Time sliced optical burst switching
read more
Citations
Optical Burst Switched Networks
TCP implementations and false time out detection in OBS networks
Contention avoidance and resolution schemes in bufferless all-optical packet-switched networks: a survey
Maximizing throughput for optical burst switching networks
Maximizing throughput for optical burst switching networks
References
Terabit burst switching
Transparent optical packet switching: network architecture and demonstrators in the KEOPS project
Photonic packet switches: architectures and experimental implementations
Routing, wavelength and time-slot assignment in time division multiplexed wavelength-routed optical WDM networks
A generalized framework for analyzing time-space switched optical networks
Related Papers (5)
Frequently Asked Questions (16)
Q2. What is the key determinant of performance for TSOBS?
Variable length bursts lead to higher discard probabilities, but the number of timeslots per frame remains the key determinant of performance.
Q3. How many timeslots would be needed to support a single channel rate of 100 Mb?
A 1 µs time slot would allow roughly 1100 bytes of user data to be sent in a single time slot, assuming a transmission rate of 10 Gb/s per wavelength, or 4400 bytes of user data, assuming a transmission rate of 40 Gb/s. With a 1 µs timeslot duration and 40 Gb/s transmission speeds, a system with 350 timeslots per frame would support an individual channel rate of about 100 Mb/s.
Q4. What is the key to implementing the search procedure?
The key to implementing the search procedure is a schedule that allows us to keep track of which delay lines are available for use at each point in time.
Q5. How many delay lines does the OTBS router need to use to switch the burst?
Using a set of short and long delay lines allows the number of switching operations needed for synchronization to be cut to 3, but requires more delay lines (14 in the example).
Q6. How many timeslots can provide excellent statistical multiplexing performance?
Their performance results show that a system with as few as 64 timeslots can provide excellent statistical multiplexing performance, even with blocking OTSIs with just four delay lines.
Q7. What is the control information needed to switch the data bursts?
The control information needed to switch the data bursts is sent in Burst Header Cells (BHC), which are carried on separate control wavelengths.
Q8. How many timeslots can be separated by a guard time?
To allow for timing uncertainties, timeslots must be separated by a guard time of at least 10 ns and possibly as large as 100 ns.
Q9. What is the effect of the limit on the number of switching operations?
For loads up to 85% the limit has a negligible effect on the number of switching operations, but for loads greater than 90% it produces a significant reduction.
Q10. What is the average number of bursts that pass through two delay lines?
For loads up to about 70% the average number remains below 2, meaning that the average burst passes through just one delay line and for loads up to about 90% the average number remains below 3, meaning that the average burst passes through two delay lines only.
Q11. What is the average number of bursts that require more than two switching operations?
For N = 64, less than 45% of the bursts require more than two switching operations, so almost 55% use at most two, meaning they only use a single delay line and less than 0.5% of the bursts require more than three switching operations, so almost 99.5% use at most three, meaning they only use two delay lines.
Q12. What is the maximum delay needed for a signal?
The maximum the authors can delay a signal using this configuration is (B−1)A+(A−1) and since the maximum delay needed is N − 1, this gives us the relation AB ≥ N .
Q13. What is the effect of limiting the number of bursts?
Fig. 10(a) shows that if the authors restrict the number of switching operations too much, the authors cause a large increase in the burst discard probability, but with a limit of 3, the burst discard probability is almost the same as when there is no limit.
Q14. What is the effect of reducing the number of delay lines?
This is significant, since with four delay lines, the total delay line length is reduced by a factor 4. Fig 11(b) shows the effect on the number of switching operations when the number of delay lines is limited.
Q15. How many timeslots per frame would allow a TSOBS router to be reduced?
Of course, a shorter timeslot duration or a smaller number of timeslots per frame would allow this maximum delay for a TSOBS router to be reduced proportionally.
Q16. What is the potential for optical burst switching?
This may allow optical burst switching to become cost-competitive with electronic packet switching, potentially a very significant development, since no previous optical packet switching architecture has shown any real promise of becoming cost-competitive with electronic alternatives.