Analysis of Changes in PMSE Spectrum and Transmission Technology - Our Prediction for PMSE, Operated Under Changed Conditions - Update 2017 WM-01 Current and Future Use of Spectrum by PMSE 4th PMSE Workshop at EuMW Prof. Dr.-Ing. Georg Fischer Institute of Electronics Engineering The birthplace of
34
Embed
Analysis of Changes in PMSE Spectrum and Transmission ... · Analysis of Changes in PMSE Spectrum and Transmission Technology -Our Prediction for PMSE, Operated Under Changed Conditions
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Analysis of Changes in PMSE Spectrum and Transmission Technology - Our Prediction for PMSE, Operated Under Changed Conditions - Update 2017WM-01 Current and Future Use of Spectrum by PMSE
4th PMSE Workshop at EuMWProf. Dr.-Ing. Georg FischerInstitute of Electronics Engineering
The birthplace of
1. What’s hot?
2. Review of Digital transmission technology
3. Crest factor and Pulsing of interferers
4. Airband use?
5. How to prepare for the future?
6. Conclusions
2
Content
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
1. What’s hot?
3
4EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
What’s hot?Interpretation
Transmission Technology• Digital PMSE has arrived in daily use• Practical experience tells that Digitization of PMSE does not allow for less spectrum• Digitization of PMSE is no relief to spectrum shortage by Digital Dividend I+II• Transition to Digital PMSE brings more quality but also new challenges
New Frequency Bands• Duplex gaps of cellular systems get more used, however crackling noise• Airband is under discussion, however compatibility unclear
Convergence in Media transmission• Broadcast and cellular are in fruitful talks• Broadcast has brought in requirements catalogue into 3GPP standardization for 5G• 5G MEDIA initiative founded in Munich 2017• Research: Can PMSE also join the Media convergence? See PMSE-xG project
New Physical layer for PMSE?• Can the 5G physical layer be used for PMSE? Research ongoing…• How does a wideband air interface behave relative to todays narrowband interface?• Some parties are advocating a wideband air interface for PMSE…
2. Review of Digital transmission technology
5
Availability• Drop-outs not acceptable, 100% reliable (telecoms 95%) • Historic events cannot be repeated “Daß Ihre Ausreise…(Genscher)”• No Clicks acceptable, especially in quiet periods (e.g. Digital out of sync…)
Quality• High quality must be captured in production archive as otherwise it is not possible to
derive the various quality levels in distribution• Large compression is thus not possible on production, only on distribution
Latency• The “drummer problem”, roundtrip from wireless mic to wireless IEM should be < 5ms
Telephone 200ms, Cellular 20ms, Satcom 500ms• Communication theory wise: Information source and information sink are collocated!
Battery operation• PMSE equipment (mics, IEM) at artist operated from battery• Part of costumes
6
Review of Digital transmission technologyPMSE specifics
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
PMSE has special demandsExpectations regarding spectrum savings from digitization of PMSE too high
Findings• Wireless Technology on Production 2x (MIC+IEM)• Wireless Technology on Distribution• Large variety of distribution CODECs• Life productions – it must work reliable!
7
Review of Digital transmission technologyPMSE specifics
Production Distribution
High Quality AudioArchive
(Former Master Tape)
Broadcast
CD/DVD
InternetMP3 downloadMusic streaming
Roundtrip Latency?
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
IEM
PMSEwireless
wireless
wireless
wireless
What’s not captured in the archive,cannot be distributed later!
8
Review of Digital transmission technologyInformation theoretical view
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Model• Information from information source (microphone) is sent to information sink
(mixing console, archive, distribution, IEM)• Goal is to minimize transinformation, it defines spectrum need• Redundancy: we can strip this off without loss of quality• Irrelevance: debatable …who decides? Big brother? The receiver should...• The more irrelevance is identified, the less transinformation, the less spectrum
is needed…
Information source
Information sink
IrrelevanceRedundancy
Transinformation
Error
Spectrum need defined by transinformation
CompressionSource coding
9
Review of Digital transmission technologyDigital wireless transmission system
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Information source
Source codingCompression
Channel CodingProtection
Channel Sounding Training sequence
Synchronisation information
Multiplexer
Transmitter
Overhead
ProtocolSignalling
Digital Transmission comes at the cost of overhead
10
Review of Digital transmission technologyPrice to pay for digital – overhead with digital
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Digital transmission• Source information is sourced coded (compressed)• Channel coding needed for robustness of transmission• Overhead not negligible!
