Transcript Shivkumar Kalyanaraman

Beyond (Basic) Cellular Networks: MultiHop/Meshed, Ad-Hoc, DTNs, White Space,
Wireless Cloud …
Shivkumar Kalyanaraman
shivkumar-k AT in DOT ibm DOT com
http://www.shivkumar.org
Google: “shivkumar ibm rpi”
Based in part upon slides of Bhaskaran Raman, Kameswari Chebrolu, Mihail L. Sichitiu, Hari Balakrishnan …
Shivkumar Kalyanaraman
IBM Research - India
1
Outline
Multi-hop (meshed) networks
 Dynamic Ad-Hoc Networking & Weak State Routing
 Delay/Disruption Tolerant Networks (DTNs)
 Opportunistic access / offload
 Software Radio & Wireless Network Cloud
 Cooperative MIMO & other cooperative techniques
 White Space Networking

Shivkumar Kalyanaraman
IBM Research - India
2
Recall: Wireless: A Short Technical summary
Rate
4G
802.11b
WLAN
3G
2G
1. Scarce bandwidth
2. Spectral Efficiency:
(10-100 MHz/operator) MHz -> Mbps (signal to noise ratio is key!)
Other Tradeoffs:
Rate vs. Coverage
Rate vs. Delay
Rate vs. Cost
Rate vs. Energy
2G Cellular
Mobility
3. Tradeoffs: Rate vs X
(no free lunch!)
Today
With femto cells & MIMO
antennas
Wireless networks are designed to maximize spectral efficiency, support mobility,
coverage, and Quality-of-Service under severe spectrum/bandwidth constraints
Shivkumar Kalyanaraman
IBM Research - India
3
Wireless IT convergence
Spectrum Scarcity & Solutions
2km
CDMA
Frequency Reuse
(reuse 3)
F2
F1
F3
F3
F2
F1
50
m
F3
F2
F1
F3
F2
BS
F1
F1
F3
F2
F1
BS
OFDM
F2
F1
F3
Frequency Reuse:
CDMA/OFDMA &
Scheduling:
• Static spectrum sharing;
• Spread spectrum in time-
TDM within each cell (GSM)
• Adjacent cells use different
frequency
• Widely used for voice
communication (GSM)
• Coverage
radius reduced
IBM Research
- India
and low spectral efficiency
Pico/Femto Cells &
(WiFi) Overlays/
Offload:
domain (CDMA) or frequency
domain (OFDMA)
• Smaller Cells, lower power
• Statistical multiplexing of
time-frequency
• Indoor access
characteristics: offload onto
Wifi or Femto cells
• Dynamic tradeoff of
power/interference: i/f limited
• More signal processing
(3G/3.5G): voice/data
• SON management
4
• Macro cell overlays
MIMO & CoMP:
• Multiple antennas: Spatial
degrees of freedom
• Collaborative MIMO (CoMP) to
manage inter-cell interference
• Intensive signal processing,
channel estimation, BS
coordination
Shivkumar Kalyanaraman
• Easier with BS Pools/Cloud
Spectrum Scarcity & Solutions (contd)
Spectrum Scarcity
Underlay
Occupied
spectrum
Opportunistic Access
Occupied
spectrum
Underlay
UWB
Spectrum resource
Spectrum resource
Multi-hop/Meshed
Networks / 60 GHz/FSO:
Space-Time-Frequency
Shifting of Workloads:
UWB (Underlay):
• Smaller hops: O(sqrt(N)) capacity
• Mobile mini-base station fleet
• Using ultra wide band
• Opportunistic access to smaller
cells (associations for ~10s)
spectrum without disturbing
the occupied users
increase
• Multi-route diversity; resilient
transport protocols; routing
advances
• Backhaul or shared spectrum
• Spectrum at higher reaches
(60GHz,
less regulated;
IBMFSO)
Research
- India
dense spatial reuse
• Multi-homed mobile devices
• Time-shifting of content delivery;
sophisticated traffic shaping at
peak times.
• Delay / disruption-tolerant/ad-hoc
networks
5
• Need to control the
transmitted power below
noise level to avoid
interference
• Generally for short range
transmission
White space/Cognitive
Radio/Opportunistic
Spectrum Access:
• Using white space of spectrum
opportunistically
• Dynamic spectrum scheduling
and management
• Need complex technologies for
detecting the white spectrum space
and management policies
Shivkumar Kalyanaraman
• Fit into cellular model TBD
Taxonomy
Wireless
Networking
Single
Hop
Infrastructure-based
(hub&spoke)
802.11
802.16
Cellular
Networks
Multi-hop
Infrastructure-based
(Hybrid)
Infrastructure-less
(ad-hoc)
802.11
Infrastructure-less
(MANET)
Bluetooth
Wireless Sensor
Networks
Wireless Mesh
Networks
Car-to-car
Networks
(VANETs)
Shivkumar Kalyanaraman
IBM Research - India
6
Meshed Networks
Shivkumar Kalyanaraman
IBM Research - India
7
Mesh vs. Ad-Hoc Networks
Wireless Mesh Networks
Ad-Hoc Networks


