1. Wireless: Facts and Fiction Benjamin Friedlander Department of Electrical Engineering University of California at Santa Cruz Wireless Communications and Signal Processing Lab Phone: 831-459-5838 friedlan@ee.ucsc.edu

2. Focus of Talk n Fixed broadband wireless access n Physical layer (range, throughput) n Performance claims vs. “Physics of wireless” n Wireless systems vs. single links n Guidelines for sorting fact from fiction

3. How to estimate range?

4. Link Budget - #1 n Radiated Power: 50dBm u Transmit power: 30dBm u Transmit Antenna Gain: 20dB n Required Power at Receiver: -83dBm u SNR for 70MBPS: 22dB u Noise floor: -94dBm (NF = 5dB) u Receive Antenna gain: 16dB (-) u System losses: 5dB u Link Margin: 0dB n Maximum Pathloss: 133dB n Free space 30miles: 133.7dB

5. From “WiMAX’s technology for LOS and NLOS environments” a WiMax Forum publication

6. Urban & Suburban Scenarios n Free space conditions are rare n Most links will not have line of sight

7. Non Free Space Pathloss

8. Conclusion #1 n Range of 30 miles may be achieved in special / rare situations n Typical ranges will be MUCH smaller

9. Range vs. Coverage n Range u Point-to-point communication u Basestation to a specific subscriber n Coverage u Point-to-multipoint u Subscribers in random locations u Specified probability of service

10. Cell Size vs. Range 6 2 7 7 3 5 2 6 4 4 5 3 4 5 3 1 6 2 7

11. Building Penetration n Additional propagation loss for building penetration n Typically: 10 – 20 dB n Impact on range: reduce by x2 – x3

12. Conclusion #2 n To serve a large population of users with high reliability the density of base stations needs to be significantly higher than predicted by the range calculated earlier (fractions of mile)

13. So what about throughput?

14. Rate vs. Range

15. Throughput vs. Cell size

16. Factors further reducing effective throughput n Interference n Bursty traffic n Other …..

17. Conclusion #3 n Effective throughput is much smaller than indicated by maximum data rates n Performance is dominated by low data rate users

18. So how can we measure system performance?

19. Measuring System Throughput

20. Measuring Throughput n Monitor data flowing in/out over time and record peak throughput (heavily loaded system) [bits/sec] n Normalize throughput by u Total coverage area [bits/sec/m 2 ] u Total bandwidth [bits/sec/Hz] u Total number of users [bits/sec/user] n Cost per MBPS (equipment, install, operation)

21. Range & Throughput Summary n Always ask for link budget n 30 mile range at 70MBPS possible (barely) n Backhaul range likely to be on order of a mile (highly dependent on actual locations) n Typical cell size a fraction of a mile n Cell throughput (point to multipoint) will be much less than 70MBPS

22. Some final words n Communication systems (modulation, coding, etc.) are well developed. n Future improvement are likely to be incremental, not revolutionary n For given resources (BW, power) it is practically impossible to have 10X improvement n We have more technology than we know what to do with

23. For additional information n Please contact friedlan@ee.ucsc.edu n Related talks: u Smart antennas, MIMO, transmit diversity – so how many antennas do we need? u Multiaccess methods: TDMA, FDMA, CDMA, OFDMA – so what comes next? u Wireless in the wild west: operating in the unlicensed spectrum. u Communicating on the move – mobility and its limitations u The amazing story of ultra-wideband