Directional Antennas for Wireless Networks Romit Roy Choudhury

Directional Antennas for Wireless Networks Romit Roy Choudhury 1

Several Challenges, Protocols Applications Internet 2

Omnidirectional Antennas Internet 3

IEEE 802. 11 with Omni Antenna RTS = Request To Send CTS = Clear To Send M S Y RTS D CTS X K 4

IEEE 802. 11 with Omni Antenna silenced M S Data Y D silenced ACK X silenced K 5

IEEE 802. 11 with Omni Antenna E silenced M A silenced F C silenced Y `` Interference management `` silenced S Data D silenced challenge for dense multihop networks A crucial G X B silenced ACK silenced D silenced K silenced 6

Managing Interference n Several approaches § Dividing network into different channels § Power control § Rate Control … New Approach … Exploiting antenna capabilities to improve the performance of wireless multihop networks 7

From Omni Antennas … E silenced M A C silenced F silenced D S silenced G X B silenced Y silenced D silenced K silenced 8

To Beamforming Antennas E M A C F S Y D G X B D K 9

To Beamforming Antennas E M A C F S Y D G X B D K 10

Today n Antenna Systems A quick look n New challenges with beamforming antennas 11

Antenna Systems n Signal Processing and Antenna Design research § Several existing antenna systems • Switched Beam Antennas • Steerable Antennas • Reconfigurable Antennas, etc. § Many becoming commercially available For example … 12

Electronically Steerable Antenna [ATR Japan] n Higher frequency, Smaller size, Lower cost § Capable of Omnidirectional mode and Directional mode 13

Switched and Array Antennas n On poletop or vehicles § Antennas bigger § No power constraint 14

Antenna Abstraction n 3 Possible antenna modes § § § Omnidirectional mode Single Beam mode Multi-Beam mode n Higher Layer protocols select § § Antenna Mode Direction of Beam 15

Antenna Beam n Energy radiated toward desired direction Main Lobe (High gain) A A Sidelobes (low gain) Pictorial Model 16

Directional Reception n Directional reception = Spatial filtering § Interference along straight line joining interferer and receiver C A Signal Interference C B Signal A Interference B D D No Collision at A 17

Will attaching such antennas at the radio layer yield most of the benefits ? Or Is there need for higher layer protocol support ? 18

We design a simple baseline MAC protocol (a directional version of 802. 11) We call this protocol DMAC and investigate its behavior through simulation 19

DMAC Example Y S D X n Remain omni while idle § Nodes cannot predict who will trasmit to it 20

DMAC Example RTS Y S D X n Assume S knows direction of D 21

DMAC Example RTS Y CTS S RTS DATA/ACK D X X silenced … but only toward direction of D 22

Intuitively Performance benefits appear obvious 23

Throughput (Kbps) However … Sending Rate (Kbps) 24

Clearly, attaching sophisticated antenna hardware is not sufficient Simulation traces revealed various new challenges Motivates higher layer protocol design 25

n Antenna Systems A quick look n New challenges with beamforming antennas 26

New Challenges [Mobicom 02] Self Interference with Directional MAC 27

Unutilized Range n Longer range causes interference downstream § Offsets benefits A Data B C D route § Network layer needs to utilize the long range § Or, MAC protocol needs to reduce transmit power 28

New Challenges II … New Hidden Terminal Problems with Directional MAC 29

New Hidden Terminal Problem n Due to gain asymmetry CTS A B RTS Data C n Node A may not receive CTS from C § i. e. , A might be out of DO-range from C 30

New Hidden Terminal Problem n Due to gain asymmetry CTS A Carrier Sense B RTS Data C n Node A later intends to transmit to node B § A cannot carrier-sense B’s transmission to C 31

New Hidden Terminal Problem n Due to gain asymmetry Collision A RTS B Data C n Node A may initiate RTS meant for B § A can interfere at C causing collision 32

New Challenges II … New Hidden Terminal Problems with Directional MAC 33

New Hidden Terminal Problem II Y S Data D X n While node pairs communicate § X misses D’s CTS to S No DNAV toward D 34

New Hidden Terminal Problem II Collision Y S Data D RTS X n While node pairs communicate § X misses D’s CTS to S No DNAV toward D § X may later initiate RTS toward D, causing collision 35

New Challenges III … Deafness with Directional MAC 36

Deafness n Node N initiates communication to S § S does not respond as S is beamformed toward D § N cannot classify cause of failure § Can be collision or deafness M S Data D S T R N 37

