Trends in Future RF Applications

Trends in Future RF Applications Neil C. Bird Philips Research Europe May 15 th, 2006

Outline Technical Trends Next Generation Wireless Communication in the Home Conclusions 2

Scope of RF Future Mobile Systems Car-to-X Connectivity Software Defined Radio CMOS CMOS ADC EVP ADC Cognitive Radio Lighting Networks Mobile Digital TV Disposable Radios Next Generation WPAN Wireless Sensor Networks 3

Key Technical Trends 1. Higher RF Frequencies 2. Digitisation 3. Multi-mode/Multi-band Transceivers 4

Frequency Trends 100 Frequency range [GHz] 10 1 Future (Microwave and mm-wave) Current Applications TV UHF-Band Sat LNB C-Band Satellite Front-end Mobile GSM Sat LNB Ku-Band Mobile DCS1800 WLAN 2.5 GHz WLAN 5 GHz 3G Mobile UMTS Sat Radio 2.3 GHz TV VHF-Band 0.1 AM Radio FM Radio 20 30 40 50 60 70 80 90 2000 10 20 5

Ku Band LNB Discrete implementation Expensive solution (e.g. dielectric resonators) Requires manual set-up. Standard package Alignment-free application 6

Frequency Trends Frequency range [GHz] 100 10 1 Future (Microwave and mm-wave) Current Applications TV UHF-Band Sat LNB C-Band Satellite Front-end Mobile GSM Sat LNB Ku-Band Mobile DCS1800 WLAN 2.5 GHz LDMS 28-30 GHz WLAN 5 GHz Satellite 29 GHz 3G Mobile UMTS Sat Radio 2.3 GHz Car Radar 77 GHz WLAN 60 GHz Car Radar 24 GHz TV VHF-Band 0.1 AM Radio FM Radio 20 30 40 50 60 70 80 90 2000 10 20 7

Digitisation Is both and enabled and driven by the switch to CMOS for virtually all RF applications below 6GHz In deep sub-micron CMOS Area of a microprocessor == Area of an inductor!!! (G. Moore, ISSCC 2003) 8

Digitisation Digitisation of existing analog functions Use digital circuits for Calibration Compensation Adaptability Rx BB RF IF Programmable Divider Fixed Divider VCO LO Gen DSP PFD PA Tx BB Xtal Loop Filter Analog Cal Comp Adapt Digital 9

Multi-band Transceivers and Front-ends Clear trend towards increased (RF) functionality in the handset Driven by explosion of standards and mobile applications Leads to significant RF challenges in the short and long term Broadcast Positioning Cellular Connectivity 10

A solution? Reconfig. RF Pipe Handset comprises a number of identical RF pipes, each re-configurable over digital interface Reconfig. RF Pipe Reconfig. Modem Application Engine User Interfaces USB NFC Reconfig. RF Pipe Source: P. Hooijmans Reconfigurable transceiver and RF front-end 11

But the real problems are at the front-end How to cover the complete 470MHz to 6GHz range? Broadcast Cellular Conn WiMax UWB Conn Antenna design for multi-band operation is possible 5/6-band operation has been shown to be feasible Requires several passive front-end components 1 2 3 4 5 GHz 6 Source: F van Straten Source: K Boyle 12

The Bigger Picture ~2005-2008 ~2009-2012 ~2013+ CMOS RF Trx Multi-mode Trx Multi-band Trx/FE Reconfigurable/SDR Cognitive Radio mwave Applications (Sat LNB, 24GHz Car Radar) mmwave Applications (60GHz WPAN, 77GHz Car Radar, RF Imaging) THz Imaging RFID Ultra Low Power Radio Invisible Radio 13

Some Future Challenges ~2005-2008 ~2009-2012 ~2013+ CMOS RF Trx Multi-mode Trx Multi-band Trx/FE Reconfigurable/SDR Cognitive Radio mwave Applications (Sat LNB, 24GHz Car Radar) mmwave Applications (60GHz WPAN, 77GHz Car Radar, RF Imaging) THz Imaging RFID Ultra Low Power Radio Invisible Radio 14

Outline Technical Trends Next Generation Wireless Communication in the Home Conclusions 15

Ambient Intelligence 16

Data Rates Sensors and Control A/V Streaming Data Transfer 10k 100k 1M 10M 100M 1G 10G Data Rate (bit/s) HDTV Streaming, Burst mode Bulk data transfer 6 orders of magnitude 17

Today s Wireless Standards ZigBee 20/40kbit/s 250kbit/s Bluetooth (1Mbit/s) 802.11b (11Mbit/s) 802.11a/g (54Mbit/s) 802.11n (200Mbit/s) UWB (480Mbit/s) 10k 100k 1M 10M 100M 1G 10G Data Rate (bit/s) Low(er) Power Vanishingly small Zero cost Multi-Gbit/s 18

