Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
1
Dynamic Spectrum Management
for Multi-Radio Environments
Fabian Quint, Andreas Klein, Hans Schotten
Fachtagung Mobilkommunikation Osnabrück, May 15, 2013
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
2
Contents
• Motivation
• Dynamic Spectrum Management Concepts
• DSA Algorithm
• Simulator Model
• Analysis of the Results
• Conclusion
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
3
Motivation
• Current wireless technologies are facing limited resources in the frequency domain.
• Today, spectrum is allocated statically due to regulatory constraints.
• Dynamic Spectrum Management (DSM) can increase system performance by managing spectrum in a flexible way.
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
4
DSM Concepts
• Different approaches how to achieve Dynamic Spectrum Access (DSM):
Dynamic Spectrum Management
Dynamic Exclusive Use Model Open Sharing Model Hierarchical Access Model
Spectrum Property Rights Dynamic Spectrum Allocation Spectrum Underlay Spectrum Overlay
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
5
DSM Concepts
• Allow selling and trading of spectrum.
• Freely choose technology.
Dynamic Spectrum Management
Dynamic Exclusive Use Model Open Sharing Model Hierarchical Access Model
Spectrum Property Rights Dynamic Spectrum Allocation Spectrum Underlay Spectrum Overlay
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
6
DSM Concepts
• Dynamic spectrum assignment to different services.
• Exploit spatial and temporal traffic variations.
Dynamic Spectrum Management
Dynamic Exclusive Use Model Open Sharing Model Hierarchical Access Model
Dynamic Spectrum Allocation Spectrum Underlay Spectrum Overlay Spectrum Property Rights
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
7
DSM Concepts
• Open sharing among peer users.
Dynamic Spectrum Management
Dynamic Exclusive Use Model Open Sharing Model Hierarchical Access Model
Dynamic Spectrum Allocation Spectrum Underlay Spectrum Overlay Spectrum Property Rights
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
8
DSM Concepts
• Hierarchical Access with primary and secondary users.
Dynamic Spectrum Management
Dynamic Exclusive Use Model Open Sharing Model Hierarchical Access Model
Dynamic Spectrum Allocation Spectrum Underlay Spectrum Overlay Spectrum Property Rights
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
9
DSM Concepts
• Constrained transmission power.
• Operate below primary users noise floor (UWB).
Dynamic Spectrum Management
Dynamic Exclusive Use Model Open Sharing Model
Dynamic Spectrum Allocation Spectrum Underlay Spectrum Overlay Spectrum Property Rights
Hierarchical Access Model
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
10
DSM Concepts
• Constraint on where and when to transmit.
• Exploit spatial and temporal white spaces.
Dynamic Spectrum Management
Dynamic Exclusive Use Model Open Sharing Model
Dynamic Spectrum Allocation Spectrum Overlay Spectrum Property Rights
Hierarchical Access Model
Spectrum Underlay
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
11
Dynamic Spectrum Allocation
• Contiguous DSA: Contiguous blocks of spectrum are allocated to different RANs.
• Fragmented DSA: Any arbitrary piece of spectrum can be allocated.
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
12
DSA Algorithm
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
13
Simulator Model
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
14
Simulator Model
• Spectrum Allocation based on predefined bandwidth chunks with size of 1.4 and 5 MHz.
• Calculation of interference events based on SEAMCAT methodology.
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
15
Simulator Model
• Quality of bandwidth chunks can be assessed by the introduced Spectrum Quality Indicator (SQI):
• Global and position dependent SQI values are considered to avoid interference.
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
16
Simulator Model
• Parameters Desired RAN LTE Interfering RAN WLAN
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
17
Simulator Model
• Exemplary temporal graph of DSA algorithm:
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
18
Analysis of the Results
• Assess benefit from operator point of view with success rate of bandwidth (BW) allocations:
• And by measuring the user satisfaction:
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
19
Analysis of the Results
• Success rate of chunk allocation for traffic model (HTTP/FTP) scenarios:
• Significant gains through interference avoidance approaches (based on SQI values).
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
20
Analysis of the Results
• User satisfaction for full buffer and 25% base load scenarios:
• Significant gain with shared band usage.
• Impact of SQI consideration is neglectable.
Reference case with disabled shared band
Lehrstuhl für
Funkkommunikation und Navigation
Chair for
Wireless Communication and Navigation
21.05.2013
21
Conclusion
• A basic temporal DSA algorithm was developed and introduced.
• The interference avoidance techniques based on SQI values showed respectable gain for traffic model scenarios.
• The usage of a shared band results in a significant gain for user experience.