Transcript Title

Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
Challenges and Impact of User-provided
Networking Technology
Shivendra S. Panwar
Polytechnic Institute of NYU
Brooklyn, NY
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
Some historical trends
• Computing: From (centralized) mainframes to
(distributed) personal computing, but… browsers
and cloud computing
• Network protocols: From (centralized) circuitswitching and SNA, to (end-to-end principle
based) TCP/IP, but …. consider the complexity of
Cisco’s IOS
• What about using user nodes or so-called “end
systems” as Layer X forwarding nodes?
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
The last mile will soon not be a
problem!
• Indeed, we may have an “all but the last
mile problem”
• The spread of FTTX, 4G networks, IEEE
802.11n, femtocells
• Still some work (and monetary investment),
but this is the trend
• This trend will further encourage userbased networking
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
Good and bad user-provided
networking
• ISP’s will encourage user-provided networking as
long as they do not lose control
• P2P is generally bad, except when controlled by
the ISP, then it is good
• WiFi is good if part of wireless carrier’s service;
otherwise viewed as a competitor to 3G/4G
• Femtocells are by definition good!
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
Some user-provided networking research
projects at NYU-Poly
• VoD service on fiber to the home, using ISP controlled
P2P technology; relieves traffic on backbone network
• FemtoHaul, a technique to use femtocells to relieve
pressure on the cellular backhaul
• Cellular networks have poor coverage and data rates at
their edges; cooperative two-hop relaying with distributed
space time coding can double or triple cell capacity
- Using other devices’ battery still an issue; security is
not an issue.
- Being considered in IEEE 802.16m, LTE Advanced,
IEEE 802.11 standards
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
P2P for VoD on FTTX
FTTX networks can provide VoD
services to customers. The use of
P2P technology on Set Top Boxes
substantially reduces the traffic on
the core network ring.
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
FemtoHaul
•
•
•
•
Current 3G/4G base station capacity: ~ 100Mbps
Four T1 lines typical for base station backhaul: ~ 6Mbps
Increasing number of data intensive applications
Adding backhaul infrastructure is very expensive (>15% of OPEX)
• Normal Scenario:
• All data comes from
the base station
backhaul
• FemtoHaul:
• Obtain the data from
an access point (WiFi,
Femtocell) through
another device
Femtocell
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
Spatial multiplexing gain for cooperative MIMO
• The number of antennas integrated on portable devices is
limited
• However, there might be several antennas at the base
station
• Randomized processing provides a robust scheme for
distributed cooperation
• Instead of diversity gain (Randomized Distributed Space
Time Coding), can we achieve spatial multiplexing gain?
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
Randomized Distributed Spatial Multiplexing
• Randomized Distributed Spatial Multiplexing (R-DSM) is based on the
Bell Lab Layered Space Time (BLAST) scheme
• Assuming each mobile station equipped with only one antenna and
base station has L antennas
• The channel capacity between the relays and the destinations scales
linearly with min(N,L), where N is the number of relays
• How does R-DSM work in PHY?
– Two-hop network: SISO transmission from source to relays first, followed
by relays transmitting together to the destination using R-DSM.
– Each relay independently generates a random coefficient and then
transmits a weighted sum of the signals for each antennas in BLAST
scheme
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
Using cooperative R-DSM on the second hop
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
Performance
• Our results demonstrate that R-DSM scheme mimics a real MIMO
system
• Provides additional capacity when and where you need it
Wireless Internet Center for Advanced Technology
NSF Industry/University Cooperative Research Center
Our Team
• Multiple Poly faculty and students working on different
aspects
– Erkip (PHY, MAC, Video, Security, Implementation)
– Knox (PHY, Implementation)
– Korakis (MAC, Video, Implementation)
– Memon (Security)
– Liu (MAC, Video, Implementation)
– Panwar (PHY, MAC, Video, Security, Implementation)
– Wang (PHY, MAC, Video, Implementation)
• Collaboration with industry
• Funded by NSF, WICAT, CATT, Poly Angel Funds
• See http://coop.poly.edu for more details