1. Data density for Light Communications
May 2015
doc.: IEEE /0496r1
November 2017
Data density for Light Communications
Date:
Author:
Name
Company
Address
Phone
Cheng Chen
University of Edinburgh
Prof. Harald Haas
pureLiFi Ltd.
Cheng Chen (University of Edinburgh)
Edward Au (Marvell Semiconductor)

2. November 2017
Abstract
This presentation discussed the data density capabilities relative to existing and emerging technologies.
Cheng Chen (University of Edinburgh)

3. 802.11ax is going to address dense scenario:
November 2017
802.11ax is going to address dense scenario:
Dynamic sensitivity control
Transmit power control
Multiple network allocation vectors
Cheng Chen (University of Edinburgh)

4. Pathloss exponent at 1-6 GHz in indoor environment: 2-3.5
November 2017
Pathloss exponent is a metric determines the decay rate of wireless signal power with propagation distance
Pathloss exponent at 1-6 GHz in indoor environment: 2-3.5
Sparse AP deployment: too many users per cell
Dense AP deployment: excessive interference causes low spatial reuse
Cheng Chen (University of Edinburgh)

5. Electrical signal Path loss exponent of LC with IM/DD: 4 - 8
November 2017
Electrical signal Path loss exponent of LC with IM/DD: 4 - 8
Can be deployed densely without causing excessive interference
Cheng Chen (University of Edinburgh)

6. Achievable single user data rate in a single cell deployment
November 2017
Achievable single user data rate in a single cell deployment
802.11ax single user data rate:
Coverage radius: 50m
160MHz modulation bandwidth
MU-MIMO with 8 spatial streams
OFDMA
LiFi single user data rate:
Coverage radius: 5.2m
180MHz, 625MHz modulation bandwidth
OFDMA
Blue filter
A user is randomly located in the coverage area of an access point. The user achieves various data rate due to different path loss and fading. This figure shows the probability of a user achieving a certain level of data rate.
Cheng Chen (University of Edinburgh)

7. Achievable cell data rate with multi-cell deployment
November 2017
Achievable cell data rate with multi-cell deployment
WiFi APs severely interfere with each other when separation between APs is less than 70m.
Each LC AP is not able to cover a large area when separation between APs is greater than 10m.
No severe interference between LC APs and each LC AP is able to cover a small when separation between APs is in the range of 6m to 10m.
Cheng Chen (University of Edinburgh)

8. Data density comparison
November 2017
Data density comparison
WiFi: Large coverage, but wider interference spread
LC: Smaller coverage, but very little interference spread
Little interference spread means capability of dense spatial reuse of transmission resources, which leads to significantly higher data density.
Cheng Chen (University of Edinburgh)

9. Transmission delay November 2017
WiFi: Massive number of stations (STAs) compete for time resource
LC: a few number of STAs compete and less delay
Cheng Chen (University of Edinburgh)

10. Simultaneous transmit and receive
November 2017
Simultaneous transmit and receive
WiFi:
Only possible when back off count down of AP and a STA end simultaneously
Requires synchronization and changes in the CSMA mechanism
LC:
Use different spectra for uplink (infrared) and downlink (visible light)
No synchronization between uplink and downlink is required.
Cheng Chen (University of Edinburgh)

11. November 2017
MU-MIMO
802.11ax:
Requires massive amount of overhead for channel estimation. Channel sounding interval is from 10 ms to more than 100 ms [1]
Synchronization between users are required when uplink MU-MIMO is required
LC:
Unidirectional source
Fixed grid of ‘beam’ can be arranged with appropriate lenses
‘Beam forming’ becomes ‘beam selection’
[1] G. Redieteab, L. Cariou, P. Christin and J. F. Hélard, "PHY+MAC channel sounding interval analysis for IEEE ac MU-MIMO," 2012 International Symposium on Wireless Communication Systems (ISWCS).
Cheng Chen (University of Edinburgh)