E225C Lecture 1 Wireless Systems Overview

1
E225C Lecture 1Wireless Systems Overview
Bob Brodersen
2
Course Outline

• Goal 1 The implementation of signal processing
  systems in CMOS technology
• A design methodology starting from a high level
  description through to an implementation
  optimized for hardware constraints.
• Goal 2 To understand the issues involved in the
  design of wireless systems
• Wireless systems will be used as a design driver
  to understand how to make tradeoffs in signal
  processing implementation

3
Homework and Projects

• First part of the semester (up to the break) will
  be approximately (bi)weekly homeworks that will
  implement each block of a wireless transceiver
• A final project will be to put a complete system
  together and demonstrate it on BEE

4
Lots of new radio systems being developed now
(Actually some not so new.just a long time
coming)

• WiFi 10-100Mbits/sec unlicensed band
• OFDM, M-ary coding
• 3G .1-2 Mbits/sec wide area cellular
• CDMA, GMSK
• Bluetooth .8 Mbit/sec cable replacement
• Frequency hop
• ZigBee .02-.2 Kbits/sec low power, low cost
• QPSK
• UWB Recently allowed by FCC
• Short pulses (no carrier), bi-phase or PPM

5
Communication systems Major technology driver
Digital Cellular Market (Phones Shipped)
6
Why so many new systems?

• The availability of unlicensed spectra
• Licensed
• 2G
• 3G
• Is this exploitation of unlicensed bands a
  temporary aberration or the new reality???

• Unlicensed
• WiFi
• Bluetooth
• ZigBee
• UWB

7
FCC Chairman Powell statement

• We are still living under a spectrum "management"
  regime that is 90 years old. It needs a hard
  look, and in my opinion, a new direction.
• Historically, I believe there have been four core
  assumptions underlying spectrum policy
• Unregulated radio interference will lead to
  chaos
• Spectrum is scarce
• Government command and control of the scarce
  spectrum resource is the only way chaos can be
  avoided
• The public interest centers on government
  choosing the highest and best use of the
  spectrum.

8
Powells statement (cont.)

• Today's environment has strained these
  assumptions to the breaking point.
• Modern technology has fundamentally changed the
  nature and extent of spectrum use. So the real
  question is, how do we fundamentally alter our
  spectrum policy to adapt to this reality?
• The good news is that while the proliferation of
  technology strains the old paradigm, it is also
  technology that will ultimately free spectrum
  from its former shackles.

9
Sharing

• So it looks like we are moving into a new regime
  that will have an ever larger number of competing
  radio systems that will require new technological
  solutions.

10
The FCC has been following this strategy
UWB
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
Reference Part 15 of the FCC Rules, September
2000.
11
Comparison

• Now for a quick description of the various
  technical differences between these new radio
  systems.
• These show the range of design constraints that
  we will need to address

12
Data rate
UWB
100 Mbit/sec
802.11g
802.11a
802.11b
10 Mbit/sec
1 Mbit/sec
3G
Bluetooth
ZigBee
100 kbits/sec
ZigBee
10 kbits/sec
UWB
0 GHz
2 GHz
1GHz
3 GHz
5 GHz
4 GHz
6 GHz
13
Range
10 km
3G
1 km
100 m
802.11b,g
802.11a
Bluetooth
10 m
ZigBee
ZigBee
UWB
UWB
1 m
0 GHz
2 GHz
1GHz
3 GHz
5 GHz
4 GHz
6 GHz
14
Power Dissipation
10 W
802.11a
802.11bg
3G
1 W
100 mW
Bluetooth
UWB
ZigBee
10 mW
ZigBee
UWB
1 mW
0 GHz
2 GHz
1GHz
3 GHz
5 GHz
4 GHz
6 GHz
15
Cost (projections)
1000
3G
100
802.11a
802.11b,g
UWB
10
Bluetooth
ZigBee
ZigBee
1
UWB
.10
0 GHz
2 GHz
1GHz
3 GHz
5 GHz
4 GHz
6 GHz
16
Infrastructure cost
3G
1000
802.11a
100
802.11b,g
10
UWB
Bluetooth
ZigBee
ZigBee
1
UWB
.10
0 GHz
2 GHz
1GHz
3 GHz
5 GHz
4 GHz
6 GHz
17
60 GHz???
Oxygen absorption band
Prohibited
Space and fixed mobile apps.
Wireless LAN
Japan Europe U.S.
Radar
Test
Unlicensed Pt.-to-Pt.
Wireless LAN
Mobile ICBN
Road Info.
Unlicensed
ISM
56 57 58 59 60 61 62 63
64 65 66
Frequency GHz
(Gary Baldwin)
18
CMOS can do it
f
t
100GHz
0.13u
0.18u
30GHz
0.25u
0.35u
0.5u
10GHz
0.6u
0.8u
1u
3GHz
1.5u
1GHz
2u
3u
Slope is 1/l2
CMOS
75
79
81
83
85
87
89
91
93
95
97
99
01
03
77
Year
19
Applications

• Of course the most critical issue is what are
  these various radio systems useful for and who
  will buy them!

20
Issues in System Implementation
21
Wireless System Design Technologies

• It is now possible to use CMOS to integrate all
  analog and digital radio functions.
• New theories of wireless signal processing.
• What makes an algorithm appropriate for
  implementation is rapidly changing
• Complex analog circuits linearly degrading
• Digital computation exponentially improving
• Low power consumption has become increasingly
  important.

22
Potential System Limitations

• Analog impairments digital compensation and
  signal processing.
• Multiple access and interference code diversity
  (CDMA), time diversity (TDMA), frequency
  diversity (OFDM), or spatial diversity (MIMO)
• Multipath frequency spreading, time-domain
  equalization, or frequency-domain equalization.
• Integration with existing wired infra-structures.
• Protocol efficiency to QoS or not to QoS?