Transmission Distance Requirement Varying data rates over 6 orders of magnitude Transmission distance varies from 1m to coverage of the entire home Issue Structure of the home walls and floors Transmission through walls implies low RF frequencies (less than 6 GHz) Consequence for achievable data rate and transmit power Interference (within the home, and from neighbours) 19

Homogeneous Network Characteristics All wireless links require through-wall capability Even for Gbit/s links (e.g. for uncompressed 10 HDTV, bulk data transfer) Implication (apartment) Neighbours receive 1 your network EDGE You can receive neighbours networks Example Data Rate [Mbit/s] 100 GPRS ZigBee 2.5 b/s/hz Upper boundary DECT Bluetooth 0.1 GSM 0.1 1 10 802.11b 8 nearest neighbours, all streaming Occupied 2 Gbit/s Bandwidth [MHz] Aggregate data rate is ~30 Gbit/s (9x2 Gbit/s + 50%) In a bandwidth of 1 GHz (if available below 6GHz) implies a required spectral efficiency of 30 bit/s/hz!! HP/2 0.5 b/s/hz Safe operation 802.11a/g 20

Heirarchical Structure with Local Data Storage Access Point WLAN WPAN 1 WPAN 2 Network takes care of data synchronisation Example 1 Gbyte of data synchronises in 13 minutes over a 10 Mbit/s room-to-room link Broadcast mode allows parallel synchronisation to all rooms 21

Data Rates - WPAN ZigBee 20/40kbit/s 250kbit/s Bluetooth (1Mbit/s) 802.11b (11Mbit/s) 802.11a/g (54Mbit/s) 802.11n (200Mbit/s) UWB (480Mbit/s) 10k 100k 1M 10M 100M 1G 10G Data Rate (bit/s) Low(er) Power Vanishingly small Zero cost Multi Gbit/s (3-10 Gbit/s) In-room only required Exploit higher frequencies 60GHz Technology 22

The Need for Speed Driver applications Wireless HDMI (requires 3+ Gbit/s) High-speed bulk data transfer HD Capacity 4GB 10GB 40GB 20GB 13min/0.4min 27min/0.9min 80GB 106min/3.6min 53min/1.8min 5min/0.2min 2005 2006 2007 2008 2009 Year Data transfer time at 100Mbit/s Data transfer time at 3Gbit/s 23

Which Technology? Classic wireless standards below 6GHz not feasible Required spectral efficiency is too high Microwave bands allowed for WLAN/WPAN usage 17GHz (Europe), 24GHz (US) Limited bandwidth (~200MHz) spectral efficiency issue Ultra Wide-Band Currently targeting 480Mbit/s for wireless USB Limited by regulations on transmit power levels Max throughput = 1+ Gbits and very short range (10s of cm)? 60GHz 24

60GHz Characteristics High oxygen absorption 15dB/km High rain attenuation Suitable for Short range applications Indoor applications Secure communications High attenuation by walls Indoor links are limited to in-room High transmit power compared to UWB Flexibility in the selection of waveforms/modulation 25

Regulatory Aspects Current regulatory situation Data Rate [Mbit/s] 1000 100 10 1 60GHz 0.1 0.1 1 10 100 Occupied Bandwidth [MHz] 1000 Worldwide overlap of 3 GHz (59-62 GHz) But channelisation of 1.75GHz possibly required 26

Data Rates - WPAN ZigBee 20/40kbit/s 250kbit/s Bluetooth (1Mbit/s) 802.11b (11Mbit/s) 802.11a/g (54Mbit/s) 802.11n (200Mbit/s) UWB (480Mbit/s) 10k 100k 1M 10M 100M 1G 10G Data Rate (bit/s) Low(er) Power Vanishingly small Zero cost Multi Gbit/s (3-5 Gbit/s+) In-room only required Exploit higher frequencies? 60GHz 27

What is Ultra Low Power (ULP) Radio? Low Data Rate (kbit/s) Low power consumption (mw) Ultra Low Power Radio Low cost (< $1) Small size (low volume) 28

ULP Typical Applications Sensing/Sensor Networks Intelligent Tagging Medical (Implantables) Ambient Intelligence Heating control Lighting control Location of objects Sensing 29

ULP What is available today? Typically one or more ICs and other components on a PCB Parameters Power consumption 10s of mw for e.g. 250kbits/s. Improvements are coming External components 10s off Cost >> $1 However, still a long way to from the ideal ULP concept! 30

ULP Radio Target? Disclaimer: depends on application, but in general, smaller is better All of the things that have been mentioned previously, but improved in all aspects Think of a postage stamp! Example: for medical sensors just rip-and-stick! 32

Outline Technical Trends Next Generation Wireless Communication in the Home Conclusions 33

Last Words No shortage of research challenges! For today Efficient and effective implementation of highly digitised transceivers in CMOS For tomorrow Towards mm-wave frequencies (60GHz and beyond) Multi-band problem High-Q tunable passive filtering Invisible Radio New materials and devices Just to be able to solve the problems But also to provide differentiation (after all, everyone is using the same CMOS processes ) 35