Raw data
source coded data
Channel codeddata
Coded data includingoverhead
Digital only pays off if we gain more by compression,than we pay by overhead
11
Review of Digital transmission technologyPrice to pay for digital – overhead with digital
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Today’s digital PMSE does not provide Studio quality
0,0
0,5
1,0
1,5
2,0
2,5
3,0
3,5
4,0
4,5
5,0
0,10,30,51,02,04,08,016,032,064,0
Mbi
t/s
Compression factor
Codec quality & datarate
CD (reference)
Studio I24bit@96 kSa/s
Studio II24bit@192 kSa/s
PMSE(standard)PMSE
(robust)DABDAB+DRM
12EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Review of Digital transmission technologyTransmission robustness (Channel coding)
Diversity• Makes transmission robust, various flavours
polarisations at one location• Pattern diversity - different antenna types
Temporal diversity• Channel coding, block codes, interleaving• 2G GSM spreading over 40 ms (20 ms blocks), 3G UMTS 10 ms blocks, 4G LTE
ACK/NACK protocol 4..8 ms• PMSE demands 1..2 ms one way!
Frequency diversity• Frequency hopping in narrowband systems, e.g. GSM across 35…75 MHz• Frequency hopping also provides interference diversity• Wideband interfaces, e.g. W-CDMA/UMTS or OFDM/LTE, width 1.25….20 MHz• Coherence bandwidth of channel typ. 1.5 MHz • PMSE narrowband, only 200 kHz channel, historic reasons, similar analogue FM Radio
time
frequency antenna(space, polarization,
pattern)
13EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Review of Digital transmission technologyDiversity schemes
Antenna diversity• Receiver diversity already in place with PMSE receivers• Could migrate to 2x2 MIMO ?• Complexity impact to Wireless mic?• Massive MIMO? Even more antennas, partly done with distributed antenna systems
Temporal diversity• Not a realistic option for PMSE due to stringent latency requirement• Cellular much less latency critical, so they can use this for robustness
Frequency diversity• Today’s PMSE is very narrowband • Wideband techniques would also provide interference diversity• Cognitive spectrum management already leads to frequency handovers due to
interference
Summary• Only antenna and frequency/interference diversity are realistic options as they do not
imply large latency• Temporal diversity (interleaving) no option due to latency hit!
3. Crest factor and Pulsing of interferers
14
15
Crest factor and Pulsing of interferersProblem description
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Characteristics of interferers to PMSE• Often use advanced air interfaces, higher order modulation format• strong fluctuation of momentary power, no constant interferer power• Have Crest, 12..15 dB peak to average
Interference mechanism• If channel sounding training sequence or sync symbols are hurt, then also data is hurt• Chanel sounding and syn symbols are central points of failure• No similar mechanism on analogue• Loss of sync leads to long drop out - 500 ms?
Consequence• Stay away from loss of sync• Apply sufficient margin to noise and interference• Practically digital does not allow for more dense packing of frequencies than analogue
Resync on digital implies long drop out
16
Crest factor and Pulsing of interferersAnalogue versus digital
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Digital is better here
Source: Sennheiser
Digital is much worse here
The digital cliff
The analogue cliff
Digital cliff -versus- analogue soft degradation
17
Crest factor and Pulsing of interferersInterference into PMSE
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Characteristics of PMSE• PMSE as victim is very vulnerable by interference• Due to high audio quality requirements in PMSE, it is operating at a fairly high SNR
around 20…40 dB, other systems typically operate at low SNR e.g. 7 dB• Due to stringent latency requirements in PMSE (5ms round trip), channel coding by
interleaving can only by spread across a block of 1 ms, thereby not delivering strong temporal diversity gains
• Interference to PMSE through strong peaks will make block decoding fail
Regulation• So far protection levels given as average power levels (interference temperature)• Better specify peak levels
The strong latency and quality requirementsmake PMSE very vulnerable to peaky interference
18
Crest factor and Pulsing of interferersInterference into PMSE
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Interference from LTE shoulder• PMSE might be operated next to LTE basestations and terminals, Duplex gap• Side emissions are specified as average power limits, however peak levels are
crucial for PMSE• If quantifying interference peak levels should be assessed
Crest factor at channel 7…10 dB
Crest factor at shoulder 40 dB
Average power on shoulder not meaningful!
