PacketBand-TDM General FAQs - For additional help please see our Movie.
Please see here for technical FAQs

All TDM Versions:

1. Q.What is TDM over Ethernet or TDM over IP and why is it needed?
   
   
2. Q.What versions are available?
   
   
3. Q.What are PacketBand-TDM's particular strengths?
   
   
4. Q.How accurate is the clock recovery?
   
   
5. Q.Which PacketBands can inter-work?
   
   
6. Q.What is a Logical Link?
   
   
7. Q.What are the best Frames per Packet settings to use?
   
   
8. Q. Can I use PacketBand on the public Internet and what are the issues?
   
   
9. Q.Can I use a satellite link?
   
   
10. Q.Will the link carry any traffic transparently, even signaling protocols like DPNSS and QSIG?
    
    
11. Q.What is a Lost packet and a Late packet?
    
    
12. Q.Does PacketBand support VLAN Tagging?
    
    
13. Q.Does PacketBand support quality of service options?
    
    
14. Q.What are the extra Ethernet ports for?
    
    
15. Q.What is the latency added by PacketBand?
    
    
16. Q.Why do I need a PacketBand at both ends of the circuit?
    
    
17. Q.How much bandwidth will my PacketBand occupy?
    
    
18. Q.How do I know the best Jitter Buffer setting to use?
    
    
19. Q.How do I know where the clocking will come from for the system?
    
    
20. Q. There are firewalls on the sites where PacketBand will be used. Will it still work?
    
    
21. Q. I have an application with two separate clock sources. Can PacketBand cope with this?
    
    
22. Q. I already have the same clock source at both locations, so I don't need clock recovery. Is this OK?

PacketBand-V11/V35 Only

1. Q. I only want capacity a certain times, how can I do that?
   
   
2. Q. I want some PacketBand VX units, but I'm not sure whether I need DTE or DCE. What are the differences?
   
   
3. Q. What is "Control Signal Dialling "?
   
   
4. Q. What speeds do the X.21 and V.35 versions run?
   
   
5. Q. Can the V.35 and X.21 PacketBand communicate with
   E1 PacketBands ?
   
   
6. Q. What are the Model numbers I need to order?
   

PacketBand TDM

Q1. What is TDM over Ethernet or TDM over IP and why is it needed?

Packet networks have no concept of clocks or synchronisation. Data is transmitted at one location and arrives at the other at some unknown time. You will have seen when downloading files on the public Internet that the data rate varies and the arrival of the data goes faster/slower. This is due to loadings on the network and the number of switches the data has to pass through. The variation in packet arrival time is known as jitter or PDV (Packet Delay Variation). Jitter is far more pronounced and noticeable on the Internet compared to private networks, but the concept is the same.

Packet-based applications are designed to handle this varied and non-deterministic data arrival rate. They will also normally deal with any lost data by automatically requesting a retransmission, so the Internet looks to users as though it is error-free, which it generally is not.

All of the above works brilliantly, but, not for all applications. There are many devices around that are designed primarily to work with traditional leased lines which are point-to-point non-switched circuits - just like a dedicated wire with no jitter.

This equipment generally needs to transmit/receive data at a certain rate and may be synchronised with other devices so they all transmit or receive at the same rate, for example mobile phone transmitters.

In these cases, the equipment expects to have a steady, accurate clock source (not the time of day but a steady "heartbeat") with which to align its transmit and receive data.

Packet networks have no concept of clocks or clocking, PacketBand provides the necessary clock recover to enable TDM over IP or TDM over Ethernet services. At one end of the link PacketBand will be given a reliable clock source (external or internal) which is recovered by the PacketBand at the other side of the packet network.

PacketBand provides the TDM circuit across the packet network and ensures the attached TDM equipment receives the necessary steady clocks to enable a synchronised connection. The recovery and stability of these clocks, which is the whole basis and reason for using this technology, is a major strength of the PacketBand range.

In addition to the clocks, PacketBand also buffers the some incoming data before it transmits to the attached TDM device. This is necessary to ensure any “late” will have arrived in time for transmission on to the TDM device.

With PacketBand the TDM device sees what it believes to be a continuous steady clocked stream of traffic with no delays - just as if it was connected on a point-to-point wire or leased line. The products are highly-functioned and benefit from Patapsco's great support. Back to top.

Q2. What versions are available?

For E1/T1 G.703/G.704 clear/channelised products there is:
A single port unit
A 4-port unit available with 1, 2 or 4 ports enabled. Upgrading from 1 or 2 ports can be done in the field and remotely via a software key.
An 8-port chassis with dual hot-swappable PSUs
A chassis that supports 32 ports with the option of a single 16-port card.
For V.35 and X.21/V.11 the PacketBands have a single interface with a choice of the type of physical presentation. Back to top.

