Medium Access and Interference Cancellation: Protocol and Evaluation - PowerPoint PPT Presentation

Medium Access and Interference Cancellation: Protocol and Evaluation Abishek Sankararaman and François Baccelli 1

Introduction • Focus: Medium Access problem in Ad-hoc networks. � � � • Aim: Propose simple implementable protocols by incorporating observations and results from Information Theory. 2

Motivation • Some key features of emerging wireless networks Dense Decentralized Control D2D Communication Vehicular Communication (802.11p) 3

Motivation • Some key features of emerging wireless networks Dense Decentralized Control D2D Communication Vehicular Communication (802.11p) Managing Interference is a key challenge - primarily handled through Medium Access Control algorithms in ad-hoc networks. 4

Popular Medium Access Solution • CSMA (Carrier Sense Multiple Access) - 802.11 standards • ‘Interference as Noise’ (IAN) paradigm. 5

Popular Medium Access Solution • CSMA (Carrier Sense Multiple Access) - 802.11 standards • ‘Interference as Noise’ (IAN) paradigm. R No Interfering Transmitters T Rxr Guard Zone around a scheduled receiver Txr CSMA/CA Schematic • Simple Distributed Implementation (RTS/CTS) 6

Ad-hoc Network - Interference Channel 1 T 1 R 1 a a 2 user interference channel T 2 R 2 1 • Capacity and achievability is unknown in general. 7

Results from Information Theory 1 T 1 R 1 a a 2 user interference channel T 2 R 2 1 • Capacity and achievability is unknown in general. • , IAN is optimal. a → 0 • , SIC (Successive Interference Cancellation) a → ∞ decoding is optimal. (Receivers treat the transmitters as a MAC channel). 8

Successive Interference Cancellation T 1 T 2 Rxr Gaussian Codebook T 3 Received Powers P i, Rates R i ⇣ ⌘ P i ≥ R i , i ∈ { 1 , 2 , 3 } . C N 0 + P 3 j = i +1 P j where , C ( x ) = 1 2 log 2 (1 + x ) P i > P j ∀ i < j 9

Successive Interference Cancellation T 1 T 2 Rxr Gaussian Codebook T 3 Received Powers P i, Rates R i ⇣ ⌘ P i ≥ R i , i ∈ { 1 , 2 , 3 } . C N 0 + P 3 j = i +1 P j where , C ( x ) = 1 2 log 2 (1 + x ) P i > P j ∀ i < j • Separation of Powers needed to ensure decodability ! 10

SIC - Separation of Powers T 1 T 2 Rxr T 3 Received Powers P i, Symmetric Rate R P i j = i +1 P j ≥ Q P i > P j ∀ i < j N 0 + I + P k 11

SIC - Separation of Powers T 1 T 2 Rxr T 3 Received Powers P i, Symmetric Rate R P i j = i +1 P j ≥ Q P i > P j ∀ i < j N 0 + I + P k , • Separation of Powers needed to ensure decodability ! needsto be significantly larger than P i +1 P i 12

Main Idea of an Improved Protocol R 3 T 2 General capacity region is unknown T 3 R 2 T 4 T 5 R 4 T 1 R 5 R 1 13

Main Idea of an Improved Protocol R 3 T 2 General capacity region is unknown T 3 R 2 T 4 T 5 R 4 a 55 T 1 a 51 Any pair of links form a 2 user interference channel. R 5 a 15 a 11 R 1 14

Main Idea of an Improved Protocol R 3 T 2 General capacity region is unknown T 3 R 2 T 4 T 5 R 4 a 55 T 1 a 51 Any pair of links form a 2 user interference channel. R 5 a 15 a 11 R 1 If and , then a 51 >> a 11 a 15 >> a 55 • CSMA/CA will schedule at most one link. 15

Main Idea of an Improved Protocol R 3 T 2 General capacity region is unknown T 3 R 2 T 4 T 5 R 4 a 55 T 1 a 51 Any pair of links form a 2 user interference channel. R 5 a 15 a 11 R 1 If and , then a 51 >> a 11 a 15 >> a 55 • CSMA/CA will schedule at most one link. • However if the receivers can perform SIC, then both links could potentially be scheduled. 16

Main Idea of an Improved Protocol R 3 T 2 General capacity region is unknown T 3 R 2 T 4 T 5 R 4 a 55 T 1 a 51 Any pair of links form a 2 user interference channel. R 5 a 15 a 11 R 1 R 1 If and , then a 51 >> a 11 a 15 >> a 55 • CSMA/CA will schedule at most one link. • However if the receivers can perform SIC, then both links could potentially be scheduled. Need to define when a cross interference is ‘strong’. 17

CSMA 1-SIC Protocol R 2 R 2 T 2 T 2 R 3 R 3 T 3 R 1 Guard Zone T 3 R 1 Guard Zone T 4 around a receiver around a receiver R 4 T 1 T 1 Schematic of proposed Schematic of CSMA/CA CSMA 1-SIC protocol. 18

CSMA 1-SIC Protocol R 2 R 2 T 2 T 2 R 3 R 3 T 3 R 1 Guard Zone T 3 R 1 Guard Zone T 4 around a receiver around a receiver R 4 T 1 T 1 Schematic of proposed Schematic of CSMA/CA CSMA 1-SIC protocol. Separation of Received Powers - Donut Shaped Guard Zone. 19

CSMA 1-SIC Signaling Assume time-slotted system. Rx Tx 20

CSMA 1-SIC Signaling Each link (Tx) samples a Random Timer Value in say [0,1] Tx ‘ senses ’ channel till timer expires. t 3 t 1 t 6 Rx Tx t 0 t 5 t 2 t 4 21