Multihop
Nodes are wireless,
possibly mobile






May rely on
infrastructure
Most traffic is user-touser
Multihop
Nodes are wireless, some
mobile, some fixed
It relies on infrastructure
Most traffic is user-togateway
Shivkumar Kalyanaraman
IBM Research - India
8
Mesh vs. Sensor Networks
Wireless Sensor Networks





Wireless Mesh Networks
Bandwidth limited (tens of kbps)
In most applications, fixed nodes
Energy efficiency is an issue
Resource constrained
Most traffic is user-to-gateway





Bandwidth is generous
(>1Mbps)
Some nodes mobile, some
fixed
Normally not energy limited
Resources are not an issue
Most traffic is user-to-gateway
Shivkumar Kalyanaraman
IBM Research - India
9
Lots of long distance links, adapted from WiFi
Shivkumar Kalyanaraman
IBM Research - India
10
Goals: variety of apps, QoS, scalable operation (100-200 nodes)
Shivkumar Kalyanaraman
IBM Research - India
11
Meshed Networks: Issues
The “link” abstraction may or may not hold depending
upon the type of links designed (directional vs omni)
 Using 802.11 MAC for multi-hop does not work
 Usually meshes are ~3 hops diameter
 TDMA / OFDMA extensions to handle 802.11 issues
 Per-hop losses: overcome via loss-tolerant TCP or
multi-path LT-TCP



Community meshes havent been too successful
 Niches in smart grids etc.
Shivkumar Kalyanaraman
IBM Research - India
12
Shivkumar Kalyanaraman
IBM Research - India
13
Links in A Backhaul Mesh
Shivkumar Kalyanaraman
IBM Research - India
14
Shivkumar Kalyanaraman
IBM Research - India
15
FRACTEL vs Roofnet
Shivkumar Kalyanaraman
IBM Research - India
16
Shivkumar Kalyanaraman
IBM Research - India
17
Shivkumar Kalyanaraman
IBM Research - India
18
MAC – Multichannel

Increases network capacity
1
2
Ch-1
1
Ch-1
2
3
2
3
4
Ch-1
User bandwidth = B/2
3
4
1
Ch-2
User bandwidth = B
Chain bandwidth = B
B = bandwidth of a channel
Shivkumar Kalyanaraman
IBM Research - India
19
MAC – Multichannel
Standard MAC – Multiple Radios




A node now can receive
while transmitting
Practical problems with
antennas separation (carrier
sense from nearby channel)
Optimal assignment – NP
complete problem
Solutions
 Centralized
 Distributed
GW
GW
GW
Shivkumar Kalyanaraman
IBM Research - India
20
MAC – Multichannel
Custom MAC – Multiple Radios



Nodes can use a control
channel to coordinate and
the rest to exchange data.
In some conditions can be
very efficient.
However the control
channel can be:
 an unacceptable
overhead;
 a bottleneck;
GW
GW
GW
Shivkumar Kalyanaraman
IBM Research - India
21
Routing for Meshes and MANETs