LTE BS
4. Airband use?
19
Airband use?Interference situation
New interference situation at L-Band• Multiple systems present in L-Band• DME Distance Measurement Equipment (civil and military versions)• LDACS1 digital avionic radio coming up• Military Radio Systems: JTIDS/MIDS
DME• Double Pulses (pulse pairs), ideally Gaussian shaped• Realistically sharper pulses are used, occupying larger bandwidth• Repetition rates: Ground station 2700 ppps (pulse pairs per second), Military version
TACAN 3600 ppps, Aircraft 150 ppps in the search mode and 30 ppps in the track mode
12 μs
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017 20
Source: Hirschbeck Martin, Interference Mitigation and Channel Estimation for Digital Aeronautical Communication Systems based on OFDM, FAU, July 2016
21
Airband use?Interferers
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
LDACS• L-Band Digital Aeronautical Communication System• OFDM System• Interleaved with DME
Notations• UAT=Universal Access Transceiver• SSR=Secondary Radar System• JITDS=Joint Tactical Information Distribution System• MIDS=Multi-functional Information Distribution System
Source: Hirschbeck Martin, Interference Mitigation and Channel Estimation for Digital Aeronautical Communication Systems based on OFDM, Dissertation to be defended, FAU, July 2016
22
Airband use?Spectrum assignments in L-Band
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Coordination• Forward link FL and reverse link RL coordinated between LDACS and DME
Source: Hirschbeck Martin, Interference Mitigation and Channel Estimation for Digital Aeronautical Communication Systems based on OFDM, Dissertation to be defended, FAU, July 2016
23
Airband use?What about LDACS interference by DME?
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
LDACS being interfered by DME• Time domain view
• On average every third OFDM Symbol is hurt• Block length in OFDM (e.g. 20 ms?) much higher than in PMSE (2ms)• Larger block length gives higher interleaving gain• Larger block length possible in LDACS as less latency critical
Source: Hirschbeck Martin, Interference Mitigation and Channel Estimation for Digital Aeronautical Communication Systems based on OFDM, Dissertation to be defended, FAU, July 2016
OFDM Symbol120 μs
24
Airband use?Time domain view
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
DME Interference• DME consist of double pulses with 12us distance• Pulses are nearly Gaussian• DME runs at 2700…3600 ppps (pulse pairs per second)
PMSE as victim• Goal 5 ms roundtrip, 2 ms one way plus 1ms for mixing console• Then in 2 ms there are 7 double pulses• A double pulse occupies roughly 20 us duration of interference• A duration of 20 us * 7= 140 us out of 2 ms would be interfered• 7% of PMSE transmission interfered• If DME pulses hit pilot symbols in PMSE transmission, whole frame would be
immediately lost• In case of incoherent PMSE transmission there would be no pilots, e.g. CPM
incoherent, however incoherent CPM is limited to about 3 bit/s/Hz
Conclusion• DME pulses happen several times in a PMSE frame
The assumption of seldom hits is not true!
25
Airband use?Power levels
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Transmit power• TACAN (military version) ground max 4 kW=+66 dBm• DME ground 1 kW• DME Board 100W…500W (2kW militärisch)• PMSE max 50mW, typ 10 mW =+10 dBm• Sensitivity PMSE typ -95 dBm
Relations• DME transmission power factor 1000…40,000 equal 56 dB higher than that of PMSE
Interpretation• PMSE is a system with short frame length and short interleaving• Temporal diversity through interleaving is rarely in place• For quantifying DME interference into PMSE receiver, peak and not average receive
level of interference should be used.
Note• The Shannon bound is valid only for long (infinite) time interleaved systems• PMSE as a latency critical system cannot be described by Shannon and is vulnerable
through peak interference• PMSE cannot profit from interleaving as a protection mechanism
26
Airband use?SNR hit by latency constraint
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Theoretical background• Required SNR predicted by Shannon bound assumes infinitely long interleaving• No analytic formula to derive SNR hit by latency constraint• Nevertheless theoretical bounds on hit can be identified• Hit on the order of extra 5 dB to meet latency constraint, under certain cases may go
up to 10 dB
Source: Rachinger Ch., Huber J., Müller R., Comparison of Convolutional and Block Codes for Low Structural Delay, IEEE Transactions on Communications,, Volume: 63, Issue: 12, Pages: 4629 - 4638, 2015
SNR hit5 dB
5. How to prepare for the future?(media convergence…)
27
28
How to prepare for the future?PMSE and broadcast
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
• Production companies• Need for PMSE
Content& Event
production
Content management
Content distribution
• Broadcaster’s core business!
