Cellular Wireless Capacity & Scheduler/Medium Access Control (MAC)

1. Cellular Wireless Capacity & Scheduler/Medium Access Control (MAC) Shivkumar Kalyanaraman shivkumar-k AT in DOT ibm DOT com http://www.shivkumar.org Google: “shivkumar ibm rpi” Based upon slides of Nitin Vaidya, Sorour Falahati, Timo O. Korhonen,P. Viswanath/Tse, A. Goldsmith, & textbooks by D. Mackay, A. Goldsmith, B. Sklar & J. Andrews et al.

2. Outline • Overview of key ideas from Pt-Pt & Multi-user Information Theory for Fading channels • Cellular MAC Protocols: • Scheduled MAC • Proportional Fairness & Interference Management

3. Pt-pt (fading) channel Capacity • A slow fading channel is a source of unreliability: very poor outage capacity. Diversity is needed. • A fast fading channel with only receiver CSI has a capacity close to that of the AWGN channel. Delay is long compared to channel coherence time. • A fast fading channel with full CSI can have a capacity greaterthan that of the AWGN channel: fading now provides more opportunities for performance boost. • The idea of opportunistic communication is even more powerful in multi-user situations.

4. Capacity of AWGN Channel Capacity of AWGN channel If average transmit power constraint is watts and noise psd is watts/Hz,

5. Power and Bandwidth Limited Regimes Bandwidth limitedregime capacity logarithmic in power, approximately linear in bandwidth. (eg: WLANs) * 2-3dB gains matter! 3dB => double capacity! Power limitedregime capacity linear in power, insensitive to bandwidth. (eg: cell-edge in WMAN cells)

6. Application: Freq Reuse in Cellular Interference is modeled as noise For hexagonal topology: optimal reuse factor is 1

7. Fading Channels: Traditional Approach to (Multi-user) Wireless System Design Compensates for channel fluctuations. I.e. treats a multi-user channel like a set of disjoint single-user (or pt-pt) channels. Examples: interference averaging; near-far power control, fixed coding/modulation rates

8. Example: CDMA Systems Two main compensating mechanisms: 1. Channel diversity: • frequency diversity via Rake combining • macro-diversity via soft handoff • transmit/receive antenna diversity 2. Interference management: • power control • interference averaging

9. What Drives this Approach? Main application is voice, with very tight latency requirements. Needs a consistent channel.

10. Opportunistic Communication: A Different View Transmit more when and where the channel is good. Exploits fading to achieve higher long-term throughput, but no guarantee that the "channel is always there". Appropriate for data with non-real-time latency requirements (file downloads, video streaming).

11. Point-to-Point (Flat) Fading Channels Capacity-achieving strategy is waterfilling over time.

12. Variable rate over time: Target BER • In the fixed-rate scheme, there is only one code spanning across many coherence periods. • In the variable-rate scheme, different codes (distinguished by difference shades) are used depending on the channel quality at that time. • For example, the code in white is a low-rate code used only when the channel is weak.

13. Performance over Pt-Pt Rayleigh Channel Not much bang-for-buck for going to (sender) CSI from CSIR @ high SNR

14. Performance: Low SNR At low SNR, capacity can be greater (w/ CSI) when there is fading. Flip side: harder to get CSI at low SNR 

15. Hitting the Peaks @ Low SNR: Hard in Practice! (High SNR) Fixed power almost as good as waterfilling At low SNR, one can transmit only when the channel is at its peak. Primarily a power gain. In practice, hard to realize such gains due to difficulty in tracking the channel when transmitting so infrequently. (Low SNR) Waterfilling helps, But CSI harder & users pay delay penalties

16. Multiuser Opportunistic Communication Multiple users offer new diversity modes, just like time or frequency or MIMO channels

17. Performance AWGN Increase in spectral efficiency with number of user at allSNR’s, not just low SNR!

18. Multi-user w/ CSI: Low SNR case

19. Multiuser Diversity Total average SNR = 0 dB. • In a large system with users fading independently, there is likely to be a user with a very good channel at any time. • Long-term total throughput can be maximized by always serving the user with the strongest channel.

20. Sum Capacity: AWGN vs Ricean vs Rayleigh • Multiuser diversity gain for Rayleigh and Ricean channels ( = 5); KP/N0 = 0 dB. • Note: Ricean is less random than Rayleigh and has lesser sum capacity!

