Wireless interference limited high-throughput access technologies and applications, Ove Edfors, Lund University
1 like405 views
1. The WILATI+ project established channel measurement facilities and collected multi-link MIMO channel data to develop models and simulation tools. 2. System simulations using real and modeled channels evaluated MU-MIMO detection schemes and spectrum sensing policies. 3. Cooperative transmission schemes were proposed and accurate synchronization/ranging techniques were developed for OFDM.
![Achievement 4 WILATI
• Multi-link MIMO channel model - COST2100 channel model
Comparison of meas. and model Scatterer
example: delay spread MS cluster
1
0.8
0.6
CDF
0.4
Visibility
Measurement region
0.2 CM trial 1
CM trial 2
BS
CM trial 3
0
0 5 10 15 20 25 30
Delay spread [ns]
Practical and reliable tool for
radio link and network design
NORDITE Conference, June 14, 2011 7](https://image.slidesharecdn.com/wirelessinterference-limitedhigh-throughputaccesstechnologiesandapplicationswilati-110704064107-phpapp02/85/Wireless-interference-limited-high-throughput-access-technologies-and-applications-Ove-Edfors-Lund-University-7-320.jpg)
Wireless interference limited high-throughput access technologies and applications, Ove Edfors, Lund University
- 1. NORDITE2 Technology Programme WIreless interference-Limited high-throughput Access Technologies and applicatIons (WILATI+) Prof. Ove Edfors (Lund University, Sweden) NORDITE Conference, Oslo, Norway June 14, 2011
- 2. Interference Limited Scenarios WILATI Terminal cooperation among multiple cells BS BS MS BS Co-channel interference reduction among multiple cells Multiple access BS in a single cell (x,y,z) = ? Broadcasting Terminal in a single cell localization NORDITE Conference, June 14, 2011 2
- 3. Project Organization & Focuses WILATI Radio channels ? Rx Tx ? Rx Lund Aalto Network & Applications NTNU d? M1 (x,y,z) = (3.0,7.0,5.0) ! θ? M2 (x,y,z) = (1.0,4.0,2.0) ! Modem NORDITE Conference, June 14, 2011 3
- 4. Achievement 1 WILATI • Established a dual-link MIMO channel sounder Lund University Aalto University receiver: 30 antennas receiver: 30 antennas Multipath Multipath channel1 channel2 Access point 1 Access point 2 Lund University transmitter: 32 antennas Mobile station Enabling measurements of broadband real-time multi-link MIMO radio channels NORDITE Conference, June 14, 2011 4
- 5. Measurement, measurement, and measurement… WILATI Rx of Lund University Tx of Lund University Rx of Lund University Established extensive basis for analyzing multi-link MIMO radio systems Rx of Aalto University Rx of Lund University NORDITE Conference, June 14, 2011
- 6. Achievement 3 WILATI • Understanding of radio wave propagation mechanisms Identifying reasons of multipath propagation z 2 1 1 y 1 Reliable estimator of Increased understanding of x propagation paths capabilities and limits of radio systems NORDITE Conference, June 14, 2011 6
- 7. Achievement 4 WILATI • Multi-link MIMO channel model - COST2100 channel model Comparison of meas. and model Scatterer example: delay spread MS cluster 1 0.8 0.6 CDF 0.4 Visibility Measurement region 0.2 CM trial 1 CM trial 2 BS CM trial 3 0 0 5 10 15 20 25 30 Delay spread [ns] Practical and reliable tool for radio link and network design NORDITE Conference, June 14, 2011 7
- 8. Achievement 5 WILATI • Multi-user MIMO simulation chain C++/IT++ based simulation environment • Free open-source software/libraries • Simulations using both channel models and measured channel data • A range of iterative receivers implemented • Simulations run in parallel on a cluster BER versus corridor position BER vs. channel capacity SIMULATION CLUSTER (for two users and different APs) 216 nodes containing two 64-bit, quad core Intel Xeon E5520 (2.26 GHz), corres- ponding to a total of 1728 processors User 1 User 2 System performance evaluation tool for real indoor propagation environments NORDITE Conference, June 14, 2011 8
- 9. Achievement 6 WILATI • Comparative study of MIMO user detection schemes Which receiver algorithm combination achieves performance requiremnent with least complexity? Practical insights on the tradeoff between complexity and performance NORDITE Conference, June 14, 2011 9
- 10. Achievement 7 WILATI • Acquired insights of MU-MIMO performance in practice Decreasing inter-user channel correlation Minor gain of dirty paper as increasing # of BS antennas coding over linear precoding Understanding of how to deploy efficient multi-user MIMO networks in indoor scenarios NORDITE Conference, June 14, 2011 10
