We use the concept of Containers (fixed size macro data unit) in a photonic burst switch for the core transport network. This Container is tailored to integrate the existing multiple protocol stacks as an intermediate transport mechanism between optical and link layers. The objective of our container based optical burst switch is to eliminate the multiple O/E/O conversions in the core network and create a collision-free, stateless, optical core by integrating optical transport and switching. This minimizes the necessity of costly Fiber Delay Lines (FDLs) in a full-mesh topology photonic network taking advantage of the fiber link delays. The concept uses the high capacity of optical transport technology and low cost of optical transportation to avoid the packet level management of the optical bandwidth. This is made possible by a centralized scheduler that periodically assigns optical burst slots to the edge nodes to carry the containers. The architecture is analyzed for its delay performance and performance dynamics of the container size is evaluated to determine the optimal container size. We also report the design and performance of our centralized scheduling algorithms.
[1]
N. Sahri,et al.
A highly integrated 32-SOA gates optoelectronic module suitable for IP multi-terabit optical packet routers
,
2001,
OFC 2001. Optical Fiber Communication Conference and Exhibit. Technical Digest Postconference Edition (IEEE Cat. 01CH37171).
[2]
George N. Rouskas,et al.
Packet scheduling in broadcast WDM networks with arbitrary transceiver tuning latencies
,
1997,
TNET.
[3]
Sajal K. Das,et al.
A photonic container switched transport network to support long-haul traffic at the core
,
2005,
2nd International Conference on Broadband Networks, 2005..
[4]
V. S. Tanaev,et al.
Scheduling Theory: Multi-Stage Systems
,
1994
.
[5]
Chase Cotton,et al.
Packet-level traffic measurements from the Sprint IP backbone
,
2003,
IEEE Netw..