Physical layer modelling for future wireless networks. 6- wlan planning

Transcription

1 Physical layer modelling for future wireless networks 6- wlan planning

2 wlan planning Aim : planning wlans Phenomena to be Considered propagation radio link interference How? propagation coverage estimation Path-Loss, Radiation pattern, Epower,3D, radio link QoS estimation Collisions, PER (1st order properties) interference system constraint / QoS Switched interferences 2

3 wlan planning References D. Stamatelos and A. Ephremides, Spectral efficiency and optimal base placement for indoor wireless networks, IEEE J. Select. Areas Commun., H. Sherali, M. Pendyala, and T. Rappaport, Optimal location of transmitters for micro-cellular radio communication system design, IEEE J. Select. Areas Commun., M. D. Adickes, R. Billo, B. Norman, S. Banerjee, B. Nnaji, and J. Rajgopal, Optimization of indoor wireless communication network layout. IIE Transactions, F. Aguado-Agelet, A. M. Varela, L. Alvarez-Vazquez, J.-M. H. Rabanos, and A. Formella, Optimization methods for optimal transmitter locations in a mobile wireless system, IEEE Trans. Veh. Technol, Y. Lee, K. Kim, and Y. Choi, Optimization of AP placement and channel assignment in wireless LANs, in 27th IEEE Conference LCN 02, Tampa, Florida, USA, 2002 J. He et al., Globally Optimal Transmitter Placement for Indoor Wireless Communication Systems, IEEE Trans. Wireless Commun., 2004 E. Amaldi et al., Algorithms for WLAN Coverage Planning, LNCS, 2005 A. Bahri and S. Chamberland, On the wireless local area network design problem with performance guarantees, Computer Networks, O. Molina Lopez and I. Alonso Gonzalez, Automatic Planning Optimal Quality-Cost Wireless Networks: The Indoor Pareto Oriented ABSPAD Approach, Proc. IEEE PIMRC

4 wlan planning II-1) Coverage predictions Deterministic vs empirical methods Results / discussion II-2) QoS estimation A perf eval model, with PER Interference free II-3) Interference A standard FAP Interference impact on perfs II-4) Solving the problem A multi-objective framework Tabu 4

5 Usual approaches Coverage Empirical (COST 231) single path Free-space, multi-wall, Not site specific S. Seidel and T. Rappaport, 914 MHz path loss prediction models for indoor wireless communications in multifloored buildings. IEEE transactions on Antennas and Propagation,1992. S. Seidel and T. Rappaport, Site specific propagation prediction for wireless in-building personal communication system design. IEEE transactions on Vehicular Technology, vol. 43, no. 4, K.-W. Cheung, J. H.-M. Sau, and R. Murch, A new empirical model for indoor propagation prediction, IEEE trans on Vehicular Technology, M. Hassan-Ali and K. Pahlavan, A new statistical model for site-specific indoor radio propagation, IEEE trans. on Wireless Communications, 2003 Geometrical Optic (Ray Tracing) Fast if only few rays and reflections Trade-off J. McKnown and R. Hamilton, Ray tracing as design tool for radio networks, IEEE Network Mag., A. Neskovic, N. Neskovic, and G. Paunovic, Modern approaches in modeling of mobile radio systems propagation environment, IEEE com. surveys, C.-F. Yang, et al, A ray-tracing method for modeling indoor wave propagation IEEE trans AP, H. Suzuki and A. S. Mohan, Measurement and prediction of high spatial resolution indoor radio channel characteristic map, IEEE Trans on Vehic Tech, G. E. Athanasiadou and A. R. Nix, A novel 3D indoor ray-tracing propagation model: The path generator and evaluation of narrow-band and wide-band predictions, IEEE trans on Vehic Tech, G. Wolfle, R. Wahl, P. Wildbolz, and P. Wertz, Dominant path prediction model for indoor and urban scenarios, in 11th COST 273, F. A. Agelet and al., Efficient ray-tracing acceleration techniques for radio propagation modeling, IEEE trans on Vehic Tech T. Imai and T. Fujii, Fast algorithm for indoor microcell area prediction system using ray-tracing method, Electronics and Communications in Japan, Z. Chen, H. L. Bertoni, and A. Delis, Progressive and approximate techniques in ray-tracing-based radio wave propagatio prediction models, IEEE trans on Ant & Prop 2004.