Routing consists of two fundamental steps
 Data plane: Forwarding packets to the next hop (from an
input to an output interface in a traditional wired network)
 Control plane: Determining how to forward packets
(building a routing table or specifying a route)
Forwarding packets is easy, but knowing where to forward
packets (especially efficiently) is hard
 Reach the destination
 Minimize the number of hops (path length)
 Minimize delay
 Minimize packet loss
 Minimize cost
Shivkumar Kalyanaraman
IBM Research - India
22
MANET vs. Traditional Routing






Every node is potentially a router in a MANET, while most
nodes in traditional wired networks do not route packets
Topologies are dynamic in MANETs due to mobile nodes, but
are relatively static in traditional networks
Channel properties, including capacity and error rates, mostly
static in traditional networks, but vary in MANETs
 Routing in MANETs could consider both Layer 3 and Layer
2 information: L2 can indicate connectivity and interference
Interference is an issue in MANETs, but not in traditional
networks
Channels can be asymmetric with some Layer 2 technologies
Traditional routing protocols for wired networks do not work
well in most MANETs: too dynamic
Shivkumar Kalyanaraman
IBM Research - India
24
Types of MANET Routing
MANET Routing Protocols
Proactive
Example:
OLSR
Reactive
Hybrid
Example:
AODV
Shivkumar Kalyanaraman
IBM Research - India
25
Common Features



MANET routing protocols must…
 Discover a path from source to destination
 Maintain that path (e.g., if an intermediate node moves and
breaks the path)
 Define mechanisms to exchange routing information
Reactive protocols
 Discover a path when a packet needs to be transmitted and
no known path exists
 Attempt to alter the path when a routing failure occurs
Proactive protocols
 Find paths, in advance, for all source-pair destinations
 Periodically exchange routing information to maintain paths
Shivkumar Kalyanaraman
IBM Research - India
26
Geographic Routing
Geographic Routing:
Compared to topology-based routing schemes, geographic routing schemes
forward packets by only using the position information of nodes in the vicinity
and the destination node.
Thus, topology change has less impact on the geographic routing than other
routing protocols.
Early geographic routing algorithms are a type of single-path greedy
routing schemes in which packet forwarding decision is made based on
the location information of current forwarding node, its neighbors,
and the destination node.
However, all greedy routing algorithms have a common problem, i.e., delivery is
not guaranteed even if a path exists between source and destination.
Shivkumar Kalyanaraman
IBM Research - India
27




Routing table entries: “state” =
indirections from persistent names
(ID) to locators
Due to dynamism, such indirections
break
Problematic in two dimensions
 Dynamism/mobility => frequent
update of state
 Dynamism + large scale => very
high overhead, hard to maintain
structure
Proposed solution:
 Probabilistic and more stable
state WEAK STATE
 Use of unstructured methods
Node Mobility
Challenges in Routing for Large-scale &
Dynamic Networks
Number of Nodes
Shivkumar Kalyanaraman
IBM Research - India
28
Weak State: A New Type of State

Strong State
 Deterministic
 Requires control traffic
to refresh
 Rapidly invalidated in
dynamic environments

Weak State
 Probabilistic hints
 Updated locally
 Exhibits
persistence
Shivkumar Kalyanaraman
IBM Research - India
29
Hard, Soft and Weak State
REMOVE
REFRESH
INSTALL
s
r
B
A
A with
A
B
probability 
Time elapsed since
Confidence in state
state
information ()
installed/refreshed
Weak
Hard
Soft State
State
Weak State is natural generalization of
Soft State
Shivkumar Kalyanaraman
IBM Research - India
30
An Instance of Weak State
SetofIDs
GeoRegion
{a,b,c,d,e,f}
Probabilistic
in terms of
membership


Probabilistic
in terms of
scope
The uncertainty in the mappings is captured by locally
weakening/decaying the state
Other realizations are possible
 Prophet, EDBF etc…
Shivkumar Kalyanaraman
IBM Research - India
31
Weak State Routing for MANETs
Random Directional Walk (RDW)
RDW used to announce location information (“put”) and
forward
Shivkumar Kalyanaraman
IBM Research
- India packets (“get”)

