Building modern networks for voice and data traffic with microwave radios

It is well recognized that modern point-to-point microwave radio systems are providing excellent building blocks for discussed application and are future proof for next 5 to 10 years.

It is important to look at the mentioned solutions from several viewpoints in order to assess suitability for particular task in hand:
Architecture;
Capacity;
Interfaces.

Architecture
Historically microwave systems have evolved from fully indoor systems into typical split mount systems of today with terminal consisting of IDU and ODU, this solution, no doubt, is here to stay.
Over last decade another type of solution has been offered to the market with mixed success, namely Full Outdoor systems. It is quite possible that Full Outdoor system share will increase in the future as migration to full packet/IP traffic will take place. The progress could be assisted by developments of technology, like ever higher capacity available at lower power consumption level and lowering equipment cost. In the future Full Outdoor system could most probably look like Power Over Ethernet powered compact all-in-one radio with a single Ethernet cable as the only connection to the user network, thus pretty much hiding the whole radio part of the network from the users if needed.

Capacity

While in the past, it was simple and straightforward regarding capacity of microwave systems:

Traditional well established PDH equipment providing capacity up to 40Mbps (2E1/4E1/8E1/16/17E1) at the low to medium capacity part of the network;
SDH systems providing STM-1 capacity in the backbone part of the network.

Today, the lines are blurred already with the arrival of new radio systems providing ‘anything in between’ type of capacity (often recognized as ‘SuperPDH’) and capable to fill much wider scope of tasks within any given network. We can definitely expect the process to continue and probably soon to observe the melting of different flavors of radio into one, capable to provide any capacity between 4 and 180Mbps (or even more) per radio.

Interfaces

As discussed, while many networks are still built by employing just traditional PDH and SDH interfaces, we see more and more widespread use of packet data in a form of Ethernet/TCP/IP, thus changing the requirement. Luckily enough there are more and more modern radio products on the market allowing user to choose the right blend of interfaces and capacity for particular task. To make it even more attractive any modern radio should be capable provide the flexibility to be reconfigured in both interface and capacity terms, as well as upgrade the system functionality in case of change of demand or during relocation.

Designing and building packet data / TCP/IP networks with microwave radios

At first glance, building packet network by employing P2P radios may seem simple and straightforward; unfortunately, as with any technology, many obstacles should be minded and sometimes overcome in order to do achieve success.
As majority of the equipment available today have Ethernet ports implemented as bridge or mapper, often with integrated L2 Ethernet switch, reasonable degree of care should be devoted to planning, testing and monitoring of the network.

It is of highest importance to know the technical parameters of the equipment in use, architecture of it and performance possible to achieve.
Knowledge of traffic patterns of particular network, applications employed is of equal importance as both together they hold the key to successful design of the network.

In particular as buffering is necessary to match the capacity of LAN side of the radio system to the capacity available from ODU to the far end terminal, it is very important to know and understand the behavior of bridge/mapper, like capacity of the buffer memory, method of packet buffering, latency introduced, etc. It is even more important to correctly measure the actual performance of Eth port of particular terminal, as this is the only way to provide benchmark for both comparing of different models of equipment and to measure the actual performance of the system during possible troubleshooting in the future. Nevertheless, it is next to impossible to have exact measurements of Ethernet performance over radio system by employing software tools/PC but this is the way to provide relative result for nonconclusive assessment of performance. Much better and stable results are to be achieved by employing leading hardware Eth testers.

In order to plan the network properly the traffic patterns should be foreseen and monitored later on, it is so much better not to overload the network than to fight traffic losses afterwards. It is paying off not to allow overly burst applications to be deployed or it should be planned appropriately in order to have the right degree of stability of packet network, hence traffic should be shaped as much as possible.

By knowing the applications well it is possible to do the right planning beforehand, will the real life latency be tolerated appropriately? This may influence the equipment choice as not all Ethernet bridges/mappers are equally regarding this metric.

To conclude – the task may not be trivial and easy one, but the effort spent will reward by network performance and stability up to expectations if the above-mentioned aspects of design and planning will be minded.

Click here to view a Case Study SAF did for Cable& Wireless Seychelles: “Reliable PDH – SAF CFM links for Cable & Wireless on Seychelles”