• Public address• TV/Radio• Internet• Mediathek• CD/DVD
Situation has changed• Broadcasters no longer are present throughout the whole business chain• Broadcasters focus on core business - content management• Broadcasters no longer represent PMSE and Broadcast spectrum inside one institution
• Downloads• Internet• Satellite• Cable• Shop...
The Classical Model for PMSE coexistence with Broadcastno longer works
29
How to prepare for the future?Broadcast via eMBMS inside 5G
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Source: BFS IMB5
Broadcast over cellular• New business opportunities• TV Distribution• Software Update• Local advertisements• Fleet Management• Car info• Tourist/Travel info
Research• Project in Germany BFS IMB5• Partners: IRT, NOKIA, FhG IIS, R&S, BR, BMW
Technical approach• Longer CP as in DVB-T/T2• 16.6 / 33.3 us, equal to 5 / 10 km• SFN (Macro Diversity gain)
30
How to prepare for the future?Broadcast via eMBMS inside 5G
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Source: BFS IMB5
2030…
DVB-T2 Introduction
2020201920182017201620152014
5G-Standardisation in 3GPP
Pre 5G introducedat Olympic Winter Games in southKorea
5G at OlympicSummer Games in Japan
Follow up projectBFS IMB5
Introduction of convergent
system
Schedule• DVB-T and DVB-T2 unquestioned• Convergence of Broadcast and cellular
31
How to prepare for the future?Some thoughts…
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Today’s PMSE• Narrowband PMSE in 200 kHz does not allow for frequency diversity• Time diversity cannot be explored due to stringent latency requirements• Antenna diversity is already explored• Migration to MIMO e.g. 2x2• Massive MIMO for large events?
Economies of scale• For historic reasons PMSE uses analogue FM and Stereo as in analogue FM broadcast
with 200 kHz channel• Allows for reuse of Integrated Circuits ICs: FM Demodulator, Stereo decoder with IEM• Instead of reusing technology from Broadcast market, reuse technology from cellular
market, Basestations, handset modems (see PMSE-xG project)• Reusing cellular technology would allow for: MIMO, Wideband air interface, interference
diversity, QoS management, Macrodiversity by SFN, eMBMS, new services…
PMSE equipment vendors• Have partially distinguished themselves based on radio performance• Reusing cellular technology would leave their USP in the area of acoustic and sound
properties – acceptable?
Media convergence Cellular, Broadcast, PMSE?
6. Conclusions
32
33
Conclusions
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
Digital transmission• PMSE has unique demands (Latency, quality, availability, drop outs, clicks, compression,
battery operation• Robustness comes through diversity
Coexistance with other systems• With interferers having fluctuating power, peak power relevant, not average• Examples: Airband DME, LTE Duplex gap, shoulders of BS and MS
Diversity• The root of robustness with digital transmission• Time diversity cannot be explored in PMSE due to stringent latency requirement• Frequency and interference diversity not in place with todays 200 kHz narrowband PMSE • Wideband transmission desirable• Antenna diversity already in use today, however room for improvement, MIMO, massive
MIMO
Media Convergence• An option on the horizon thanks to 5G• Cellular and Broadcast already in talks on convergence through eMBMS, derivation ofa
catalogue of requirements• Cellular and PMSE already in research activities regarding necessary adaptations on 5G
standard
34
Conclusions
EUMW2017, WM01, Workshop on PMSE, Nuremberg, 9th Oct. 2017
And the PMSE vendors?• May have to adjust their business• Radio no longer a distinguishing feature• Concentrate on sound and acoustic properties• Capsule, transducer competence is key• Uncompressed and studio quality will get a realistic option, compression technology no
longer a USP?
And the PMSE users?• Get familiar with cellular technology• Operational service, local MNO?• More diversity, more robustness• New services to be administered
Opportunities by 5G tremendous!Can it deliver what PMSE requires?