Performance Analysis Of Bufferless 802.11 MAC Queues Ashwin Rao - - PowerPoint PPT Presentation

Introduction System Model Numerical Results Future Work & Conclusions References

Performance Analysis Of Bufferless 802.11 MAC Queues

Ashwin Rao 2006SIY7513 Supervisors

• Dr. Anirban Mahanti
• Dr. Arzad Kherani

Introduction System Model Numerical Results Future Work & Conclusions References

Outline

Introduction System Model Numerical Results Future Work & Conclusions References

Introduction System Model Numerical Results Future Work & Conclusions References

Introduction

• Safety applications [8]
• Main motivation for vehicular networks
• Use broadcast services of Medium Access Control(MAC) layer
• MAC of vehicular communication stack (1609.4 [1]) similar to

IEEE 802.11 [2]

• IEEE 802.11 MAC based on CSMA/CA

Introduction System Model Numerical Results Future Work & Conclusions References

IEEE 802.11 MAC: An Overview

Unicast Transmission

• Sample backoff value in the range (0, w)
• Decrement counter on channel idle
• Transmit packet when counter reaches 0
• Positive acknowledgement from destination =

⇒ no collision

• On collision repeat cycle for w = 2mCWmin

where, m ← backoff stage(no. of previous attempts) CWmin ← minimum contention window

Broadcast Transmission

• No acknowledgements
• No retransmission

Introduction System Model Numerical Results Future Work & Conclusions References

Motivation

Figure: Queuing Model Considering Security

Probability of collisions determine arrivals from lower layers Need to analyse P(Coll)

Introduction System Model Numerical Results Future Work & Conclusions References

Related Work

• Bianchi [3] - Markov chain for saturated MAC queues
• Malone et. al [7] - Extenstion for unsaturated queues (Unicast

Traffic)

• Kumar et. al [6] - generalisation of [3]
• Chen et. al [4] - extension of [3] for saturated MAC queues

and broadcast traffic

• Moreno et. al [9, 10] - simulations using probability of

reception as a performance metric

• Choi et. al [5] - 2 State markov chain for wireless channel to

study hidden node problem.

Introduction System Model Numerical Results Future Work & Conclusions References

Outline

Introduction System Model Numerical Results Future Work & Conclusions References

Introduction System Model Numerical Results Future Work & Conclusions References

Assumptions

Figure: Broadcast of fixed size packets in a single cell.

• n homogeneous nodes placed in a single cell
• The data exchanged is of a fixed size =

⇒ busy periods of a fixed number of slots (L).

• The nodes having a packet attempt to transmit with a

probability β in each slot.

• The arrival rate λ is very small resulting in at most one packet

at the MAC queue of each node.

• Cycle of busy and idle periods

Introduction System Model Numerical Results Future Work & Conclusions References

Markov Chain At Each Node

(a) Busy and Idle States of channel (b) Markov Chain at each node

Consider a node at the end of a busy period

• State 0 =

⇒ no packet

• State 1 =

⇒ packet to transmit, undergoing backoff process

• π =

⇒ Steady state probability of having a packet

• At each node if a packet
• arrives in a cycle =

⇒ 0 to 1 transition

• is transmitted in a cycle =

⇒ 1 to 0 transition

Introduction System Model Numerical Results Future Work & Conclusions References

Steady State Transition Probabilities

P10

• Transition from State 1 at beginning of cycle to State 0 at

end of cycle

• P10(k) =

⇒ Transition from 1 to 0 when k other nodes have a packet to transmit

• k nodes having packet =

⇒ j (0 ≤ j ≤ n − k − 1) other nodes can have arrival in next slot

• No transmission attempt (1 − β) by given node =

⇒ P10(k + j) in next slot

P10(k) =

n−k−1

X

j=0

n − k − 1 j ! λj (1 − λ)n−k−1−j (1 − β)k+1 P10(k + j) + β P10 =

n−1

X

k=0

n − 1 k ! πk (1 − π)n−1−k P10 (k)

Introduction System Model Numerical Results Future Work & Conclusions References

Steady State Transition Probabilities ... contd.

P01

• Transition from State 0 at beginning of cycle to State 1 at

end of cycle

P01(k) = “ 1 − (1 − β)k” “ 1 − (1 − λ)L+1” +

n−k−1

X

j=0

n − k − 1 j ! λj (1 − λ)(n−k−1−j) (1 − β)k ((1 − λ) P01 (k + j) + λP11 (k + j)) where P11(k) = 1 − P10(k) P01 =

n−1

X

k=0

n − 1 k ! πk (1 − π)n−1−k P01(k)

Introduction System Model Numerical Results Future Work & Conclusions References

Steady State Transition Probabilities ... contd.

Figure: Markov Chain at each node

πP10 = (1 − π)P01 ∴ π = P01 P01 + P10

Introduction System Model Numerical Results Future Work & Conclusions References

Outline

Introduction System Model Numerical Results Future Work & Conclusions References

Introduction System Model Numerical Results Future Work & Conclusions References

Simulation Parameters

Parameter Value Number of nodes (n) 3 to 250 Probability of transmit 1/16 and 1/8 in a given slot β (CW = 32, 16) Probability of packet arrival 1/1000, 1/2500 and 1/5000 in a given slot λ

100∗10−3 20∗10−6 = 5000

Number of slots required for 25, 50 and 100 a transmission L (busy slots)

250∗8 6∗106∗20∗10−6 ≈ 17

Table: Simulation Parameters

Simple discrete event simulator written that abstract the busy and idle cycles.

