![Скачать презентацию Approximating Radio Maps Boaz Ben-Moshe SFU 1 28 7 05 Скачать презентацию Approximating Radio Maps Boaz Ben-Moshe SFU 1 28 7 05](https://present5.com/wp-content/plugins/kama-clic-counter/icons/ppt.jpg)
b41fb986af13ad78d65f57cd0e6ef496.ppt
- Количество слайдов: 30
Approximating Radio Maps Boaz Ben-Moshe (SFU) 1 28/7/05
Talk outline • • • 2 Definition, Motivation Known methods New Radar like Algorithm Experimental results Future work 28/7/05
Approximating Radio Maps Goal: Given an antenna (on a terrain T) compute it's approximated signal strength over T (it's Radio Map): 3 28/7/05
Definition & Motivation Clients: • any wireless device Base Station: • Type, patterns Motivation: • Wi. Max - 802. 16 4 28/7/05
Definition & Motivation Base Station: • Clients ant’ • Microwave links 5 28/7/05
Definition & Motivation Applications (of radio maps) • Locating Base Stations: – Guarding like. – Complex objective function. • Frequency Assignment: 6 28/7/05
Definition & Motivation Applications (of radio maps) • Locating Base Stations: • Frequency Assignment: – Conflict free frequency 7 28/7/05
Common method General Frame work: • Given a terrain and an antenna on it • Sample the area of 'interest' (SP) • For each p in SP: compute the signal strength • Compute a interpolation DS for SP Note: Propagation model: SKE, MKE 8 28/7/05
Approximating Radio-Maps General Frame work: Sampling Set (SP) 9 Interpolation DS 28/7/05
Approximating Radio-Maps Sampling Methods: • • Random, Grid Client oriented Terrain simplification Hybrid main problem: runtime! 10 28/7/05
Main Obstacle run-time: computing radio maps is often the runtime bottle-neck of wireless networks facility location algorithms. Existing radio maps methods are often too slow or not accurate enough. Solution: Radar-Like-Algorithm (RLA)… 11 28/7/05
Approximating Radio Maps Related work [BCK 04]: Visibility Approximating: Given a terrain T and a view point p compute the set of points on the surface of T that are visible from p. 12 28/7/05
Radar-like: Pizza slice Radar DS: {pizza-slice} – fast query 13 28/7/05
Radar-like: Pizza slice left & right cross-sections pizza slice. 14 28/7/05
Radar-like generic algorithm 1. Given Terrain (T), view point (vp), and fixed angle (a=A): 2. while(int i=0; i<360) { 3. S 1=cross-section(i); 4. S 2=cross-section(i+a); 5. if(close enough(S 1, S 2)) { 6. interpolate(S 1, S 2); 7. a = A; i = min(360, i + a); } 8. 15 else a = a/2; } 28/7/05
Approximating Radio-Maps Generalizing radar-visibility to RF propagation model: ● ● 16 Discrete visibility (boolean) continues Visibility a long a ray RF sampling 28/7/05
Approximating Radio-Maps 100*100 km elevationmap (of southern Israel) the brighter the higher. Antenna, 30 km radius. 17 28/7/05
Approximating Radio-Maps • Compute two consecutive crosssections. 18 28/7/05
Approximating Radio-Maps • Compute a sample set along the each cross-section: using 2 D terrain simplification methods. 19 28/7/05
Approximating Radio-Maps • Compute the signal strength along the sample set – using pipe-line method. 20 28/7/05
Approximating Radio-Maps • Compute the distance between the two signal-sections: • average / max / RMS distance 21 28/7/05
Approximating Radio-Maps Putting it all together: • Sensitive Radar algorithm • Sensitive 2 D Simplification • Robust distance norm Fine Tuning: • None grid sampling (2 D) • Parameters (terrain independent) 22 28/7/05
Radio-Maps: results Methods: • • 23 Random, Grid, TS F-Radar: fixed angle S-Radar: sensitive angle A-Radar: advance sampling 28/7/05
Approximating Radio-Maps Grid Random TS • 5000 samples per radio-map 24 F-Radar S-Radar 28/7/05
Approximating Radio-Maps Grid • 5000 samples per radio-map 25 S-Radar 28/7/05
Radio-Maps: results Run time for the same size sampling. • The radar is 3 -15 times faster than the regular sampling Radio Map methods. • More accurate. 26 28/7/05
Radio-Maps: results 27 28/7/05
Radio-Maps: results 28 28/7/05
Future-work • More testing • Advance interpolation methods • Interferences http: //www. cs. bgu. ac. il/~benmoshe/Radio. Maps/ 29 28/7/05
Fin Questions? 30 28/7/05
b41fb986af13ad78d65f57cd0e6ef496.ppt