Channel Underutilized n Collision: N must attempt less often n Deafness: N should attempt more often § Misclassification incurs penalty (similar to TCP) M S Data D S T R N Deafness not a problem with omnidirectional antennas 38

Deafness and “Deadlock” n Directional sensing and backoff. . . § Causes S to always stay beamformed to D § X keeps retransmitting to S without success § Similarly Z to X a “deadlock” Z S RTS DATA D RTS X 39

New Challenges IV … MAC-Layer Capture The bottleneck to spatial reuse 40

Capture n Typically, idle nodes remain in omni mode § When signal arrives, nodes get engaged in receiving the packet § Received packet passed to MAC § If packet not meant for that node, it is dropped Wastage because the receiver could accomplish useful communication instead of receiving the unproductive packet 41

Capture Example A C C D D B Both B and D are omni when signal arrives from A A B B and D beamform to receive arriving signal 42

Take Away Message n Technological innovations in many areas § Several can help the problem you are trying to solve § Although gains may not come by plug-and-play § New advances need to be embraced with care. n Directional/Beamforming antennas, a case study § Gains seemed obvious from a high level § Much revisions to protocols/algorithms needed § Some of the systems starting to come out today n Above true for projects you will do in class 43

Rate Control in Wireless Networks Romit Roy Choudhury 44

Recall 802. 11 n RTS/CTS + Large CS Zone § Alleviates hidden terminals, but trades off spatial reuse E CTS A B C RTS F D 45

Recall Role of TDMA 46

Recall Beamforming Omni Communication Directional Communication e Silenced Node • f a b f a c b c d d No Simultaneous Communication Ok • Simultaneous Communication 47

Also Multi-Channel n Current networks utilize non-overlapping channels § Channels 1, 6, and 11 n Partially overlapping channels can also be used 48

Also Data Rates Benefits from exploiting channel conditions – Rate adaptation – Pack more transmissions in same time 49

What is Data Rate ? Number of bits that you transmit per unit time under a fixed energy budget Too many bits/s: Each bit has little energy -> Hi BER Too few bits/s: Less BER but lower throughput 50

802. 11 b – Transmission rates Highest energy per bit 1 Mbps 2 Mbps 5. 5 Mbps 11 Mbps Time Lowest energy per bit Optimal rate depends on SINR: i. e. , interference and current channel conditions 51

Some Basics n Friss’ Equation n Shannon’s Equation C = B * log 2(1 + SINR) n Bit-energy-to-noise ratio Eb / N 0 = SINR * (B/R) Leads to BER Varying with time and space How do we choose the rate of modulation 52

Static Rates SINR Time # Estimate a value of SINR # Then choose a corresponding rate that would transmit packets correctly (i. e. , E b / N 0 > thresh) most of the times # Failure in some cases of fading live with it 53

Adaptive Rate-Control SINR Time # Observe the current value of SINR # Believe that current value is indicator of near-future value # Choose corresponding rate of modulation # Observe next value # Control rate if channel conditions have changed 54

Rate and Range Rate = 10 B A C E D 55

Rate and Range Rate = 10 B A C E D Rate = 20 There is no free lunch talking slow to go far 56

Any other tradeoff ? Will carrier sense range vary with rate 57

Total interference Rate = 10 B A C E D Rate = 20 Carrier sensing estimates energy in the channel. Does not vary with transmission rate 58

Bigger Picture n Rate control has variety of implications § Any single MAC protocol solves part of the puzzle n Important to understand e 2 e implications § Does routing protocols get affected? § Does TCP get affected? § … 59

A Rate-Adaptive MAC Protocol for Multi-Hop Wireless Networks Gavin Holland HRL Labs Nitin Vaidya UIUC Paramvir Bahl Microsoft Research MOBICOM’ 01 Rome, Italy © 2001. Gavin Holland 60

Background n Current WLAN hardware supports multiple data rates § 802. 11 b – 1 to 11 Mbps § 802. 11 a – 6 to 54 Mbps n Data rate determined by the modulation scheme 61

Problem Modulation schemes have different error characteristics • 8 Mbps • BER • 1 Mbps • SNR (d. B) • But, SINR itself varies • With Space and Time 62

Impact Large-scale variation with distance (Path loss) • Distance (m) • Mean Throughput (Kbps) • SNR (d. B) • Path Loss • 8 Mbps • 1 Mbps • Distance (m) 63

Impact Small-scale variation with time (Fading) • SNR (d. B) • Rayleigh Fading • 2. 4 GHz • 2 m/s LOS • Time (ms) 64