21. Multiuser Diversity & MAC/Schedulers • Independent fading makes it likely that users peak at different times. • In a wideband system with many users, each user operates at low average SNR, effectively accessing the channel only when it is near its peak. • In the downlink, channel tracking can be done via a strong pilot amortized between all users. • Scheduler / MAC implication: it’s the job of a MAC protocol to achieve this multi-user capacity, while respecting fairness on the medium term.

22. optimal Uplink AWGN Capacity successive cancellation: cancel 1 before 2 conventional decoding cancel 2 before 1

23. AWGN Multiuser Capacity & SIC Decoder • User 1 can achieve its single-user bound while at the same time user 2 can get a non-zero rate: • Each user encodes its data using a capacity-achieving AWGN channel code. • 2-stage decoding: • 1. Decodes the data of user 2, treating the signal from user 1 as Gaussian interference. • 2. Once the receiver decodes the data of user 2, it can reconstruct user 2’s signal and subtract it from the aggregate received signal. • Then decode the data of user 1. • Only the background Gaussian noise left in the system, the maximum rate user 1 can transmit at is its single-user bound log (1 + P1/N0). • This receiver is called a successive interference cancellation (SIC)

24. MAC / Scheduling Protocols

25. MAC Protocols: a taxonomy Three broad classes: • Channel Partitioning (eg: simple TDMA) • divide channel into smaller “pieces” (time slots, frequency) • allocate piece to node for exclusive use • Random Access (eg: WiFi & some control channels in cellular) • allow collisions • “recover” from collisions • “Taking turns” (eg: used in cellular schedulers) • Tightly coordinate shared access to avoid collisions • Use this paradigm to maximize statistical mux gains, opportunistic channel access etc

26. OFDMA-based MAC OFDMA reduces the PAPR problem in OFDM:each MS transmits in a small set of sub-carriers with a lower PAPR (PAPR increases with the number of subcarriers), and also with far lower total power than if it had to transmit over the entire bandwidth.

27. Sum Capacity vs Proportional Fairness • Sum capacity / waterfilling across users is a “greedy” approach. • Some users with poor channels may be starved. • How to bring in fairness constraints? • Proportional fairness, but over time: (not in any one scheduling frame) Dynamic priority: Users consistently underserved are weighted higher

28. Tradeoff: Spectral Efficiency vs Fairness Tradeoff: ~0.25-0.4bps/Hz Can weight the PF constraints: function of relative SINRs to get bulk of multi-user diversity

29. OFDMA MAC structure

30. OFDMA: Latin Squares & Hopping Patterns • Hopping pattern matrix: (coordinated w/ neighboring BS) • Interference diversity

31. Wimax QoS Categories VoIP over WiMAX using ErtPS can fit an order of magnitude (2x) more calls/sector compared to ckt-switched methods used in 2.5G/3G

32. Extra Slides

33. MAC sub-layers in WiMAX

34. Uplink Fading Channel: Summary

35. Details: Capacity of Flat Fading Channels Three cases Fading statistics known Fade value known at receiver Fade value known at receiver and transmitter Optimal Adaptation with TX and RX CSI Vary rate and power relative to channel Goal is to optimize ergodic capacity

36. Waterfilling: Optimal Adaptive Scheme Power Adaptation Capacity 1 g g0 g Waterfilling

37. Random access: Slotted Aloha • time is divided into equal size slots (= pkt trans. time) • node with new arriving pkt: transmit at beginning of next slot • if collision: retransmit pkt in future slots with probability p, until successful. • Maximum throughput: 37% Success (S), Collision (C), Empty (E) slots

38. Carrier Sense Multiple Access (CSMA) • In some shorter distance networks, it is possible to listen to the channel before transmitting • In radio networks, this is called “sensing the carrier” • The CSMA protocol works just like Aloha except: If the channel is sensed busy, then the user waits to transmit its packet, and a collision is avoided • This really improves the performance in short distance networks!

39. A B C Hidden Terminal Problem Nodes A and C cannot hear each other Transmissions by nodes A and C can collide at node B Nodes A and C arehidden from each other 802.11 RTS/CTS mechanism aims to tackle this issue, but it is often turned off in deployments!