32
Dissemination/Proactive Phase: (put)

When a node receives a
location announcement, it
 creates a ID-to-location
mapping
 aggregates this mapping
with previously created
mappings if possible
C
B
A
Shivkumar Kalyanaraman
IBM Research - India
33
Forwarding Packets (get)
A
S
Confidence
0.71
B
WSR involves unstructured, flat, but
C
Confidence
scalable routing
; no flooding !
0.84
E
1.0
Confidence
1.0
D
Forwarding decision: similar to longest-prefix-match.
“strongest semantics match” to decide how to bias the random walk.
IBM Research - India
Details in ACM Mobicom 2007 paper
34
Shivkumar Kalyanaraman
Packet Delivery Ratio – Fixed Density
1
WSR with vmin=5, vmax =10
0.9
Packet Devivery Ratio
0.8
0.7
GLS-GPSR with vmin=5, vmax =10
WSR always achieve high OLSR with v =5, v =10
min
max
GLS ratio
works fine at low mobility
delivery
WSR with vmin=10, vmax =20
but fails to maintain
structure
at highGLS-GPSR
dynamismwith vmin=10, vmax=20
0.6
0.5
0.4
0.3
OLSR delivers only a small
fraction even at low
dynamism
0.2
0.1
0
400
IBM Research - India
500
600
700
800
Number of Nodes
35
900
1000
Shivkumar
Kalyanaraman
Control Packet Overhead
Total Overhead per Second (Number of Packets)
12000
WSR with vmin=5, vmax =10
GLS-GPSR with vmin=5, vmax =10
10000
OLSR with vmin=5, vmax =10
WSR with vmin=10, vmax =20
8000
OLSR overhead increases GLS-GPSR with vmin=10, vmax=20
exponentially
GLS works fine at low mobility but
requires superlinearly increasing
overhead to maintain structure at
high mobility
6000
4000
2000
0
400
IBM Research - India
500
600
700
800
Number of Nodes
36
900
1000
Shivkumar
Kalyanaraman
Transport: TCP Solutions


Focus on eliminating the
uncertainty between congestion
loss and all other reasons
Many approaches developed for
single-hop wireless systems



Snoop
I-TCP
M-TCP

End to end




SACK
Explicit error notification
Explicit congestion
notification (e.g. RED)
New solutions for multi-hop
 Loss-Tolerant TCP
 Multi-path LT-TCP (MPLOT)
Shivkumar Kalyanaraman
IBM Research - India
37
Loss-Tolerant TCP (LT-TCP) vs TCP-SACK
Maximum
Goodput
Missing
Goodput!
Shivkumar Kalyanaraman
IBM Research - India
38
Single path: limited capacity, delay, loss…
High Delay/Jitter
Low
Capacity
Lossy
Network paths usually have:
• low e2e capacity,
• high latencies and
• high/variable loss rates.
IBM Research - India
39
Time
Shivkumar Kalyanaraman
Idea: Aggregate Capacity, Use Route Diversity!
Low Perceived
Loss
High Perceived
Capacity
IBM Research - India
Scalable Performance
Boost Delay/Jitter
with ↑
Low Perceived
Paths
Shivkumar Kalyanaraman
40
Multi-path LT-TCP (ML-TCP): Structure
Socket
Buffer
Map pkts→paths intelligently
based upon Rank(pi, RTTi, wi)
Per-path congestion control
(like TCP)
Reliability @ aggregate, across paths
(FEC block = weighted sum of windows,
PFEC based upon weighted average loss rate)
Note: these ideas can be applied to other link-level multi-homing,
Network-level virtual paths, non-TCP transport protocols (including video-streaming)
Shivkumar Kalyanaraman
IBM Research - India
41
Delay/Disruption Tolerant Networks (DTNs)
Shivkumar Kalyanaraman
IBM Research - India
42
DTN Examples
Delay and Disruptions are first-class issues
End-to-end path may never exist at any instant in time,
but may emerge only over time
Shivkumar Kalyanaraman
IBM Research - India
43
Overview of Routing Issues for DTNs

State vs stateless Routing:
 Stateless: completely depend upon mobility, local storage at nodes &
replication/coding
 Eg: “Spray-and-wait” , “Spray-and-focus”
 Scaling challenges.
 Stateful: How to maintain useful state info despite disconnections.
 Weak state can again help (eg: WSR-D protocol), with “osmosis” of
state across connectivity clusters.

