Cooperative and Opportunistic Wireless Network Behnaam Aazhang University of Oulu
Cooperative and OpportunisticWireless Network
Behnaam AazhangUniversity of Oulu
Wireless Network
• Infrastructure• Ad hoc• Mesh network
Engineering Wireless Network
• Wireless links– Per link design
Wireless Network
• Collection of links– Interference– Resource allocation
• Orthogonal?
Scalability
• Number of users– Active flows
• Throughput• Delay• Energy
Wireless Network
• Improve scalability– Exploit network topology– Location
At the Physical Layer
• Architecture examples• Gains
– Achievable rates– Outage
R1
R2
D
S D
R
DataData S
Relay Channels
• Practice of multiuser information theory– Distributed coding– Finite feedback
• Relay selection• Hardware implementation
– Amplify and forward– Detect and forward
S D
R
Data
Wireless “Cooperative” Network
• Gains– Throughput– Power– Outage
• Overhead– Discovery– Access– Latency– Interference 2 8
1
3
5 7
D
S
6
4Data
Multiple Flows
• Distributed resource management– Access– Interference
• Priorities– Primary– Secondary
2 8
1
3
5 7
D2S1
6
4
S2D1
Data
Data
Research Agenda
• Distributed node discovery– Source initiated– State information
• Nodes and channels– Inspired by
• 802.11 RTS/CTS• Multihop route discovery (e.g., DSR)• Timer based relay selection
Research Agenda
• Cooperative path construction– Trellis– Graph theory
Example 1
• Two competing flows• Complete network state information
– Managing data with a trellisData
S2 D2
R
Data
D1S1
TrellisFlow 1
Flow 2
Meta node
Interference
Discovery
• The discovery for flow 1 is standard• R in the range of D1 receives D1’s CTS
and AckData
S2 D2
R
S1 D1
RTSRTS CTSCTS DataData AckAck
Discovery
• The discovery for flow 2– S2 sends RTS– R in the range of D1 will interfere with flow 1– D2 sends CTS– A noncooperativeflow 2 begins
Data
S2 D2
R
Data
S1 D1
Discovery
• If flow 2 is high priority– S2 sends RTS– R sends RTS– D2 sends CTS– R sends CTS torelay in half duplex
• Flow 1 will reducerate to half
Data
S2 D2
R
Data
S1 D1
Cooperative Network
• Developed protocols• Initial simulations• “In scale” experimental validation
– Throughput– Overhead– Complexity
Applications
• “Edge” solutions• Mesh networks• Sensor networks