Upside-Down World

The main track for 5G seems to be FDD for “old bands” below 3 GHz and TDD for “new bands” above 3 GHz (particularly mmWave frequencies). But physics advices us to the opposite:

  • At lower frequencies, larger areas are covered, thus most connections are likely to experience non-line-of-sight propagation. Since channel coherence is large (scales inverse proportionally to the Doppler), there is room for many terminals to transmit uplink pilots from which the base station consequently can obtain CSI. Reciprocity-based beamforming in TDD operation is scalable with respect to the number of base station antennas and delivers great value.
  • As the carrier frequency is increased, the coverage area shrinks; connections are more and more likely to experience line-of-sight propagation. At mmWave frequencies, all connections are either line-of-sight, or consist of a small number of reflected components. Then the channel can be parameterized with only few angular parameters; FDD operation with appropriate flavors of beam tracking may work satisfactorily. Reciprocity certainly would be desirable in this case too, but may not be necessary for the system to function.

Physics has given us the reciprocity principle. It should be exploited in wireless system design.

4 thoughts on “Upside-Down World”

  1. As mmwaves cannot penetrate walls and buildings, so we can have separate base stations for indoor and outdoor. If mmwave communication is always for short range then why in satellite communications we goto mmwave bands which cover much longer distance ?

    1. but also all existing cellular networks sub 3 GHz send most of their RF energy in the wrong direction which means they are neither power efficient or capacity efficient – also to achieve multiplexing efficiency you need RF bandwidth ratios that introduce additional inefficiencies – and antennas at 700 MHz in smart phones can have a negative gain of -10 dB. In contrast centimetre and millimetre band can deliver over 40 dBi of isotropic gain and much reduced delay spread. Military radio systems >70 GHz achieve 60 kilometre line of sight and free space loss is of the order of half a dB per kilometre – if you consider 5G as a progressive point to point system it makes a lot more sense economically and technically – also RF hardware can be potentially shared across mobile, back haul and satellite (mobile and fixed) and of course narrow band and wide band radar at 77 GHz – not so good when it rains but even this can be characterised and accommodated as fade margin in the link budget.

  2. Search for ‘coexistence TDD’ and you will find out why TDD may not have been the first choice of duplexing.

    1. Coexistence issues clearly need be carefully addressed but they are not fundamental technical problems.

Leave a Reply

Your email address will not be published. Required fields are marked *