Introduction System Model Numerical Results Future Work & Conclusions References

π and fraction of packets undergoing collision when L = 25

0.1 0.2 0.3 0.4 25 50 75 100 125 150 175 200 π Number of Nodes π when L = 25

λ=1/1000, β=1/ 8 λ=1/1000, β=1/16 λ=1/2500, β=1/ 8 λ=1/2500, β=1/16 λ=1/5000, β=1/ 8 λ=1/5000, β=1/16

(a)

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 25 50 75 100 125 150 175 200 Fraction of packets undergoing collision Number of Nodes Fraction of packets undergoing collisions when L = 25

λ=1/1000, β=1/8 λ=1/1000, β=1/16 λ=1/2500, β=1/8 λ=1/2500, β=1/16 λ=1/5000, β=1/8 λ=1/5000, β=1/16

(b)

Collisions and π increase as

• Arrival rate lambda increases
• Probability of transmission attempt β decreases

Adapting λ has a greater impact on P(coll) than adapting n and β.

Introduction System Model Numerical Results Future Work & Conclusions References

π and fraction of packets undergoing collision when λ = 1/5000

0.1 0.2 0.3 0.4 25 50 75 100 125 150 175 200 π Number of Nodes π when λ=1/5000

L = 100, β = 1/ 8 L = 100, β = 1/16 L = 50, β = 1/ 8 L = 50, β = 1/16 L = 25, β = 1/ 8 L = 25, β = 1/16

(c)

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 25 50 75 100 125 150 175 200 Fraction of packets undergoing collision Number of Nodes Fraction of packets undergoing collisions when λ = 1/5000

L = 100, β=1/8 L = 100, β=1/16 L = 50, β=1/8 L = 50, β=1/16 L = 25, β=1/8 L = 25, β=1/16

(d)

Collisions and π increase as

• L increases

Increasing L is detrimental to system performance

Introduction System Model Numerical Results Future Work & Conclusions References

Outline

Introduction System Model Numerical Results Future Work & Conclusions References

Introduction System Model Numerical Results Future Work & Conclusions References

Future Work

• Equation of π and P(coll) in terms of λ, β, n, L
• Sensitivity Analysis of π and P(coll)
• Using equation of P(coll) in network of queues

Introduction System Model Numerical Results Future Work & Conclusions References

Conclusions

• Adapting λ has a greater impact on P(coll) than adapting n

and β.

• Increasing L is detrimental to system performance

Introduction System Model Numerical Results Future Work & Conclusions References

Outline

Introduction System Model Numerical Results Future Work & Conclusions References

Introduction System Model Numerical Results Future Work & Conclusions References

References I

IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) - Multi-channel Operation, IEEE Std 1609.4-2006. 2006. IEEE Standard for Information technology-Telecommunications and information exchange between systems-Local and metropolitan area networks-Specific requirements - Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications. June 12 2007, pp. C1–1184. Bianchi, G. Performance analysis of the IEEE 802.11 distributed coordination function. Selected Areas in Communications, IEEE Journal on 18, 3 (2000), 535–547.

Introduction System Model Numerical Results Future Work & Conclusions References

References II

Chen, X., Refai, H. H., and Ma, X. Saturation performance of IEEE 802.11 broadcast scheme in ad hoc wireless lans. In Vehicular Technology Conference, 2007. VTC-2007 Fall. 2007 IEEE 66th (Sept. 30 2007-Oct. 3 2007), pp. 1897–1901. Choi, J.-M., So, J., and Ko, Y.-B. Numerical analysis of IEEE 802.11 broadcast scheme in multihop wireless ad hoc networks. Information Networking (2005), 1–10. Kumar, A., Altman, E., Miorandi, D., and Goyal, M. New insights from a fixed-point analysis of single cell ieee 802.11 wlans. IEEE/ACM Trans. Netw. 15, 3 (2007), 588–601.

Introduction System Model Numerical Results Future Work & Conclusions References

References III

Malone, D., Duffy, K., and Leith, D. Modeling the 802.11 distributed coordination function in nonsaturated heterogeneous conditions. IEEE/ACM Trans. Netw. 15, 1 (2007), 159–172. The CAMP Vehicle Safety Communications Consortium. Vehicle safety communications project task 3 final report identify: Intelligent vehicle safety applications enabled by dsrc.

• Tech. Rep. 809859, National Highway Traffic Safety

Administration, U. S. Department of Transportation (USDOT), Mar. 2005.

Introduction System Model Numerical Results Future Work & Conclusions References

References IV

Torrent-Moreno, M., Corroy, S., Schmidt-Eisenlohr, F., and Hartenstein, H. IEEE 802.11-based one-hop broadcast communications: Understanding transmission success and failure under different radio propagation environments. In MSWiM ’06: Proceedings of the 9th ACM international symposium on Modeling analysis and simulation of wireless and mobile systems (New York, NY, USA, 2006), ACM Press,

• pp. 68–77.

Torrent-Moreno, M., Jiang, D., and Hartenstein, H. Broadcast reception rates and effects of priority access in 802.11-based vehicular ad-hoc networks. 10–18.