Simple situations such as “data mule” (getting e-mail from a village, or
synchronizing photos) involve 1-hop DTN routing etc.

Vehicular DTNs (eg: for an entire city) to provide useful complementary
communication services to cellular: not yet fully solved.
 Interesting small-scale testbeds: DieselNet (UMass)
Shivkumar Kalyanaraman
IBM Research - India
44
Opportunistic Offload via Small Cells
Opportunistic traffic offload
Eg:
Aruna Balasubramanian, Ratul Mahajan, Arun Venkataramani Augmenting Mobile 3G Using WiFi:
Measurement, Design, and Implementation In Proceedings of ACM MobiSys, San Francisco, USA, June 2010.
Vladimir Bychkovsky, Bret Hull, Allen K. Miu, Hari Balakrishnan, Samuel Madden, “A Measurement Study of Vehicular
Internet Access Using In Situ Wi-Fi Networks,” Proceedings of ACM Mobicom 2006, Los Angeles, 2006. (best paper)
Shivkumar Kalyanaraman
IBM Research - India
45
Wi-Fi Is Everywhere (in developed urban markets)
IBM Research - India
46
Shivkumar Kalyanaraman
Images from WiGLE.net and CarTel
Opportinistic Offload: The Opportunity

Today:
 Broadband connections are often idle
 65% of on-line households have Wi-Fi

What if …
 … home users open up their APs …
 … and share/sell the spare bandwidth?

Cellular complement for mobile users:
 Messaging (multimedia, e-mail, text)
 Location-aware services
 Mobile sensor networks (e.g. MIT project CarTel )
Shivkumar Kalyanaraman
IBM Research - India
47
Wi-Fi For Mobile Messaging



Wi-Fi cells are smaller than cellular cells
 Is density sufficient? Are connections too short?
Organically grown, unplanned deployments
 Uneven densities, AP churn, unpredictable
Back-of-the-envelope:
 55 km/hour: ~15 meters/s
 ~150 meter AP coverage [Akella’05]
 ~10 sec connectivity


What about connection overhead?
 scan, associate, get IP, etc.
Current stacks too slow
Shivkumar Kalyanaraman
IBM Research - India
48
Fraction of connections
CarTel Expt: Bytes Uploaded Per Connection
Non-trivial amount of data:
Median: 200 KBytes per
connection
Mean: 600 KBytes
Consistency check:
600 KBytes / 24 sec = 25 KBps
IBM Research - India
Shivkumar Kalyanaraman
Bytes received
49 on server (KBytes)
The Future of Software Radio:
Wireless Network Cloud
Parul Gupta, Smruti Sarangi, Shivkumar Kalyanaraman [IBM Research – India]
Zhen Bo Zhu, Lin Chen, Yong Hua Lin, Ling Shao [IBM Research – China]
Shivkumar Kalyanaraman
IBM Research - India
50
2G-3G
4G
Wireless
wireless
Network
network
over
architecture
Wireless Network Cloud
PSTN
Cloud
Access
of Network
Wireless Access Network
Core Network
+ Core Network
SMS/MMS
SMS/MMS
Mobile switch center
IMS WAP GW
controller
WiMAX
GSM
GSM
BS cluster
Radio network
TD-SCDMA
BS
Content Service
Edge gateway
Management
Server
BS
BS
Service Network
Web Service
Service support
Radio network
Gateway
node Edge
controller
Billing
gateway
Internet
BS
LTE
LTE
WiMAX
WiMAX
BS cluster
Shivkumar Kalyanaraman
IBM Research - India
51
2G/3G/4G Wireless over Wireless Network Cloud
PSTN
Cloud of Wireless Access Network + Core Network
Service Network
SMS/MMS
Service on Edge
IMS
WiMAX
GSM
GSM
TD-SCDMA
BS cluster
Content Service
Edge gateway
Management
Server
Billing
Web Service
Service support
node Edge
gateway
Internet
LTE
LTE
WiMAX
WiMAX
BS cluster
Shivkumar Kalyanaraman
IBM Research - India
52
Various Forms of Infrastructure Sharing in
Wireless Networks
Network Sharing
Base Station Sharing
Owner #1
Network
Owner #1
Retail
MSC
BSC
BSC
SDR
RRU
BTS
Owner #2
Network
Owner #2
Retail
Base Band Unit
BSC
BTS
Antenna Sharing
Owner #1
Network
Tower Sharing
BSC
Owner #1
Network
BTS
BSC
BTS
BSC
BTS
O
Owner #2
Network
BSC
Owner #2
Network
BTS
Shivkumar Kalyanaraman
IBM Research - India
53
Towards Active Sharing: Unbundling Base
Stations: RRU + BBU
 Distributed
base station
 RRU (Remote Radio Unit)
 BBU (Base Band Unit)
 Two
key standards enable distributed
base station development
 CPRI
 OBSAI
Traditional Integrated Macro BS
 Benefits
of distributed base station
 Reduce cost of facilitate
infrastructure
 Reduce power consumption
 Easy of installation
 Flexible deployment model
RRU
BBU
DistributedShivkumar
BS: RRU + BBU
Kalyanaraman
IBM Research - India
54
Multi-Technology Software Radio
Shivkumar Kalyanaraman
IBM Research - India
55
Unbundled SDR BS w/ Open Wireless Interfaces & IT
Platforms
...
RRU adaptor
RRU
GPS module
Base band
processing server
(PHY, MAC, C&M)
...
Base band
processing server
(PHY, MAC, C&M)
GE switch
RRU adaptor
PCIe/IB switch
General purpose servers
E1/T1/STM-1
Base band
processing
accelerators
SWR Base Station
CPRI/
OBSAI/
Ir
GE/E1/
T1/STM-1
IO & GPS
module
To RNC/ASN-GW/
AGW
Shivkumar Kalyanaraman
IBM Research - India
56
Distributed base station #2: distributed RRU +
centralized BBU pool
Scenario #2: central deployment