Question • SNR (d. B) Which modulation scheme to choose? • 2. 4 GHz • 2 m/s LOS • Distance (m) • Time (ms) 65

Answer Rate Adaptation • Mean Throughput (Kbps) n Dynamically choose the best modulation scheme for the channel conditions • Desired • Result • Distance (m) 66

Design Issues n How frequently must rate adaptation occur? n Signal can vary rapidly depending on: § § carrier frequency node speed interference etc. n For conventional hardware at pedestrian speeds, rate adaptation is feasible on a per-packet basis • Coherence time of channel higher than transmission time 67

Adaptation At Which Layer ? n Cellular networks § Adaptation at the physical layer n Impractical for 802. 11 in WLANs Why? 68

Adaptation At Which Layer ? n Cellular networks § Adaptation at the physical layer Why? n Impractical for 802. 11 in WLANs RTS/CTS requires that the rate be known in advance • C • Sender • A • RTS: 10 • CTS: 8 • Receiver • 8 • B • D • 10 n For WLANs, rate adaptation best handled at MAC 69

Who should select the data rate? • A • B 70

Who should select the data rate? n Collision is at the receiver n Channel conditions are only known at the receiver § SS, interference, noise, BER, etc. • A • B n The receiver is best positioned to select 71

Previous Work n PRNet § Periodic broadcasts of link quality tables n Pursley and Wilkins § RTS/CTS feedback for power adaptation § ACK/NACK feedback for rate adaptation n Lucent Wave. LAN “Autorate Fallback” (ARF) § Uses lost ACKs as link quality indicator 72

Lucent Wave. LAN “Autorate Fallback” (ARF) • A • 2 Mbps • Effective Range • 1 Mbps • Effective Range • 2 Mbps • DATA • B n Sender decreases rate after § N consecutive ACKS are lost n Sender increases rate after § Y consecutive ACKS are received or § T secs have elapsed since last attempt 73

• SNR (d. B) Performance of ARF • Time (s) • Rate (Mbps) • Dropped Packets • Failed to Increase • Rate After Fade • Time (s) • Attempted to Increase • Rate During Fade – Slow to adapt to channel conditions – Choice of N, Y, T may not be best for all situations 74

RBAR Approach n Move the rate adaptation mechanism to the receiver § Better channel quality information = better rate selection n Utilize the RTS/CTS exchange to: § Provide the receiver with a signal to sample (RTS) § Carry feedback (data rate) to the sender (CTS) 75

Receiver-Based Autorate (RBAR) Protocol • 1 Mbps • 2 Mbps • C • RTS (2) • 1 Mbps • A • CTS (1) • DATA (1) • D • 2 Mbps • B • 1 Mbps n RTS carries sender’s estimate of best rate n CTS carries receiver’s selection of the best rate n Nodes that hear RTS/CTS calculate reservation n If rates differ, special subheader in DATA packet updates nodes that overheard RTS 76

• Rate (Mbps) • SNR (d. B) Performance of RBAR • Time (s) • ARF • Time (s) 77

Question to the class n There are two types of fading § Short term fading § Long term fading n Under which fading is RBAR better than ARF ? n Under which fading is RBAR comparable to ARF ? n Think of some case when RBAR may be worse than ARF 78

Rate Selection and Fairness 79

Motivation n Consider the situation below A B C 80

Motivation n What if A and B are both at 56 Mbps, and C is often at 2 Mbps? n Slowest node gets the most absolute time on channel? A B Timeshare A B C C Throughput Fairness vs Temporal Fairness 81

Rate Selection and Scheduling Goal n Exploit short-time-scale channel quality variations to increase throughput. Issue n Maintaining temporal fairness (time share) of each node. Challenge n Channel info available only upon transmission 82

Approaches n RBAR picks best rate for the transmission § Not necessarily best for network throughput n Idea: Wireless networks have diversity § Exploiting this diversity can offer benefits § Transmit more when channel quality great § Else, free the channel quickly 83

Opportunistic Rate Control Idea (OAR) n Basic Idea § If bad channel, transmit minimum number of packets § If good channel, transmit as much as possible D A C C A B Data Data 84

Why is OAR any better ? n 802. 11 alternates between transmitters A and C § Why is that bad D A C C Data A Data B Data Data Is this diagram correct ? 85

Why is OAR any better ? n Bad channel reduces SINR higher Tx time § Fewer packets can be delivered D A C C Data A Data B Data 86

These are only first cut ideas … Much advanced research done after these, covered in wireless courses However, important to understand these basics, even for mobile application design 87