This type of collision prevention is very simple, but can only be applied in situations where you can guarantee that no messages will ever need to be sent from one group to another. This is not the case in all systems, so an alternative method (random back-off) is also available.
Jump Tables
In some systems, there is an advantage to enforcing incompatibility between different groups of radios. This sounds a little strange, so let's imagine a system with only four radios. For simplicity, we'll arrange these radios into two groups of two.
Radios in our first group (1:1 and 1:2) will only need to talk to each other. Radio 1:1 will send all of its messages to 1:2 and radio 1:2 will send all of its messages to 1:1. The same is true for our second group (2:1 and 2:2).
Technically speaking, you do not have to change the radio's default jump table for this to happen. If radio 1:1 sends a packet addressed to 1:2, then this packet will be ignored by radios 2:1 and 2:2. However, by using different jump tables, we can decrease the amount of interference between groups, which increases reliability and throughput speeds.
So what is a jump table and why would you want to change it?
Radios such as those in the CDR XL radio family can transmit and receive on a number of different frequencies (channels). When transmitting data, the radio sends some data on one channel, then the transmitter jumps to a new channel, the radio sends some more data, etc. Regulatory agencies such as the FCC prohibit any radios from using the channels in sequential order, so the jump table defines a pseudo-random order that the channels should be used.
Back to our example above, let's pretend that radio 1:1 is sending a large amount of data to 1:2, while radio 2:1 is simultaneously sending a large amount of data to radio 2:2. We've already established that the packets are addressed and that radios other than the intended recipient will ignore any packets they hear, but the real problem is when the two radios transmit on the same frequency at the same time. Statistically speaking, this is bound to happen from time to time.
When multiple radios transmit on the same channel, the signals overlap, resulting in unintelligible gibberish. You can imagine that the radios are people at a party. When two of the people talk at the same time, it makes it difficult to hear what either one is saying. This is called a collision.
Collisions are not a big deal. If the radios are configured to use guaranteed-delivery mode (for example) then the transmitters will simply try the packets again later. The problem is that if radios 1:1 and 2:1 collide on a given frequency, wait for an acknowledgement, and then try again on the next frequency in the jump table, then they could collide again on that channel, and again and again until the data is discarded as undeliverable.
To prevent this from happening, we can configure the radios in group 1 to use one jump table (jump table 1, for example) and configure the radios in group 2 to use a second jump table (jump table 2, for example). The transmitters will still collide from time to time, but when the messages are retried, the transmitters will retry on different channels, and so they will not continue to collide.
Note: