TDM and Packet Resource Information

The IP Network Interface Series 2 card converts between packet and circuit switched data that consumes a finite set of resources. The host needs to be aware of how the IP Network Interface Series 2 card interfaces with the Time Division Multiplexing (TDM) and packet resources in order to efficiently provision them.

The IP Network Interface Series 2 card has a one-to-one relationship between its VoIP channels and its TDM timeslots. Packet data is defined in terms of channels and circuit data is defined in terms of timeslots. If the IP Network Interface Series 2 card supports 1024 VoIP channels, it also needs 1024 TDM timeslots.

In the CSP, the IP Network Interface Series 2 card is like a
T-ONE/E-ONE line card, they both need TDM resources that are managed by the CSP Matrix Series 3 Card. The IP Network Interface Series 2 card, like the T-ONE/E-ONE line cards, uses spans and channels. The IP Network Interface Series 2 card VoIP channels are divided into multiples of 32-channel spans. Each VoIP module as well as the entire board divides evenly into 32-channel spans.

The IP Network Interface Series 2 card Pulse Coded Modulation (PCM) Companding law is configured for m-Law, the same as the
T-ONE line card. While in Normal Mode, the CSP Matrix Series 3 Card converts between m-Law and A-Law for each channel.

Spans and channels are assigned to the IP Network Interface Series 2 card, not to a specific VoIP endpoint (module). The VoIP channels are a pooled resource that can be associated with any span/channel. For example, span 1, channel 1 can be mapped to a channel on module 1; span 1, channel 2 can be mapped to a channel on module 2. This allows the host to allocate a span/channel according to system requirements.

For the case where all of the physical timeslots are not available, the host can decide how best to utilize the ones that are. The disadvantage is that the host may exhaust a module’s resource but still have available spans and channels. The other disadvantage is during a module failure. If a module or DSP on the module fails, active connections purge, but spans and channels cannot be taken out of service.

The IP Network Interface Series 2 card has three external 100Base-T (100 Mbps) Ethernet ports located on its Multi-Function Media I/O card. The intent is for the host to trunk these ports together to achieve an effective 300 Mbps pipe. If trunking is not used, only one of the three ports can be used at a time (any port can be used). Trunking (link aggregation) the ports also provides a transparent link failure protection. If one of the links fails, the traffic is automatically diverted to another link.

The IP Network Interface Series 2 card does not support Ethernet Link Redundancy. Instead, it automatically trunks all active Ethernet ports.

It is recommended that the IP Network Interface Series 2 card packet interface be on a different subnet from the one used to manage the CSP. Its main traffic is RTP, RTCP, and T.38 packets.

Important! Ethernet link failures do not cause RTP connections to be dropped. An alarm is generated to the host and the host can purge the connections. The same occurs for the removal of the
Multi-Function Media I/O card. An alarm is generated to the host and the host can purge the connections.

Bandwidth on the External Links is not monitored. If you have one link available and you are generating over 100 Mbps of RTP data, no alarms are generated and data is not throttled. The only indication is the degradation of voice quality.

You can choose to either have the Ethernet link negotiate the link rate and duplex mode or you can specify 100 Mbps and full duplex mode.

You control the mode with dip switch 4 on the IP Network Interface main board as follows:

Setting	Function
ON (default)	Ethernet Link Auto-Negotiate Mode
OFF	Ethernet Link Force 100 Mbps/Full Duplex Mode

3. Allow the card to reconfigure. During configuration, the CSP sends the host the Card Status Report (0x00A6) with the updated dip switch settings. Refer to this message in the API Reference.

Internet Protocol. Part of the TCP/IP family of protocols used to route messages and tracks the addresses of nodes.

TDM. Technique for digital transmission of a number of separate voice and data channels simultaneously over one communications medium, where each channel is allotted a time slot for data transmission.

See Time Division Multiplexing.

Voice over Internet Protocol. The delivery of voice information in digital form in discrete packets rather than in the traditional circuit-committed protocols of the public-switched telephone network (PSTN).

See Pulse Code Modulation.

PCM and companding standard used in Europe to compress and expand the amplitude range of a signal before and after transmission. See µ-Law.

Range of channels provided by T1 or E1 transmission service. A physical span is a group of channels taking up a single physical wire. A logical span is a group of channels taking up less than a single logical wire, or carried over more than one physical wire.

Digital Signal Processor

The protocol used in most local area PC networks. Typically, most Ethernet networks support data transmission speeds up to 10 Mbps, but 100 Mbps is also possible.

Real-Time Transport Protocol. Supports transport of real-time data like interactive voice and video over packet-switched networks.