Benefits
Fit for super urban, urban
with high density of traffic
 Highly scalable
 Improve utilization by
resource sharing
 Reduce management cost

RRU
MSC
BBU
BBU
BBU
BBU
RRU
10KM
BSC
BBU Pool

Requirements & Challenges
to BBU
High density
 Resource sharing with
BBU pool
 Low power consumption


Case in China:

World largest TD-SCDMA BBU pool

Max support 72 RRUs

Power: 400W
city like
Bangalore
IBMA
Research
- India
Kalyanaraman
or Delhi could be served from <10 pooledShivkumar
sites.
57

Key barriers:
Wireless Network Cloud: Convergence of IT Platforms, SDR & RRH,
Cloud Computing Principles & Fiber-to-the-tower
Software Radio Technology/
IT & Cloud Computing
Techniques
Hybrid IT Systems
GSM RF header
GSM RF header
Server for BS
GSM
WCDMA
WCDMA RF header
Remote Radio Header Technology
WiMAX
GSM
Server for Access
GW
WCDMA
GSM
Resou
manag
LTE
Antenna + Remote
Radio Header
WiMAX
Server for Access
GW
Timing Network
over IP/Eth
WCDMA
WCDMA RF header
Fiber (> 10Km)
WiMAX RF header
BaseStation Pool
LTE RF header
End-to-End IP Infrastructure in 4G
Shivkumar Kalyanaraman
IBM Research - India
58
Wireless Network Cloud Potential: Distributed Interference Management.
Eg: Collaborative MIMO
Joint processing
Multi-cell environment with
frequency reuse factor 1
Optical fiber
Optical fiber
interference
I
I
Optical fiber
I
I
I
I
IBM Research - India
 Multiple points collaborate to mitigate ICI
or align interference for cancellation.
Shivkumar Kalyanaraman
59
Wireless Cloud/Base Station Pools Facilitate Co-operative
Techniques
Interference Avoidance
Joint scheduling and Load
Fractional Frequency Reuse (FFR),
Joint Power Control
Balancing
Base Station
Co-operation
Efficient Handovers
Interference Cancelation
To ensure quality for real-time flows like
VoIP, Video on Demand
Through Collaborative MIMO
Processing (CoMP) techniques
Shivkumar Kalyanaraman
IBM Research - India
60
TV White Space Spectrum Fact Sheet

Spectrum bands available: 54-60 MHz (TV channel 2), 76-88 MHz (TV channels 5 and 6), 174-216 MHz (TV
channels 7-13), 470-608 MHz (TV channels 14-36) and 614-698 MHz (TV channels 38-51). Channels below
21 restricted only to fixed devices.