CSMA 1-SIC Signaling Tx ‘ senses ’ channel till timer expires. t 3 t 1 t 6 Rx Tx t 0 t 5 Send RTS t 2 t 4 22

CSMA 1-SIC Signaling Rx ‘ senses ’ to hear a RTS. t 3 t 1 t 6 Rx Tx t 0 t 5 Send CTS t 2 t 4 23

CSMA 1-SIC Signaling Rx ‘ senses ’ to hear a RTS. t 3 t 1 Rx Tx t 0 t 5 Send CTS t 2 t 4 24

CSMA 1-SIC Signaling Tx broadcasts ‘ Established ’ to silence nearby receivers t 3 t 1 Rx Tx t 0 t 5 Send Established t 2 t 4 25

CSMA 1-SIC Signaling Tx broadcasts ‘ Established ’ to silence nearby receivers t 3 t 1 Rx Tx t 0 t 5 Send Established t 2 t 4 26

CSMA 1-SIC Signaling Tx transmits ‘ Established ’ signal t 3 t 1 Rx Tx t 0 t 5 Established t 2 t 4 27

CSMA 1-SIC Signaling Rx transmits ‘ Blocked ’ signal to silence all other strong interferers t 3 t 1 Rx Tx t 0 t 5 t 2 t 4 28

CSMA 1-SIC Signaling Summary • Randomized Protocol (Timers Chosen randomly). • 2 parameters to tune. � � � • Guarantees to any scheduled receiver that there will be at- most one ‘ strong ’ interfering transmitter. � 29

CSMA k-SIC Protocol One$Interfering$Transmi/er$Allowed$ No$Interfering$Transmi/er$ r 1$ r 2$ r 4$ r 3$ � • The separation of powers leads to 2k parameter protocol. • One can then develop a similar signaling algorithm. 30

CSMA /CA Versus CSMA 1-SIC Non-Monotonicity t 3 t 3 t 3 t 1 t 1 t 2 t 2 t 1 t 2 t 4 t 4 t 4 CSMA/CA CSMA 1-SIC 31

CSMA /CA Versus CSMA 1-SIC Non-Monotonicity t 3 t 3 t 3 t 1 t 1 t 2 t 2 t 1 t 2 t 4 t 4 t 4 CSMA/CA CSMA 1-SIC • Averaged over timer values however, CSMA 1-SIC schedules more links. 32

CSMA /CA Versus CSMA 1-SIC Non-Monotonicity t 3 t 3 t 3 t 1 t 1 t 2 t 2 t 1 t 2 t 4 t 4 t 4 CSMA/CA CSMA 1-SIC • Averaged over timer values, CSMA 1-SIC schedules more links. • This also means, that the interference levels are higher. 33

Performance Evaluation - Setup • A Stochastic Network Model to compare the gains in adopting the protocol. Dipole Network Model - Each Tx has an unique Rx. Tx form a PPP and the corresponding Rx is located at an uniform and independent angle away. 34

Performance Evaluation - Setup • A Stochastic Network Model to compare the gains in adopting the protocol. Dipole Network Model - Each Tx has an unique Rx. Tx form a PPP and the corresponding Rx is located at an uniform and independent angle away. No Power Control. All scheduled Tx transmit at unit power. Fading - Channel between any pair of devices is random and symmetric Path loss - l(r) = r -4 F xy l ( || x − y || ) 35

Performance Evaluation - Metrics • The metrics MAP - (Medium Access Probability (p a ) ) 36

Performance Evaluation - Metrics • The metrics MAP - (Medium Access Probability (p a ) ) SINR > Q Success Density - (Fraction of scheduled links successful (p s ) ) 37

Performance Evaluation - Metrics • The metrics MAP - (Medium Access Probability (p a ) ) SINR > Q Success Density - (Fraction of scheduled links successful (p s ) ) Throughput - (Fraction of links that get scheduled and are successful) 38

Performance Evaluation - Metrics • The metrics MAP - (Medium Access Probability (p a ) ) SINR > Q Success Density - (Fraction of scheduled links successful (p s ) ) Throughput - (Fraction of links that get scheduled and are successful) λ p s p a Throughput = 39

Performance Evaluation - MAP Rayleigh Fading No fading 0.8 0.8 CSMA IAN CSMA IAN CSMA 1 − SIC CSMA 1 − SIC 0.7 0.7 0.6 0.6 0.5 0.5 MAP MAP 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0 0.2 0.4 0.6 0 0.2 0.4 0.6 γ γ In large random networks, more links get scheduled on average . 40

Performance Evaluation - Success Probability Rayleigh Fading No Fading 0.9 1 CSMA IAN 0.95 CSMA 1 − SIC Success Probability Success Probability 0.8 0.9 0.85 0.7 0.8 0.75 0.6 0.7 CSMA IAN 0.65 CSMA 1 − SIC 0.5 0.2 0.4 0.6 0.8 0.2 0.4 0.6 0.8 γ γ CSMA 1-SIC has higher interference since it schedules aggressively ! 41

Performance Evaluation - Throughput Rayleigh Fading No Fading 0.2 0.28 0.18 0.26 0.16 0.24 Throughput Throughput 0.14 0.22 0.12 0.2 0.1 0.18 CSMA IAN CSMA IAN CSMA 1 − SIC 0.08 0.16 CSMA 1 − SIC 0.06 0 0.5 1 0 0.5 1 γ γ Nonetheless, CSMA 1-SIC has higher throughput ! 42