Cognitive Wireless Mesh Networks with Dynamic Spectrum Access

1
Cognitive Wireless Mesh Networks with Dynamic
Spectrum Access

• IEEE Journal on Selected Area in Communication,
  Jan. 2008 issue 1
• Kaushik R. Chowdhury
• Ian F. Akyildiz

2
Outline

• Introduction
• Related Work
• COMNET
• Architecture
• Spectrum sensing
• Channel switching
• Performance Evaluation
• Conclusion

3
Introduction

• Wireless Mesh Network (WMN)
• Mesh routers (MR)
• Mesh clients (MC)

4
Introduction

• Performance of WMN limited by
• Channel congestion
• WLAN, Bluetooth device, microwave ovens
• Traffic congestion
• Throughput capacity per node reduces
  significantly when node density increase
• The Capacity of Wireless Networks, IEEE Trans.
  On Information Theory, 2000

5
Introduction

• Urban areas area affected most by channel
  congestion
• - Capacity of Multi-Channel Wireless Networks
  Impact of Number of Channels and Interfaces, ACM
  MobiCom 2005

Dynamic Spectrum Access
6
Introduction

• If we have Intelligent and network aware spectrum
  selection
• Reduce channel congestion
• ? Also reduce node density in per channel reduce
  traffic congestion

7
Introduction

• Key challenge
• How to identifying which portions of spectrum are
  free to use
• Resolving contentions with licensed users
• Load balancing over the entire available spectrum

8
Related work

• Sensing techniques
• Single user detection
• Energy over time and threshold
• Primary user module
• Cooperative detection
• Topology
• TDMA/CDMA like protocol

9
COMNET

• Cognitive wireless Mesh NETwork (COMNET)
• Spectrum sensing and shared algorithm
• Easily integrated into existing mesh scenario
• A theoretical framework for identifying primary
  user frequencies through time domain sampling
• Formulate channel assignment as an optimization
  problem that is solved by MR

10
COMNET Architecture

• Objection
• Intelligent spectrum sensing capability in
  standard mesh scenario without advanced
  hardware/software techniques

11
COMNET Architecture

• Assumption
• Each MR and MC is equipped with a single IEEE
  802.11b based transceiver
• It is tunable anytime.
• No computing overhead in MR
• MR-MR multi-hop links are achieved through out of
  band communication
• No contention between MR-MR traffic with
  remaining nodes

12
COMNET Architecture

• Assumption
• MR has a fixed location that is known to all
  other MRs
• Primary transmitters as stationary
  radio/television towers
• No restriction on mobility of primary recivers
• MC can calculate its distance from primary
  stations

13
COMNET Spectrum Sensing

• Channel sensing based on time domain sampling
• Use of backoff interval for channel sensing

14
COMNET Spectrum Sensing

• Limitations of backoff interval for channel
  sensing
• Channel switch overhead

15
COMNET Spectrum Sensing

• Estimating channel occupancy by received signal
  power alone may not guarantee accurate results

16
COMNET Spectrum Sensing

• Centralized framework for time-domain sensing
• The leakage power of primary transmitters from
  other channel
• The received power is the sum of the individual
  transmit power

Channel for measurement is fixed
Leakage power for each transmitter is isolated
from aggregate received power
17
COMNET Spectrum Sensing

• The individual transmitter channels can be
  estimated
• Assume simple free space path loss model

I1,x spectral overlap factor between channels
of transmitter and receiver
18
COMNET Spectrum Sensing
Channel for measurement is fixed I1,1 I1,2
I1,3 I1,M
19
COMNET Spectrum Sensing
DX Y ? X D-1Y Assume antenna gains Gt Gr 1
20
COMNET Spectrum Sensing

• Decentralized approach gauss elimination

21
COMNET Channel Switching

• Formulate problem as an ILP.

22
COMNET Channel Switching
23
Performance Evaluation

• Effect of noise power increase

24
Performance Evaluation

• Effect of measurement set numbers

25
Performance Evaluation

• Relationship between min. required sensing nodes
  for given primary stations

26
Performance Evaluation

• Centralized and decentralized approach

27
Performance Evaluation

• Chain propagation delay

28
Performance Evaluation

• Channel selection algorithm
• topology

29
Performance Evaluation

• Before channel switching

30
Performance Evaluation

• After channel switching

31
Conclusion

• Combine wireless mesh network and cognitive radio
  approach
• Spectrum sensing method
• Decentralized approach?