The amount of spectrum available varies from market to market. In rural areas where fewer broadcasters are
operating, it can provide a substantial amount of capacity. But in dense urban areas, white spaces offer far less
capacity because more broadcasters are using the spectrum. For this reason, white-space spectrum could be
particularly valuable for providing broadband access in rural areas, where wired infrastructure doesn’t exist.

These bands was earlier used exclusively mainly by TV broadcast channels and wireless microphones

The FCC order in Nov 2008 required sensing for TV channels every 60 seconds and wireless mics and
auxiliary devices every 30 seconds. If detected, white space devices required to stop tx in 2 seconds

In the FCC Sep 2010 order, this need for spectrum sensing is removed for devices that lookup a geo-location
dbase of usage schedules to determine channel availability.

“Sensing only” devices may apply for certification which may be granted if they show high sensing accuracy

2 channels set aside for wireless mics. By default, mics will not be included in the geo-location dbase, but large
scale events organizers can petition to FCC for special inclusion

Power limits: 1W for fixed devices, 100 mW for portable (subject to interference and adjacent channel separation
conditions being met), 50mW for sensing only devices

Fixed TV bands devices can’t operate on locations where the ground level is more than 76 meters above the
average terrain level in the area.
Shivkumar Kalyanaraman
IBM Research - India
61
Opinions on FCC White Space Ruling
 FCC
ruling simplifies requirements on spectrum sensing for a class of devices (those
which can query TV databases)
 We
do not believe white-space spectrum will allow the emergence of a nation-wide
competitor, though niche players may emerge in specific markets.
 Even complementary strategies like Satellite + White Space, or WiMAX + White
space will be subject to capital investment constraints …
 … and unreliability challenges posed like meshed networks.
 Longer
transmission range + ad-hoc random access is a mixed blessing:
 … will lead to significant interference management problems for radio
access/MAC design (more complex than WiFi)
Shivkumar Kalyanaraman
IBM Research - India
62
Some Implications & Potentials

White space spectrum and “super-Wifi” hotspots that it enables are complementary
 It could significantly expand the “offload” strategy beyond indoor-Wifi in
homes, enterprises, airports, cafes etc.
 Sub-urban outdoor areas could also be served by Super-Wifi: will allow the
rapid adoption / proliferation of uplink video applications like Qik
 Delay-tolerant/opportunistic workloads could benefit if appropriate access
designs are created.
Technical trends likely to be catalyzed by White space spectrum:
 Disruption tolerance & opportunistic communication as a first class
paradigm in protocol stack design (beyond PHY/MAC layers)
 Multi-homing / multiple access (eg: 3G + WiFi + white space simultaneously
access) and complementarities leveraged
 Distributed MAC algorithms: balance needed between the two extremes of:
 scheduled MAC (for scalability/QoS) and
 random access (for spontaneous / ad-hoc access)
Shivkumar Kalyanaraman
IBM Research - India

63
White-Fi: a pre-FCC 2010 protocol
 Goal:
form network like Wi-Fi on top of white space.
 WhiteFi
 Wi-Fi
like system constructed on top of UHF white spaces
 Adaptively configures itself to operate in the most efficient
part of the available white spaces
 Techniques:
 1.
Spectrum Assignment
 2. AP discovery
 3. Handling disconnection
Shivkumar Kalyanaraman
IBM Research - India
64
64/22
Summary
Multi-hop (meshed) networks
 Dynamic Ad-Hoc Networking & Weak State Routing
 Delay/Disruption Tolerant Networks (DTNs)
 Opportunistic access / offload
 Software Radio & Wireless Network Cloud
 Cooperative MIMO & other cooperative techniques
 White Space Networking


Growth of video and data services over wireless
networks will drive future innovations in these areas.
Shivkumar Kalyanaraman
IBM Research - India
65