Q3. What are PacketBand-TDM's particular strengths?

The reason for installing a TDM over IP/Ethernet product is because you need clock synchronisation; there is no other reason to use it. So the clock recovery processes are critical to this technology.

PacketBand provides class-leading synchronisation with advanced algorithms that can be selected to match the packet network and to provide the best possible clock recovery, synchronisation and stability.

The products not only have great clock systems but also provide a very stable clock when they have to go to Hold-Over should there be a problem in the network, and when the network service resumes PacketBand is instantly ready and immediately gives service to the E1 T1 ports using the last good clock value.

You will also find in PacketBand a large number of possible configuration options. They do not have to be used but they have been incorporated over the years in response to user requests and different network types and applications.

But what if something should go wrong? PacketBand has a large number of easy-to-use diagnostic tools and statistics on network performance that prove invaluable "in the real world".

There is a wide range of products to match different applications and clock requirements, they interwork and all are easy to drive via the intuitive DbManager GUI.

You will not find a better performing unit or one with the capabilities of PacketBand in the same price bracket.

The final point we would like to cover is support. It is, of course, vital to get support when you need it. One of Patapsco's major strengths is our commitment to our in-house designed products and the technical expertise of our support team; not just on the products themselves but also on the user applications and related network setups. You will find Patapsco's support difficult to equal, indeed these days it is hard to find anyone that provides this level of commitment and expertise on the end of an email or telephone.

We are sure you will be delighted with the products and the after-sales support.
Back to top.

Q4. How accurate is the clock recovery?

Whilst PacketBand has a number of intelligent advanced Algorithms which can recover clocks extremely accurately, the final results do depend largely on the quality of the network.

PacketBand is capable of recovering clocks to within the G.823 Synchronisation Mask and down to 10ppb but this does require a high-quality network and/or the use of the Patapsco Multicast feature.

Unmanaged networks or those highly loaded without QoS (Quality of service) or those with high packet loss will not enjoy such an accurate clock, however the clock that PacketBand does recover is usually suitable for most applications.

We suggest you contact Patapsco to discuss this area and we can give you some indications once we know more about your application and network. Back to top.

Q5. Which PacketBands can inter-work?

Generally, any PacketBand TDM family device will work with any other. Some have feature not supported on others but all have the same core/base feature set and can inter-work.

Contact Patapsco if you need further information Back to top.

Q6. What is a Logical Link?

A Logical Link is a connection between two PacketBand units across a packet switched network. A single Logical Link can support up to 32 TDM channels (timeslots) between these two units. If a unit has one timeslot going to each of two PacketBands it will require 2 Logical Links. Each Logical Link can be configured with different settings (QoS and VLAN tags for example). Logical Links are sometimes referred to as Switched Virtual Circuits (SVCs). Back to top.

Q7. What are the best Frames per Packet settings to use?

The default value is 20, which is a good place to start unless bandwidth allocation is an issue. Generally speaking, the Frames per Packet value should be as low as possible where bandwidth will allow as this tends to deliver the best clock recovery. If the packet switched network is known to have an amount of packet loss, then a lower frames per packet value should be used so that a lost packet will not cause so much data to be lost. Please see the PacketBand Overhead Calculator to work out how much bandwidth a system will occupy. This can be found on the DbManager installation disk. Back to top.

Q8. Can I use PacketBand on the public Internet and what are the issues?

PacketBand can be used on the public Internet, provided there is enough network capacity available (remember an ADSL link will have a slow direction which is the limiting capacity of the link and there may also be considerable contention).
PacketBand has no control over packet loss across the link and packet loss means lost data.

The unpredictable nature of the public Internet and lack of prioritisation and management can make stable, synchronous communications systems difficult to run. Having said that, Patapsco does have some good experience in running services across the Internet.

Some actions can be taken to maximise the service and these include using the same ISP at each end of the circuit and using an ISP with QoS (Quality of Service) offerings.

National links and links within a region such as Europe will perform better than long distance inter-continental/regional links. Back to top.

Q9. Can I use a satellite link?

PacketBand has been proven to work on many satellite links with excellent results. The large amount of extra delay associated with satellite transmissions is no issue for PacketBand, provided the PDV is reasonably stable. Back to top.

Q10. Will the link carry any traffic transparently, even signaling protocols like DPNSS and QSIG?

PacketBand TDM will handle all protocol types transparently, simply passing the data to the end devices unchanged. This will work the same way in both G.703 (unstructured) and G.704 (structured or framed) modes, except that timeslot 0 is not passed on G.704 connections, but reconstituted locally.

