SOQPSK-A, Multichannel

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV26

    www.nova-eng.com513-642-3000

    9 Mbps Multi-h CPM, Multichannel

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV27

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV28

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV29

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV30

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV31

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV32

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV33

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV34

    www.nova-eng.com513-642-3000

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV35

    www.nova-eng.com513-642-3000

    Summary

    Sheet1

    ModulationChannel spacing for 1 dB loss at I/S ratio = 0 dBChannel spacing for 1 dB loss at I/S ratio = 10 dBChannel spacing for 1 dB loss at I/S ratio = 20 dBEb/N0 for BER = 1.0e-5, with I/S ratio = 20 dB

    SOQPSK-A--1.11 bit rates14.5 dB

    SOQPSK-B1.0 bit rates1.4 bit rates1.50 bit rates12.0 dB

    Multi-h CPM--0.77 bit rates12.0 dB

    Sheet2

    Sheet3

  • ITC/USA 2001October 22-25, 2001, Las Vegas, NV36

    www.nova-eng.com513-642-3000

    Conclusions• SOQPSK and Multi-h CPM offer substantial reduction in

    bandwidth over PCM/FM, with little or no penalty in BER.• SOQPSK-B is largely insensitive to IF filtering, while

    SOQPSK-A is more “delicate”• SOQPSK-B offers better detection efficiency than

    SOQPSK-A, at the expense of slightly greater bandwidth• Multi-h CPM offers almost 50% more bps / Hz than

    SOQPSK-A, and 2.5 dB better detection efficiency.• Multi-h CPM nearly doubles the bps / Hz of SOQPSK-B,

    with the same detection efficiency.

    Performance Of SOQPSK And Multi-h CPM In The Presence Of Adjacent Channel Interference���Terrance J. Hill�Nova Engineering, Inc.,�Cincinnati, OHOutlineWhy These Waveforms?CPM Waveforms Offset QPSK VariationsSOQPSKDefinition of SOQPSK-A, -BFrequency Pulse Shape, g(t)Power Spectral DensitySOQPSK-A Eye PatternSOQPSK-B Eye PatternDefinition of Multi-h CPMKey CPM ParametersFrequency Pulse and Phase TreePSD ComparisonPSD is Half the StorySlide Number 17Slide Number 18Slide Number 19Slide Number 20Slide Number 21Slide Number 22Experimental SetupSlide Number 249 Mbps SOQPSK-A, Multichannel9 Mbps Multi-h CPM, MultichannelSlide Number 27Slide Number 28Slide Number 29Slide Number 30Slide Number 31Slide Number 32Slide Number 33Slide Number 34SummaryConclusions