Half Duplex Vs Full Duplex - What is it, and why does it matter?

Duplex or duplexing although sounds like a toilet tissue brand and the word for applying it, is the networking term for the ability for two locations or devices to talk with one another. An example of this is walkie-talkies. One talks, the other listens until you say over. In popular culture geeks/nerds, whatever the colloquial term now is, have always been given grief for insisting on using 'over'. Guess which camp I'm in? That's right, how on earth are you supposed to know when it's your turn to speak?! You can shout all you want but if I havent stopped transmitting you aint getting through Jack!

So there are many types of time-divisional duplex (TDD) systems which are common place and are found in 99.99% of all homes today in one form or another. Your home Wifi, Bluetooth and DECT phone systems are prime examples of this technology. One concern my clients sometimes have is if they're considering using a half-duplex bridge solution the one thing they dont want is to keep saying 'over' when using VoIP. TDD works using a toggling system where the directions of the transmission are over a time interval. This toggling takes place very, very quickly and will not be noticed by the user. TDD can support voice and other symmetrical communication services as well as asymmetric data services. TDD can also deal with a dynamic mixture of both traffic types.

TDD uses the same frequency but transmits and receives at different times. This method is commonly used in cost-effective wireless bridging as unlike FDD (I'll get the in a sec) it doesn't need the additional costs associated with hardware and software needed to isolate the transmit and the receive antennas utilised in the FDD method. It's important to also understand that most manufacture's will give an 'aggregate' throughput. For example, an aggregate throughput is 170Mpbs, realistically this would be equal to 85Mbps in one direction.

  • It is more spectrum friendly, allowing the use of only a single frequency for operation and dramatically increasing spectrum utilization, especially in license-exempt or narrow-bandwidth frequency bands;
  • It allows for the variable allocation of throughput between the transmit and receive directions, making it well suited to applications with asymmetric traffic requirements, such as video surveillance, broadcast and Internet browsing;
  • Radios can be tuned for operation anywhere in a band and can be used at either end of the link. Therefore, only a single spare is required to serve both ends of a link.
  • The switch from transmit to receive incurs a delay that causes traditional TDD systems to have greater inherent latency than FDD systems;
  • Traditional TDD approaches yield poor TDM performance due to latency;
  • For symmetric traffic (50:50), TDD is less spectrally efficient than FDD, due to the switching time between transmit and receive; and Multiple co-located radios may interfere with one another unless they are synchronized.
FULL DUPLEX - FDD (Frequency-Division Duplex)

I didn't know people could do this but, I remember the first time I listened to my wife and her mum having a conversation. They were having two separate conversations at the same time I'm not even sure how this is possible! I can however understand FDD, well, its basics. Simply put full duplex is the networking term for describing the communication between two locations or devices in both directions simultaneously.

Unlike TDD the FDD method is achieved by using two different frequencies for transmit and receive. As briefly touched on earlier to achieve this, specific hardware and software are required to isolate both antennas which requires some substantial R&D. In wireless bridging this give numerous benefits and complexities while also adding a few pounds to the overall price tag. It's worth bearing in mind that in the wireless bridging world these abilities are only found in Millimetre technologies and frequencies. These solutions are typically enterprise or carrier class solutions specifically designed for commercial use.

  • The full data capacity is always available in each direction because the send and receive functions are separated;
  • It offers very low latency since transmit and receive functions operate simultaneously and continuously;
  • It can be used in licensed and license-exempt bands;
  • Most licensed bands worldwide are based on FDD; and Due to regulatory restrictions, FDD radios used in licensed bands are coordinated and protected from interference, though not immune to it.
  • Complex to install. Any given path requires the availability of a pair of frequencies; if either frequency in the pair is unavailable, then it may not be possible to deploy the system in that band;
  • Radios require pre-configured channel pairs, making sparing complex;
  • Any traffic allocation other than a 50:50 split between transmit and receive yields inefficient use of one of the two paired frequencies, lowering spectral efficiency; and Collocation of multiple radios is difficult.