23
Communication Algorithms and Their Implementation

• Blast algorithms (Lucent) - antenna arrays which
  have demonstrated 40 b/s/Hz (1Mb/s in 25kHz)
• Multi-user detection - eliminates interference
  from other users
• OFDM - eliminates multi-path and ISI
• Digital implementation of timing and carrier
  synchronization

Requires 100s of GOPs of processing how to
do it at the lowest energy and smallest area???
24
CMOS Radio-on-a-Chip
I
8
DAC
DAC
Q
8
P
S
Baseband Processor
D
I
8
ADC
ADC
Q
8
25
Wireless Channel Multipath Effects
Receiver
Transmitter
26
Inter-Symbol Interference (ISI)
Transmitted data
Received data

• Equalization or combining
• Complexity, performance (TDMA or CDMA)

• Code as multiple low-rate streams
• Each stream at different frequency - OFDM

27
Introduction to OFDM Modulation
X1
X2

X3
X4
Symbol

• Different data per tone
• Multipath just scales tones
• Tones remain orthogonal even with multipath

Frequency
28
Design Example 5GHz WLAN Standard
...
20 MHz
20MHz OFDM channels in 5 GHz band

• 802.11a and Hiperlan II have very similar OFDM
  PHYs
• 20 MHz channel is divided into 64 carriers
• Carriers are coded with varying modulation and
  error correction code.
• Each carrier is 300kHz wide, giving raw data
  rates from 125kb/s to 1.5Mb/s

29
Symbol Encoding
OFDM (52 of 64 sub-carriers used)
20 MHz

• Channel sampled at 20MHz
• 64-sample (3.2us) per symbol
• 16-sample (0.8us) cyclic prefix / guard interval
• 250 Ksymbols per second
• Of 64 the subcarriers
• 12 zero subcarriers (in black) on sides and
  center
• 48 data subcarriers (in green) per symbol
• 4 pilots subcarriers (in red) per symbol for
  synchronization

30
Data Encoding

• Data subcarrier encoding
• BPSK, QPSK, 16QAM, 64QAM
• 1, 2, 4, 6 bits/subcarrier
• Error corrective coding
• 1/2, 2/3, or 3/4 rate convolutional code
• Increased robustness
• Overall data rates
• 6, 9, 12, 18, 24, 36, 48, 54 Mbps
• Lowest 48 1 1/2 250K 6 Mbps
• Highest 48 6 3/4 250K 54 Mbps

31
Integrated Baseband Chip
32
A Wireless System is More Than DSP

• Analog RF circuits
• Amplifiers
• Synthesizers
• Mixers
• Passive components
• Analog baseband circuits
• Amplifiers
• Filters
• A/D and D/A converters
• Protocols

33
Transmitter Block Diagram
34
Receiver Block Diagram
LOIF(Q)
35
CMOS Integrated Analog Chip
36
CMOS Cost Model

• Cost It doesnt matter what you do on a CMOS
  chip, the cost is approximately constant and then
  reduces over time (e.g. .10 per mm2)
• Cost of different data rates will be the same
  order of magnitude from kilobits-gigabits/sec.
• Cost is weakly dependent on carrier frequency
  (actually might get cheaper as the frequency goes
  up since passives are smaller)
• Cost increases weakly as a function of range
  (power amp)
• Moores law scaling improves the digital part of
  wireless system capabilities at nearly the same
  rate as it improves microprocessors, but doesnt
  help the analog part (actually makes that part
  more expensive).

37
Protocols MAC and Network Implementation e.g.
802.11
AP
Station

• Infrastructure mode
• Access Point (AP)
• Essentially a bridge between wireless cells and
  wired infrastructure
• Provides authentication, packet forwarding
• Stations associate with a particular AP
• Stations may roam with no loss of service
• Roaming mechanism provides redundancy and
  robustness in addition to mobility
• Ad-hoc mode
• Ad-hoc mode allows operation without any AP

38
Protocol enhancements

• New capabilities
• Spatial multiplexing (beam-forming)
• Multi-hop routing
• Requires
• MAC modifications
• Coordination for multi-beam operation
• More centralized scheduling for efficiency
• Compatible with standardized protocols

39
Basestation of Today
Non-sectorized
40
Basestation of Today
Non-sectorized
41
Basestation of the Future
Multi-link beam-formed, sectorized
Multiple simultaneous packets per sector
42
Future of Spatial Multiplexing

• Multiple transceiver chains performing adaptive
  beam-forming deliver multiple independent data
  streams in the same channel at the same time
• Use both 5.7GHz and 2.4GHz bands -gt 7 channels
• Three sectors, 50 antenna element per sector
• Total capacity 7350/230Mbps 15 Gbps!
• Assumes reuse factor of one, many chips etc.
• Dynamic switching of Gbps over multiple wireless
  logical channels

43
Wireless Multi-Hop Routing

• Route communication through intermediate nodes
• Decouple capacity from coverage
• Antenna beam-forming to create spatial diversity
• Transmit power control to limit interference

44
Focus of this Course

• 3 components of the design problem
• Algorithm specification Matlab (or C)
• Floating point, implementation independent,
  system simulation
• Architecture mapping
• Simulink for data flow
• Stateflow for control
• Hardware optimizations
• Real-time emulation
• FPGA/ASIC implementation

45
Major topic areas

• System modeling
• Channels
• Interference
• Analog impairments
• Wireless system algorithms
• AGC
• Synchronization
• Modulation/Demodulation
• Error correction
• Protocols

• Computational algorithms
• FFT
• Cordic
• Viterbi
• Architectures
• Direct mapped
• Time multiplexed
• Reconfigurable
• Software programmable