- 11. Achievement 8 WILATI • Developed sensing policies in cognitive networks Effects of adaptive machine Cognitive network structure learning algorithm Throughput relative to ideal policy With learning Without learning Understanding of how to design efficient cognitive radio networks NORDITE Conference, June 14, 2011 11
- 12. Achievement 9 WILATI • Comparative study of detecting user under cooperation User Distributed Alamouti detection cooperation Orthogonal separate detection Frame error rate Joint detection User 1 Non-orthogonal separate detection User 2 Destination Understanding of low complexity user cooperation NORDITE Conference, June 14, 2011 12
- 13. Achievement 10 WILATI • Identified the best precoder for MMSE iterative receivers Transmitter block diagram Random code Symbol error rate Golden code Receiver block diagram Dispersive Nucleo Algebraic code Theoretical limit Improved transmit signal arrangements for better radio system performance NORDITE Conference, June 14, 2011 13
- 14. Achievement 11 WILATI • Feasibility study of mobile terminal localization Ranging in OFDM systems using a Antenna array development in Matrix pencil method mobile phones for localization Multipaths Wave1 angle Error in resolving two Tx-Rx distance waves Mobile locations Wave2 angle New methods for high-accuracy ranging and localization NORDITE Conference, June 14, 2011 14
- 15. Summary WILATI • Possible impacts of our achievements on wireless industries 1. Design tools for multi-user MIMO networks – World-unique channel measurement facilities and channel database – COST2100 multi-link MIMO channel model – System simulations with real and modeled multi-link channels 2. Ideas for system and network improvement – Effective iterative receivers and precoders – Benefit of user cooperation – Performance of low complexity receivers 3. Insights for practical deployment of wireless networks – Complexity vs gain in precoding – Coverage of indoor base stations 4. Ideas for new systems and applications – Spectrum sensing policy in cognitive networks – Methods for high-accuracy ranging and localization NORDITE Conference, June 14, 2011 15
- 16. WILATI THANK YOU! NORDITE Conference, June 14, 2011 16
- 17. Achievements of WP1 in WILATI WILATI 1. Establishment of a dual-link MIMO channel sounder 1. Frame and synthesizer synchronization 2. Accurate calibration of the channel sounder and antenna array 2. Extensive collection of dual-link MIMO channel data and parameter estimates 1. Extended Kalman Filtering for parameter estimation 2. Provision of reference multi-link channel data for system simulations 3. Establishment of a multi-link channel model 1. Geometry-based model 2. Analytical model 4. Contributions to the COST2100 channel model 1. Multi-link extension and provision of model parameters 2. Validation of the model 9 IEEE journals, 21 int. conf., 3 PhD theses, 16 COST TDs NORDITE Conference, June 14, 2011 17
- 18. Achievements of WP2 in WILATI WILATI 1. Establishment of two MU-MIMO simulation chains (MATLAB and C++) capable of using 1. Raw/recreated channel measurements 2. Various channel models 2. Evaluation of MU-MIMO detection schemes in terms of performance and complexity/convergence using (a combination of) 1. Theoretical studies 2. Simulation based studies (with channel measurements) 3. Establishment of a spectrum sensing policy and evaluation methods 1. Proposing a measure for spatial diversity in spectrum sensing for different fusion rules 2. A machine learning based collaborative and energy-efficient spectrum sensing policy 3 journals papers (3 in preparation), 17 international conferences NORDITE Conference, June 14, 2011 18
- 19. Achievements of WP3 in WILATI WILATI 1. Proposal of various schemes for two-user cooperative MIMO communications in quasi-static fading channels 1. Decode-and-forward coop. and space-time coding at the Tx 2. MMSE-based iterative multiuser detection at the Rx 3. Extension of the schemes to arbitrary number of users 4. Analytical conditions on the space-time rotation matrix for achieving full diversity with MMSE-based iterative detection 2. Development of accurate time-synchronization and ranging techniques for OFDM systems 1. Time-reversal synchronization 2. Validation of the concepts by measured channels 3. Cooperative localization by multi-dimensional scaling 6 scientific journals, 7 international conferences Note: Further works are planned in 2011 in Norway NORDITE Conference, June 14, 2011 19