6 Coverage Discrete approaches FDTD, TLM, Accurate but slow L. Talbi, FDTD characterization of the indoor propagation, Journal of electromagnetic waves and applications, J. Lee and A. K. Y. Lai, FDTD analysis of indoor radio propagation, in IEEE Antennas Propagation Society International Symposium, 1998, R. Sato and H. Shirai, Simplified analysis for indoor propagation of a wlan channel, in IEEE Topical Conference on Wireless Communication Technology, B. Chopard, P. Luthi, and J. Wagen, A lattice boltzmann method for wave propagation in urban microcells, in IEE Proc - Microwaves, Antennas and Propagation, 1997 P. O. Luthi, Lattice wave automata : from radiowave to fracture propagation, Ph.D. dissertation, University of Geneva, Switzerland, W. Hoefer, The transmission line matrix method - theory and applications, IEEE Transactions on Microwaves Theory Technique., J-M Gorce, K Jaffrès-Runser, G de la Roche. "A New Deterministic Approach for simulating Indoor Radio Wave Propagation" to appear IEEE Antennas and Propagation, march J-M Gorce, K Jaffrès-Runser, G de la Roche, The Adaptive Multi-Resolution Frequency-Domain ParFlow (MR-FDPF) Method for Indoor Radio Wave Propagation Simulation. Part I : theory and algorithms, technical report INRIA, RR- 5740, 57pp nov 2005 J-M Gorce, E Jullo, K Runser. «An adaptive multi-resolution algorithm for 2D simulations of indoor propagation». in Proc. 12th IEE Int conf on Ant & Prop, ICAP2003, IEE J-M Gorce, S. Ubéda «Propagation simulation with the ParFlow method : fast computation using a multi-resolution scheme», in Proc. IEEE VTC

7 Coverage Why a new method? Geometrical optic based approaches aim at starting from fast methods and improving the realism Our approach : starts from a realistic framework; tries to reach a low computational load 7

8 Coverage The main result A set of blocks, a set of candidate, for each a coverage computation : RI k k F m AP m 8

9 wlan planning II-1) Coverage predictions Deterministic vs empirical methods Results / discussion II-2) QoS estimation A perf eval model, with PER Interference free II-3) Interference A standard FAP Interference impact on perfs II-4) Solving the problem A multi-objective framework Tabu 9

10 Throughput vs received power Radio coverage -82dBm -87dBm QoS estimation WiFi : b : 4 thresholds g : 8 thresholds -91dBm 11 Mb/s 5,5 Mb/s 2Mb/s 1Mb/s 10

11 AP throuphput evaluation QoS estimation Throughput varies with: Number of users, Users' data rates A CSMA/CA: Same access probability of each user. Low bit rate frames spend more time on the channel. See first course 11Mb/s B [Bianchi, 2000] AP 5,5 Mb/s 2 Mb/s C 11

12 AP throughput evaluation QoS estimation Performance evaluation model [Bianqui, 2000][Lu & Valois, 2006]: Markovian model of the IEEE MAC backoff process [Bianchi, 00], Input: P {Retransmission} = P {Channel error} + P {Collision} The number of users transmitting at 1, 2, 5.5 and 11Mbits/s. Only uplink transmission under saturated traffic. n 5.5 n 11 AP 12

13 AP throughput evaluation QoS estimation Markov Chain resolution provides: The probability that one station transmits with success at data rate 1, 2, 5.5 or 11Mbps. The channel occupation efficiency for each type of nodes, The duration for which the channel is occupied. Final Output: The average throughput provided by the AP for each set of users working at 1, 2, 5.5 and 11Mbits/s J. Lu, K. Jaffrès-Runser, J-M. Gorce and F. Valois, "Indoor wlan Planning with a QoS constraint based on a Markovian Performance Evaluation Model", in IEEE WiMob Montréal, Québec, Canada, June

14 AP throughput evaluation QoS estimation Throughput model: Valid for a single AP, Interference free network. Predicting a realistic behaviour requires a complementary interference criterion Mitigating overall interference Reducing cells' overlapping, Improving efficiency of the Frequency Assignment Problem (FAP) solution, 14

15 wlan planning QoS estimation II-1) Coverage predictions Deterministic vs empirical methods Results / discussion II-2) QoS estimation A perf eval model, with PER Interference free II-3) Interference A standard FAP Interference impact on perfs II-4) Solving the problem A multi-objective framework Tabu 15

16 WiFi channels Interference 16

17 «FAP theory» Interference Interference level depend on the channel assignment for each cell (FAP) CCI PCI CCI

