ZL50114GAG2 ZARLINK [Zarlink Semiconductor Inc], ZL50114GAG2 Datasheet - Page 55

• Manufacturer Part Number: ZL50114GAG2
• Description: 128, 256 and 1024 Channel CESoP Processors
• Manufacturer: ZARLINK [Zarlink Semiconductor Inc]

5.5
In general, the next processing block for TDM packets is the Protocol Engine. This handles the data-plane
requirements of the main higher level protocols (layers 4 and 5) expected to be used in typical applications of the
ZL50110/11/14 family: UDP, RTP, L2TP, CESoPSN and SAToP. The Protocol Engine can add a header to the
datagram containing up to 24 bytes. This header is largely static information, and is programmed directly by the
CPU. It may contain a number of dynamic fields, including a length field, checksum, sequence number and a
timestamp. The location, and in some cases the length of these fields is also programmable, allowing the various
protocols to be placed at variable locations within the header.
5.6
Packets ready for transmission are queued to the switch fabric interface by the Queue Manager. Four classes of
service are provided, allowing some packet streams to be prioritized over others. On transmission, the Packet
Transmit block appends a programmable header, which has been set up in advance by the control processor.
Typically this contains the data-link and network layer headers (layers 2 and 3), such as Ethernet, IP (versions 4
and 6) and MPLS. Packet Reception
Incoming data traffic on the packet interface is received by the MACs. The well-formed packets are forwarded to a
packet classifier to determine the destination. When a packet is successfully classified the destination can be the
TDM interface, the LAN interface or the host interface. TDM traffic is then further classified to determine the context
it is intended for.
Each TDM interface context has an individual queue, and the TDM re-formatting process re-creates the TDM
streams from the incoming packet streams. This queue is used as a jitter buffer, to absorb variation in packet delay
across the network. The size of the jitter buffer can be programmed in units of TDM frames (i.e. steps of 125 µs).
There is also a queue to the host interface, allowing a traffic flow to the host CPU for processing. Again the host’s
DMA controller can be used to retrieve packet data and write it out into the CPU’s own memory.
5.7
At the receiving end of the packet network, the original TDM data must be re-constructed from the packets
received. This is known as re-formatting, and follows the reverse process from the Payload Assembler. The TDM
Formatter plays out the packets in the correct sequence, directing each octet to the selected timeslot on the output
TDM interface.
When lost or late packets are detected, the TDM Formatter plays out underrun data for the same number of TDM
frames as were included in the missing packet. Underrun data can either be the last value played out on that
timeslot, or a pre-programmed value (e.g. 0xFF). If the packet subsequently turns up it is discarded. In this way, the
end-to-end latency through the system is maintained at a constant value.
Protocol Engine
Packet Transmission
TDM Formatter
Zarlink Semiconductor Inc.
ZL50110/11/14
55
Data Sheet