Modeling of the Resilient Packet Ring
This
model has been created based on the specification laid out by the IEEE 802.17
Resilient Packet Ring Working Group. For more details on Resilient Packet
Rings, view the details below the model. This model has been created
to demonstrate the use of VisualSim for creating reliable and redundant bus
designs.
The VisualSim model has been created with the details of the Ring and segment specified. The rest of the system including the type of network interfaced, routing and packet rate distribution have been parametereized and kept loosely specified for the user to modify for a specific implementation. The model has been built as an hierarchical system.
The top-level model can be expanded to add more segments by simply disconnecting one of the existing segments and adding an indentical segment to the chain. Each segment module of the model contains the two rings, the details of the wire segment including packet fragmentation, queueing FIFOs and the test for destination match. The Ethernet or ATM node can be attached between the segment and the test area of the segment block. All of the segments have been instantiate from a single segment block.
All of the segments generate their own traffic using a single centralized traffic generator block. This traffic block is visible at the top-level. The traffic is generated based on a input text file and stores the packet information in a model database. The model database, Model_List, is a high-performance indexed database that can be used as a lookup table for the model. As the number of segments increase, the parameter can be annotated with the new segment name and the traffic for the additional blocks will be created. The DEMUX block, part of the Virtual Connection library, attached to the output of the Traffic block determines the original segment and sends the packet to start from that segment.
Each segment has an IN block that specifies the name of the segment and decodes the source. This Channel_Basic block handles fragmentation, number of separate channels, buffering and channel rate. IN addition, using the related Channel_Release block, blocking and non-blocking scenarios can be created by the placement of the Release block- immediately or after the transactions are completed. If the segment Channel requires the use of priority, then the Channel_Priority block can be used in place of the Channel_Basic.
A number of statistics generators have been pre-created. These include: Ring 1 and Ring 2 occupancy, and Ring utilization. Additional statistics can be created including packet latency, latency versus distance/priority and packet modified.
The model uses parameters extensively for the user to modify to evaluate different operating conditions. Example of modifiable parameters include Ring data-rate, traffic generation rate, amount of statistics to collect and link connectivity. The model uses local memory variables to store data on actions or results in each segment. These can be used during the simulation for storing intermediate or state information.
RPR is a technology developed in response to the rapid growth of data traffic in carrier networks. Currently, data networks in the metro must be transported through TDM circuits, like SONET, which are based on point-to-point, fixed bandwidth connections most appropriate for voice and constant bit rate services. An addressable, connectionless network optimized for bursty traffic best handles data communications and RPR is optimized for this traffic type running over optical networks in the metro and wide area.
RPR is optimized as a high availability protocol for transporting data, packet video and voice over ring topologies, while providing sub-50 millisecond protection switching. It is intended for carrier applications in the MAN and WAN, but can also provide reliable, high-speed campus and data center connectivity.
The VisualSim model has been created with the details of the Ring and segment specified. The rest of the system including the type of network interfaced, routing and packet rate distribution have been parametereized and kept loosely specified for the user to modify for a specific implementation. The model has been built as an hierarchical system.
The top-level model can be expanded to add more segments by simply disconnecting one of the existing segments and adding an indentical segment to the chain. Each segment module of the model contains the two rings, the details of the wire segment including packet fragmentation, queueing FIFOs and the test for destination match. The Ethernet or ATM node can be attached between the segment and the test area of the segment block. All of the segments have been instantiate from a single segment block.
All of the segments generate their own traffic using a single centralized traffic generator block. This traffic block is visible at the top-level. The traffic is generated based on a input text file and stores the packet information in a model database. The model database, Model_List, is a high-performance indexed database that can be used as a lookup table for the model. As the number of segments increase, the parameter can be annotated with the new segment name and the traffic for the additional blocks will be created. The DEMUX block, part of the Virtual Connection library, attached to the output of the Traffic block determines the original segment and sends the packet to start from that segment.
Each segment has an IN block that specifies the name of the segment and decodes the source. This Channel_Basic block handles fragmentation, number of separate channels, buffering and channel rate. IN addition, using the related Channel_Release block, blocking and non-blocking scenarios can be created by the placement of the Release block- immediately or after the transactions are completed. If the segment Channel requires the use of priority, then the Channel_Priority block can be used in place of the Channel_Basic.
A number of statistics generators have been pre-created. These include: Ring 1 and Ring 2 occupancy, and Ring utilization. Additional statistics can be created including packet latency, latency versus distance/priority and packet modified.
The model uses parameters extensively for the user to modify to evaluate different operating conditions. Example of modifiable parameters include Ring data-rate, traffic generation rate, amount of statistics to collect and link connectivity. The model uses local memory variables to store data on actions or results in each segment. These can be used during the simulation for storing intermediate or state information.
Model: Top-level of the RPR Design
Model- Details of each segment
About Resilient Packet Rings:
Resilient Packet Ring (RPR) technology is an emerging standard that enables the efficient transmission of data traffic over a SONET ring infrastructure, while continuing to leverage the network for TDM services. A resilient packet ring combines the low cost and simplicity of packet-based, connectionless networking with the reliability, bandwidth, and scalability of optical networks. The result is the best of both worlds -- a resilient, packet-oriented, ring-based solution that provides virtual mesh networking connectivity.RPR is a technology developed in response to the rapid growth of data traffic in carrier networks. Currently, data networks in the metro must be transported through TDM circuits, like SONET, which are based on point-to-point, fixed bandwidth connections most appropriate for voice and constant bit rate services. An addressable, connectionless network optimized for bursty traffic best handles data communications and RPR is optimized for this traffic type running over optical networks in the metro and wide area.
RPR is optimized as a high availability protocol for transporting data, packet video and voice over ring topologies, while providing sub-50 millisecond protection switching. It is intended for carrier applications in the MAN and WAN, but can also provide reliable, high-speed campus and data center connectivity.