18 Interference Interference are managed in two phases Positioning phase: minimize the number of interference for each receiver A posteriori FAP Reduce overlapping between CC cells. G. de la Roche, R. Rebeyrotte, K. Jaffres-Runser and J-M. Gorce, "A QoS-based FAP Criterion for Indoor Wireless LAN Optimization", in proc. IEEE ICC 2006, Istanbul, Turkey, June

19 Interference Positioning phase: For each receiver node : Get the highest power Neglect h strongest interferers Minimizes h+1 interferer a good positioning should exhibit a low (h+1) interferer in each MR-node h allowed signals F B S Noise F (dbm) Interfering signals F h + 1 B l 19

20 Interference A posteriori FAP When the APs are selected, the best server is known : In each MR-node the SNR is given by: BS SNR mw = P / The FAP selects a frequency for each AP such as minimizing the SNR: And the SINR is: BS n SINRmW = P /( P + PI) P n. 20

21 wlan planning II-1) Coverage predictions Deterministic vs empirical methods Results / discussion II-2) QoS estimation A perf eval model, with PER Interference free II-3) Interference A standard FAP Interference impact on perfs II-4) Solving the problem A multi-objective framework Tabu 21

22 Planning criteria Optimization Three planning criteria: Coverage criterion f cov Interference criterion f I Throughput criterion f QoS Generic criterion formulation For each region B l P fmes Utility Value X l Penalization function X l fmes l µ l percentage of surface area S MIN S MAX X Quadratic criterion: 22

23 Coverage criterion Optimization Coverage provision Utility value X l = F BS l, the best server received power Thresholds for IEEE802.11b: S MIN = S 1 = -94 dbm, for a 1 Mbps data rate. S MAX = S 11 = -82 dbm, for a 11 Mbps data rate. fmes P = (S 11 - S 1 ) S 1 S 11 F BS 23

24 Interference criterion Optimization Reduce overlapping between cells. Utility value X l = F l h+1, Thresholds for IEEE802.11b: S MIN = S N = -96 dbm, the noise power. S MAX = S 11 = -82 dbm, for an 11 Mbps data rate. fmes h allowed signals F B S F (dbm) P = (S 11 - S N ) Interfering signals F h + 1 Noise S N S 11 F h+1 B l 24

25 Throughput criterion Optimization Throughput guaranty. Utility value X l = d l the user throughput Target throughput D* l Thresholds: S MIN = 0 kbps S MAX = D* l kbps P = D* fmes 0 D* d 25

26 Optimization Optimization example : variables : APs : number & position 258 candidates (from MR-nodes) criteria interference (h=2) throughput (ds = 256 kbits/s, 200 users) coverage F. Glover and M. Laguna, Tabu Search, Kluwer, 1997, p

27 Multi-objective approach Optimization Finding several solutions Trade-off surface : Pareto-front : x dominates y if : Surface of Trade-off Algorithm A. Jedidi, A. Caminada, and G. Finke, 2-Objective Optimization of Cells Overlap and Geometry with Evolutionary Algorithms, Proc. EvoWorkshops Y. Collette and P. Siarry, Multiobjective Optimization. Principles and Case Studies, Springer, F. Glover and M. Laguna, Tabu Search, Kluwer,

28 0 0 Results Interference Criterion f I Optimization Coverage Criterion f cov 0 28 QoS Criterion fqos

29 Results Optimization 29

30 Conclusions WLAN planning algorithm requires QoS-based constraints. Physical layer realistic properties can be introduced in the optimization criteria What has been considered? Propagation 2,5D / asymmetry / radiation pattern / emission power Radio link PER : any model can be used, Uncertainty can be introduced Interference Introduced as a planning constraint Introduced in the final FAP for evaluation purpose K Jaffrès-Runser,J-M Gorce, S Ubeda "Mono- and Multiobjective Heuristics for the Indoor Wireless LAN Planning Problem" in revision in Journal of Computers and Operations Research. Special Issue on Telecommunications Network Engineering. K Jaffrès-Runser, J-M Gorce, S Ubeda "QoS constrained wireless LAN optimization within a multiobjective framework", IEEE Wireless Communications, Special Issue. Dec 2006 K. Jaffrès-Runser, J-M. Gorce and S. Ubéda, "Multiobjective QoS-oriented planning for indoor wireless LANs", proc. of the IEEE VTC, Fall 2006, Montréal, Québec, Canada, Sept