2020
DOI: 10.1109/jlt.2020.2974319
|View full text |Cite
|
Sign up to set email alerts
|

Clock and Data Recovery-Free Data Communications Enabled by Multi-Core Fiber With Low Thermal Sensitivity of Skew

Abstract: Optical switching has the potential to scale the capacity of data center networks (DCN) with a simultaneously reduction in latency and power consumption. One of the main challenges of optically-switched DCNs is the need for fast clock and data recovery (CDR). Because the DCN traffic is dominated by small packets, the CDR locking time is required to be less than one nanosecond for achieving high network throughput. This need for sub-nanosecond CDR locking time has motivated research on optical clock synchroniza… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
6
0

Year Published

2021
2021
2022
2022

Publication Types

Select...
3
2
1

Relationship

4
2

Authors

Journals

citations
Cited by 10 publications
(6 citation statements)
references
References 19 publications
0
6
0
Order By: Relevance
“…This worst-case overhead can be lowered by reducing the optical fiber TCD, τ , and therefore reducing the required rate of clock phase updates, f φ , by the same factor, by using low TCD fibers such as hollow core fiber (20 times smaller TCD and therefore 20 times smaller required f φ and worstcase overhead than SMF-28 [9]). This overhead can also be potentially reduced by using homogeneous multicore fibre (MCF), which features a low thermal coefficient of skew of about 40 fs/km/ • C between different cores [21]. Further modeling of system performance and scalability in a clock phase cached optical switch using SMF-28 and HCF may be found in Chapters 6 and 7 respectively of [20].…”
Section: Scalability Of Clock Phase Cached Systemsmentioning
confidence: 99%
“…This worst-case overhead can be lowered by reducing the optical fiber TCD, τ , and therefore reducing the required rate of clock phase updates, f φ , by the same factor, by using low TCD fibers such as hollow core fiber (20 times smaller TCD and therefore 20 times smaller required f φ and worstcase overhead than SMF-28 [9]). This overhead can also be potentially reduced by using homogeneous multicore fibre (MCF), which features a low thermal coefficient of skew of about 40 fs/km/ • C between different cores [21]. Further modeling of system performance and scalability in a clock phase cached optical switch using SMF-28 and HCF may be found in Chapters 6 and 7 respectively of [20].…”
Section: Scalability Of Clock Phase Cached Systemsmentioning
confidence: 99%
“…The network overhead can be reduced by reducing the TCD, 𝜏, by using low TCD fibers such as hollow core fiber (20 times lower than SMF28) [8]. It can also be potentially reduced by using homogenous multicore fibers (MCF), which features a low thermal coefficient of skew of about 40 fs/km/°C [9]. A comprehensive analytical model of system performance and scalability in a clock phase cached optical switch may be found in [10].…”
Section: System Performance and Scalabilitymentioning
confidence: 99%
“…For a given transmitter to receiver pair, the total timing jitter arises from two sources: a) random temperature variation that changes the light propagation delay in optical fibers [3], and b) jitter of the clock sources that the transmitter and receiver are both synchronized to. The temperature induced timing jitter is relatively slow and, therefore, can be compensated by using clock phase caching [2] or significantly reduced by using low thermal sensitivity fibers [4,5]. The clock source induced jitter, however, is primarily high frequency and hard to compensate.…”
Section: Introductionmentioning
confidence: 99%