Reconstituting timeslot 0 has some significant advantages when inter-connecting E1 devices on a network with packet loss. If running G.703 and packets are lost, the E1 interface of the attached devices may shut down after a number of lost timeslot 0 packets. When running G.704 mode this will not happen as the timeslot information is generated locally. Back to top.

Q11. What is a Lost packet and a Late packet?

PacketBand uses a buffer to collect and reorganize packets as they arrive from the packet switched network. The buffer can be configured to a defined size in order to catch the disorganized and transit delayed packets, without adding an unnecessary amount of extra latency to the system in processing time.

Early and Late packets are packets which arrive outside the range of the Jitter Buffer due to being very late or very early (these are rare) in crossing the packet switched network.

Lost packets are gaps in the packet sequence numbers, indicating that the packet was never received, and so was lost within the packet switched network. Early, Late and Lost packets are all unusable by PacketBand and the attached device and therefore are counted as Underruns. Back to top.

Q12. Does PacketBand support VLAN Tagging?

Yes. VLAN tags can be added on a per-Logical Link basis, and PacketBand’s onboard Ethernet switch can be configured to add and remove VLAN tags for passing traffic, as well as prioritise traffic based on its VLAN tag. Back to top.

Q13. Does PacketBand support quality of service options?

Yes. Type of Service settings for Delay, Throughput, Reliability, Cost and Precedence can be entered, or a DiffServ Codepoint value can be manually entered, which will be used for Logical Link traffic. Back to top.

Q14. What are the extra Ethernet ports for?

The extra Ethernet ports on PacketBand allow other Ethernet devices access to the packet switched network via PacketBand's onboard Ethernet switch. These ports can then be Rate Limited or the traffic prioritised so as not to interfere with the TDM packets.

Another useful feature of having spare Ethernet ports is the ability to configure a spare port as a 'sniffer' port. This port can have a PC running packet monitoring software running connected, and all activity on the port connected to the PSN is mirrored to the sniffer port. This can be very useful when debugging a system.
LACP (Link Aggregation Control Protocol) can use more than one Ethernet port to aggregate more than one network circuit together for extra capacity, load-sharing and resilience. Not all PacketBands support this option.

RSTP (Rapid Spanning Tree Protocol) both protects Ethernet networks from loops (or "short circuits") and can also be used to provide a "hot-standby" network link for resilience. Not all PacketBands support this feature. Back to top.

Q15. What is the latency added by PacketBand?

The main setting to take into account when considering the latency added by PacketBand is the length of the Jitter Buffer. The Jitter Buffer should be configured large enough that it will catch early and late packets, but not so large that an inordinate amount of latency is added to the system and not used. Please see the PacketBand manual for a full description of the Jitter Buffer settings. Some of the other Logical Link settings will affect the amount of extra latency imposed by PacketBand, and these are also explained in the manual. Back to top.

Q16. Why do I need a PacketBand at both ends of the circuit?

PacketBand is not able to 'convert' TDM data into a usable Ethernet packet. The PacketBand system encapsulates the data so that it can be passed across a packet switched network. When the packet arrives at the destination site it must be un encapsulated by another PacketBand unit before it can be used by a synchronous TDM or ISDN device. The packet is not suitable for use with another Ethernet device.

It is also vital for TDM traffic that it is delivered in the correct order, when expected and that the two ends of the TDM circuit are clock-locked and synchronised together. Back to top.

Q17. How much bandwidth will my PacketBand occupy?

The bandwidth allocation of a PacketBand system is based on many factors. As a general guide, setting a higher Frames per Packet value will generate less bandwidth overhead because there is less Ethernet/IP overhead for any given payload.

However, large or infrequent delivery of packets will impact negatively on the clock recovery processes.

Conversely, a low Frames per Packet value will require higher bandwidth but reap potentially better clocking.

It should be noted that when running at higher speeds (full T1/E1 for example) the change in Frames per Packet does not have a very large impact on overheads; it is at the low speeds that it is more noticeable.

The full user manual for PacketBand explains the settings which affect bandwidth, and the PacketBand Overhead Calculator can be used to work out how much bandwidth a system will occupy. Both of these resources can be found on the DbManager installation disk and by Registering. Back to top.

Q18. How do I know the best Jitter Buffer setting to use?

Jitter is the variation in delay time across a packet switched network. It is sometimes referred to as Packet Delay Variation (PDV). A high quality, managed packet switched network will often have a defined or maximum amount of jitter. Please consult a network administrator and ask about jitter.

If the amount of Jitter is not already predetermined, it is possible to monitor the Jitter on the network using PacketBand. PacketBand monitors its Jitter Buffer, and will report when packets fall outside the Jitter Buffer range. These are reported as Late and Early packets, and indicate that the Jitter Buffer length should be extended. Graphs of Jitter Buffer usage are also available on DbManager and alarms can be generated if the network starts to perform outside of a configured threshold.

