Αναζήτηση

Πλοήγηση

• Κοινότητες
• Συγγραφείς
• Έτος
• Τίτλοι
• Θέματα

Ο Χώρος μου

• Είσοδος

Evaluation of Kalman Filtering for NetworkTime Keeping

Aggelos Bletsas

Πλήρης Εγγραφή

---

Περίληψη

Time information is critical for a variety ofapplications in distributed environments that facilitate pervasivecomputing and communication. This work describesand evaluates a novel Kalman filtering algorithm for endto-endtime synchronization between a client computer anda server of “true” time [e.g., a Global Positioning System(GPS) source] using messages transmitted over packetswitchednetworks, such as the internet. The messages exchangedhave the network time protocol (NTP) format, andthe algorithm evaluated, is performed only at the clientside. The Kalman filtering algorithm is compared to twoother techniques widely used, based on linear programmingand statistical averaging, and the experiments involve independentconsecutive measurements (Gaussian case) or measurementsexhibiting long-range dependence (self-similarcase). Performance is evaluated according to the estimationerror of frequency offset and time offset between clientand server clock, the standard deviation of the estimatesand the number of packets used for a specific estimation.The algorithms could exploit existing NTP infrastructure,and a specific example is presented.

Υπηρεσίες