Please see the PacketBand TDM manual for more information or contact Patapsco. Back to top.

Q19. How do I know where the clocking will come from for the system?

Correct clocking for a system is absolutely vital - without it you will experience lost data (at best). Most PacketBands must have a single clock source (unless running Plesiochronous clocks) and this can be from a variety of possible sources depending upon the unit. Please consult the full PacketBand manual for details on PacketBand's clocking requirements. Back to top.

Q20. There are firewalls on the sites where PacketBand will be used. Will it still work?

Yes, PacketBand can be used provided the correct port numbers are opened on the firewall and port forwarded to the PacketBand. PacketBand uses the following default port numbers:

The Logical Link port numbers can be changed. Please see the PacketBand manual for details.
Back to top.

Q21. I have an application with two separate clock sources. Can PacketBand cope with this?

PacketBand TDM has versions with support for a PDH (Plesiochronous Digital Hierarchy) environment. This Plesiochronous or Plesiosynchronous mode allows networks with separate clock architecture to interconnect via a PacketBand system. This method sends the transmit clock from each CPE across the PacketBand system, using each CPE's clock for the transmitted data
Plesiochronous clocking is covered in more detail in the appropriate model's brochure and manual. See the PacketBand-TDM Product Comparison sheet for a summary of features. Back to top.

Q22. I already have the same clock source at both locations, so I don't need clock recovery. Is this OK?

Yes. PacketBand TDM can be configured in Loop Timing mode. Loop Timing allows PacketBand to loop the Tx and Rx clocks from a CPE back on themselves. The clock is not passed across the IP network, so no clock recovery takes place. This option should be used when both CPEs have the same clock source, so no clock recovery is required and PacketBand provides the packetisation and depacketisation and uses its jitter buffers to "smooth" the traffic compensating for fast/slow packets. Back to top.

PacketBand-V11/V35 Only

Q1. I only want capacity a certain times, how can I do that?

A schedule can be set for each Logical Link, so that it is only enabled within defined periods . If the link is only needed when calls are made between two end user devices, it may be worth considering the PacketBand ISDN product range instead, which supports fully switched, dynamic call handling for BRI and PRI. Back to top.
Back to PacketBand-V11/V35 questions.

Q2. I want some PacketBand VX units, but I'm not sure whether I need DTE or DCE. What are the differences?

The main difference is DCE port normally gives or provides clocking and a DTE unit expects to be clocked. the DCE port would normally be the carrier-delivered Data Communications equipment and the DTE the users Data Terminal Equipment.

A DCE would always be connected to a DTE.
Most PacketBand-VX units sold are DCE versions where the clock is either generated internally or recovered across the packet network.
DTE versions can accept clock from an attached DCE device and this can then be recovered at the other end of the circuit on the partner PacketBand.
The other differences are the physical connector type and its gender so please ensure you order correctly. Contact us if you need assistance. Back to top.
Back to PacketBand-V11/V35 questions.

Q3. What is "Control Signal Dialling "?

PacketBand VX has the ability to monitor its serial interface for control signals and to then establish the TDM link across the packet network in response, thus instead of a permanent 24/7 circuit in place, it is "on-demand".
The control signal which will trigger this is 'C' for X.21 units and 'DTR' for V.35 units.

If you would like more information please contact Patapsco. Back to top.
Back to PacketBand-V11/V35 questions.

Q4. What speeds do the X.21 and V.35 versions run?

PacketBand VX can be configured to run at any speed from 64kbps to 2048kbps in steps of 64kbps. Back to top.
Back to PacketBand-V11/V35 questions.

Q5. Can the V.35 and X.21 PacketBand communicate with E1 PacketBands ?

Yes, PacketBand VX can fully interwork with E1/T1 PacketBand units.
When running at less than a full E1/T1 the E1/T1 PacketBand must use its channelised or framed option so each 64kbps step on the PacketBand VX represents one timeslot or channel on the PacketBand TDM unit.
If running a full 2.048Mbps the PacketBand E1 can run unstructured G.703 or if at 1.544Mbps a full unframed T1. Back to top.
Back to PacketBand-V11/V35 questions.

Q6. What are the Model numbers I need to order?

The flexibility and combinations of ports, power options and country-specific versions means there are a substantial number of variants. There are pricing examples available on the web site once registered. A description of the products and how the Part Numbers work is on the Ordering Information pdf. Back to top.
Back to PacketBand-V11/V35 questions.

Should you be unable to find the answer you seek, or are having difficulties, please contact us referencing any Support Agreement.